2023-07-14 sql-vs-rd-pattern-analysis_executed.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b10ec5b6-efb7-4010-bede-f4f5fe9b1ce1",
   "metadata": {},
   "source": [
    "# Final analysis — SQL vs. Visual Diagrams for Matching Relational Query Patterns — June 2023 Study\n",
    "\n",
    "We are preregistering this study based on the [OSF Google Docs template](https://docs.google.com/document/d/1DaNmJEtBy04bq1l5OxS4JAscdZEkUGATURWwnBKLYxk/edit), which is one of [several preregistration templates](https://osf.io/zab38/wiki/home/?view) that [OSF](https://osf.io/) provides. Our experimental setup is inspired by [Leventidis et al. (2020)](https://doi.org/10.1145/3318464.3389767).\n",
    "\n",
    "$\\color{red}{\\text{\\textbf{See our updates post-registration}}}$ below in the [Other](#Other) section."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d3b92256-751a-4d54-9df8-8d0da7a961c2",
   "metadata": {},
   "source": [
    "## Study Information\n",
    "\n",
    "* **Title:**\n",
    "    The effect of SQL vs. Visual Diagrams on time and correctness matching relational query patterns\n",
    "\n",
    "* **Authors:** \n",
    "    *Anonymous for peer review. The online form on [osf.io](https://osf.io/) will list authors upon publication or embargo expiration.*\n",
    "\n",
    "* **Description:** \n",
    "    Pilot testing has indicated that visual diagrams (RD) improve participant speed at correctly identifying relational query patterns, contrasting with formatted SQL.\n",
    "    We will measure participant time and the proportion of correct answers for two conditions (RD and SQL) and 4 relational query patterns across 32 questions.\n",
    "\n",
    "* **Hypotheses:** We are testing for a total of `3` hypotheses: \n",
    "  * ***Time:***\n",
    "    * Let $\\theta_X$ denote the median time per question in seconds for a given condition $X$ per participant. We hypothesize that (1) $\\theta_{RD}$ / $\\theta_{SQL} < 1$, thus participants are relatively faster using `RD` compared to `SQL`.\n",
    "    * $\\color{red}{\\text{Corrected to match code \\& original intent:}}$ Let $\\theta_{X1}$ and $\\theta_{X2}$ denote the median time per question in seconds for a given condition $X$ per participant in the first $X1$ or second half $X2$. We hypothesize that (2) both $\\theta_{RD2} / \\theta_{RD1}$ and $\\theta_{SQL2} / \\theta_{SQL1} < 1$, thus participants are relatively faster in the 2nd half than the 1st.\n",
    "  * ***Correctness:*** Let $\\delta_X$ denote the mean proportion of correct responses for a given condition $X$. We hypothesize that (3) $\\delta_{RD}$ $\\simeq$ $\\delta_{SQL}$, i.e., participants make a comparable number of correct responses using `RD` or `SQL`.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "525cf552-1d10-4073-a2d9-2388f94aea72",
   "metadata": {},
   "source": [
    "## Design Plan\n",
    "\n",
    "* **Study type:**\n",
    "    Experiment.\n",
    "\n",
    "* **Blinding:**\n",
    "    No blinding is involved in this study.\n",
    "\n",
    "* **Study design:** \n",
    "    We have a within-subjects design, i.e., each participant sees questions in both of our modes: `RD` and `SQL`.\n",
    "    Each participant will see a total of `32` questions: `2` modes $\\times$ `4` patterns $\\times$ `2` instances in each half $\\times$ `2` halves.\n",
    "    I.e., they will see each pattern-mode combination `2` times per half.\n",
    "    For each question, the participant will be given a SQL query presented using one of the modes. They must choose the most appropriate of `4` natural-language interpretations of the query, each corresponding to one of our `4` patterns.\n",
    "    Their sequence number determines the mode presented to each participant for a given question — described under the randomization bullet.\n",
    "    The stimuli for each mode is:\n",
    "\n",
    "  * `SQL` — A conventional SQL representation with appropriate indentation.\n",
    "      The SQL text is indented and SQL keywords are color-coded appropriately.\n",
    "  * `RD` — A visual diagram we created of the query.\n",
    "\n",
    "    See the included `supplement/tutorial.pdf` file for a sample of the stimuli and how to read them, one page extracted here:\n",
    "\n",
    "![Example of the four patterns and two modes.](supplement/patterns.png)\n",
    "\n",
    "* **Randomization:** \n",
    "    To reduce ordering effects caused by which mode is presented first, we assign participants as they arrive to alternately start with `SQL` (sequence number `0`) or `RD` (sequence number `1`).\n",
    "    We then alternate the modes the presenter sees with each question.\n",
    "    I.e., `[SQL, RD, SQL, RD...]`.\n",
    "    We randomize the order that patterns are presented in each half separately, ensuring that each combination appears the same number of times and that both halves have the same number of each pattern-mode combination."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fbfb9ee8-3d98-441d-8594-fc6118b5f54c",
   "metadata": {},
   "source": [
    "## Sampling Plan\n",
    "\n",
    "* **Existing data:**\n",
    "    Registration before creating data: As of the date of submission of this research plan for preregistration, the data have not yet been collected, created, or realized.\n",
    "    Only pilot data has been collected.\n",
    "\n",
    "* **Explanation of existing data:**\n",
    "    N/A.\n",
    "  \n",
    "\n",
    "* **Data collection procedures:** \n",
    "  * ***Population:***\n",
    "    Participants will be recruited through [Amazon Mechanical Turk (AMT)](https://www.mturk.com/), a popular crowdsourcing website used for a variety of tasks, including academic studies.\n",
    "  * ***Recruitment efforts:*** A Human Intelligence Task (HIT) will be posted on AMT. We may repeatedly re-post the HIT as recruitment slows over time.\n",
    "  * ***Inclusion criteria:*** Participants could accept the AMT HIT if they are all of the following:\n",
    "    1. Residing in the USA and, thus, we expect English proficiency.\n",
    "    2. Adults (over 18 years old).\n",
    "    3. Experienced SQL users, self-determined with the prompt: \"Workers should be familiar with SQL at the level of an advanced undergraduate database class, in particular with nested SQL queries.\"\n",
    "    4. Have submitted over `500` approved HITs on AMT.\n",
    "    5. Have more than `97%` approved HIT assignments on AMT.\n",
    "  * ***Exclusion criteria:*** None. Pilot study participants were collected from our institution so did not need to be excluded on AMT.\n",
    "  * ***Timeline:*** Data will be collected from when we start until our **Stopping rule** below is met.\n",
    "  * ***Payment:***\n",
    "    * *AMT Rejection criteria:* A HIT will be accepted and the participant paid only if they correctly answered $\\ge$ `16/32` questions within `50` minutes. Otherwise, the HIT will be rejected.\n",
    "    * *Base pay:* `$6.00` USD.\n",
    "    * *Correctness bonus:* For every correctly answered question after the `16th` the participant receive a bonus payment of `$0.20` USD for a total pay of `$9.20` USD.\n",
    "    * *Time bonus:* Based on total test completion time, the participant will receive a percentage bonus on total pay (including the correctness bonus). Completion within `11` minutes awards a `5%` bonus for a maximum pay of `$9.66`. Each minute faster gets you an additional `5%` bonus up to `40%` for completing within `4` minutes, with a maximum pay of `$12.88`.  \n",
    "\n",
    "* **Sample size:** \n",
    "    Our target sample size is `50` participants.\n",
    "\n",
    "* **Sample size rationale:**\n",
    "    As all 13 pilot participants were faster with `RD` than `SQL`, we did not use a power analysis to choose the sample size. Instead, `50` was chosen as a meaningfully large and round number that is still a multiple of 2 to ensure that we have an equal number of participants among the sequences (see **Randomization**, above)\n",
    "\n",
    "* **Stopping rule:**\n",
    "    We will terminate data collection once our number of complete HITs has reached our maximum target sample size.\n",
    "    Given the strict ***Inclusion criteria*** in our study, it is possible that we won't be able to hit our target sample size.\n",
    "    In that case, we shall restrict our analysis to the data we can collect before paper submission.\n",
    "    We will continue collecting data until we reach the maximum target sample size or the camera-ready paper submission deadline."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fbd00368-c86e-4442-9dd3-7a7aa6646c6b",
   "metadata": {},
   "source": [
    "## Variables\n",
    "\n",
    "* **Manipulated Variables:**\n",
    "\n",
    "  * *Mode:* `[SQL, RD]`. See [Study Information](#Study-Information).\n",
    "\n",
    "\n",
    "* **Measured Variables:** \n",
    "\n",
    "  * For each participant and each question they answer, we record:\n",
    "    * ***Time (quantitative):*** The time they take to answer the question.\n",
    "    * ***Pattern (categorical/integer):*** The pattern they were provided in one of the modes for that question.\n",
    "    * ***Choice (integer):*** Their selected pattern from the `4`-option multiple choice question.\n",
    "    * ***Correct (boolean/integer):*** Whether their answer for the `4`-option multiple choice question was correct.\n",
    "\n",
    "  * For each participant, we also record:  \n",
    "    * ***Sequence (categorical):*** The sequence the participant was randomly assigned to (see [Design Plan](#Design-Plan)).\n",
    "    * ***Free-text feedback (string):***  The participant's optional answers to a feedback prompt.\n",
    "\n",
    "\n",
    "* **Indices:** From our collected study data, we will calculate:\n",
    "  * ***Indices for time***\n",
    "     \n",
    "    * **`Median time per mode per participant`:**\n",
    "    This is calculated by taking the median of the `16` ***Time*** records for each mode for each participant.\n",
    "    Using that information, we can calculate the median time across all participants.\n",
    "\n",
    "    * **`Ratio of median time of RD relative to SQL per participant:`**\n",
    "    Using the `Median time per mode per participant`, we will calculate `RD/SQL`.\n",
    "\n",
    "    * **`Quartiles and CIs of median time per mode across all participants:`**\n",
    "    Using the `Median time per mode per participant`, we will calculate the 1st, 2nd (median), 3rd quartiles.\n",
    "\n",
    "    * **`Quartiles and CIs of median time of RD relative to SQL across all participants:`**\n",
    "    Using the `Ratio of median time of RD relative to SQL per participant`, we will calculate the 1st, 2nd (median), 3rd quartiles as well as 95% Confidence Intervals using Bias Corrected and Accelerated (BCa) bootstrapping ([Efron (1987)](https://doi.org/10.2307%2F2289144)).\n",
    "\n",
    "    * **`Per-half indices:`**\n",
    "    We will compute both the `Quartiles and CIs of median time...` indices for each half.\n",
    "        \n",
    "  * ***Indices for correctness***\n",
    "\n",
    "    * **`Proportion of correct responses per mode per participant`:**\n",
    "    For each participant, calculate the proportion of correct responses per mode as: `correct responses / total questions per mode`.\n",
    "    \n",
    "    * **`Mean and CIs of proportion of correct responses per mode across all participants`:**\n",
    "    The mean proportion of correct responses across all participants is calculated by taking the arithmetic mean of all the `Proportion of correct responses per mode per participant` values for a given mode across all participants.\n",
    "    We will calculate 95% Confidence Intervals using Bias Corrected and Accelerated (BCa) bootstrapping."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e9d2853-0e21-469a-96a4-316c25dd9e42",
   "metadata": {},
   "source": [
    "## Analysis Plan\n",
    "\n",
    "* **Statistical Models:**\n",
    "\n",
    "  * ***Distribution testing:***\n",
    "    We will examine the distributions of our data visually, including for each mode, to ensure there are no problematic distributions.\n",
    "  * ***Hypothesis Testing:***\n",
    "    We will visually examine the 95% BCa bootstrapped confidence intervals for each mean and median.\n",
    "\n",
    "* **Transformations:** N/A.\n",
    "\n",
    "* **Inference criteria:**\n",
    "    We will interpret the results using Interval Estimation rather than dichotomous $p$-value cutoffs (e.g., we will not use *p*<.05 to determine statistical significance).\n",
    "    See [Dragicevik (2016)](http://dx.doi.org/10.1007/978-3-319-26633-6_13) and [Besançon & Dragicevic (2019)](http://dx.doi.org/10.1145/3290607.3310432) for a discussion of using estimation for these types of analyses.\n",
    "    \n",
    "* **Data exclusion:**\n",
    "    To perform a concrete analysis of our data, we would like to minimize the set of outlier points as they will negatively affect the quality of our statistical analysis and introduce unwanted/non-existent bias.\n",
    "    After collecting our data, we will examine the time distribution of each worker. In particular, we will examine how long each worker took to answer the question on a per-question basis. We expect two types of outlier points in our experiments. \n",
    "  1. **Speeders:** Workers that answered a question much faster than the vast majority of participants (usually in the order of a few seconds, i.e., workers tried to rush answering each question without thinking).\n",
    "      The range could vary, but a rough indication of a speeder would be if their time per question is 2-3 standard deviations lower than the mean time per question.\n",
    "  2. **Unusually delayed answers:** This refers to workers who took unusually long to answer a question.\n",
    "      This is most likely attributed to some distraction that made the worker not focus on our question while the timer was running (i.e., a phone call, text message, bathroom break, etc.).\n",
    "      As an online test, we can't know exactly what was the cause of it, but usually, we can identify such data points by noticing their much larger value in time.\n",
    "      Since we capture a time distribution, it is expected to be left-skewed, and thus a rough measure of an *unusually delayed answer* would be about $\\geq 3$ times the mean time per question.\n",
    "  3. **Cheaters:** Previous studies have indicated that workers can leak answers to enable other workers to answer all the questions correctly and quickly. We have used technical measures to limit this possibility and give each user different sets of stimuli. However, if we identify cheaters through log analysis, we will exclude them.\n",
    "  4. **Median:** To minimize the effect of outliers for all the above reasons and to provide a statistically valid unbiased estimator for the ratio of times, we use the median instead of the mean aggregation for time analysis.\n",
    "\n",
    "\n",
    "* **Missing data:**\n",
    "    We will unlikely have missing data because for a participant to submit their results, they must answer all the questions.\n",
    "    However, if we have missing data points from an individual, we will remove the individual completely from our analysis."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ebff1220-4844-4180-ab77-1b7c128a6f88",
   "metadata": {},
   "source": [
    "## Other\n",
    "\n",
    "* **Discrepancies between preregistration prose and analysis code:**\n",
    "    The intent of our study design is explained in this section.\n",
    "    In case of any discrepancy between the analysis code below and this section, what is written in this section takes precedence for the preregistration.\n",
    "\n",
    "* $\\color{red}{\\text{\\textbf{Updates post-registration:}}}$\n",
    "\n",
    "    1. Our second hypothesis was incorrectly stated above, but our intent was clear from the below text and code. We have updated it. This result should be evaluated accordingly as it is a deviation from our prergistration.\n",
    "    2. Minor error in correctness score calculation fixed.\n",
    "    3. Anonymization of MTurk worker IDs is removed from this code and now done outside this worksheet to avoid accidental release of worker IDs.\n",
    "    4. Time spent on tutorials was erroneously collected, and should not be used. We remove this column in the anonymization code.\n",
    "    5. In total, 177 participants began the study, but many quit before finishing the tutorial or after a few questions. Only 120 participants submitted the HIT. Of those, only 58 reached the 50% correctness threshold for HIT acceptance. We only select the first 50 of those 58 that were submitted to be in accordance with our preregistration.\n",
    "    6. Added visual emphasis for ratio = 1 in figure 1b.\n",
    "    7. Added user feedback printing at the end of the worksheet.\n",
    "    8. Added \"variants\" for further exploratory analysis based on elevated correctness thresholds.\n",
    "    9. Added per-pattern exploratory analysis.\n",
    "    10. Simplified the code for figure 3.\n",
    "    11. Added additional figure showing difference for question 4.\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a58c448a",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "# Scripts to analyze the study data\n",
    "\n",
    "(Q1) TIMING PER PARTICIPANT (SQL vs RD)\n",
    "1. Per participant, calculate the median time in seconds spent on SQL and RD (32/2=16 per mode and participant, irrespective of correctness)\n",
    "2. Q1a: show violin plot figure with median times per user compared via gray lines\n",
    "3. Calculate their ratio per user (also gives fraction of users who are faster with one or the other)\n",
    "4. Calculate the median of those fractions and the 95% CI\n",
    "5. Q1b: show violin plot figure with fractions, and also 95% CI\n",
    "\n",
    "(Q2) TIMING PER PARTICIPANT (SQL vs RD / 1st vs. 2nd half)\n",
    "1. Per participant, calculate the median time over all questions answered in 1st half in RD (32/2/2=8) and SQL, and in 2nd half.\n",
    "2. Q2: show repeated measure violin plot figure, showing improvements over time, of 2 halfs\n",
    "3. Calculate the relative ratio for timing 2nd/1st for RD, and SQL including 95% CI\n",
    "\n",
    "(Q3) TIMING PATTERNS ACROSS PARTICIPANTS\n",
    "1. calculate the median time per pattern (4) across the two modes (2). Thus 8 values.\n",
    "2. show repeated measure violin plot figure\n",
    "\n",
    "(Q4) CORRECTNESS (SQL vs RD)\n",
    "1. take mean correct over all questions and all users answered in SQL (32/2*13), or in RD (2 values)\n",
    "2. calculate 95% CI for each, and sampled p-value (perhaps with difference?)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11486b0f",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "## Load packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "90163bd4-f66f-4d4d-a525-a8420532f03e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "sns.set(style=\"whitegrid\",font_scale=2)\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.stats import bootstrap as scipybootstrap\n",
    "from IPython.display import display\n",
    "import matplotlib.ticker as mtick           # allows change to percentage\n",
    "\n",
    "# Tell matplotlib to export svg text as text not paths\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "plt.rcParams['axes.axisbelow'] = True   # draw axes and grids behind everything else\n",
    "\n",
    "# Set Jupyter and Pandas to show 3 decimal places, does not work for lists of numbers\n",
    "%precision 3\n",
    "pd.options.display.float_format = '{:,.3f}'.format\n",
    "np.set_printoptions(precision=3)\n",
    "# np.set_printoptions(formatter={'float': lambda x: \"{0:0.3f}\".format(x)})            # TODO: does not work for lists\n",
    "def print(*args):\n",
    "    __builtins__.print(*(\"%.3f\" % a if isinstance(a, float) else a\n",
    "                         for a in args))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60109a40-4d08-4bd6-90ed-a6788cfe728a",
   "metadata": {},
   "source": [
    "## Global Variables Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "2d1d96cd-4678-4d54-a8eb-d84cf3e14247",
   "metadata": {},
   "outputs": [],
   "source": [
    "# A set of constant global variables used throughout the notebook\n",
    "num_questions = 32\n",
    "modes = ['SQL', 'RD']\n",
    "mode_to_name = {0: 'SQL', 1: 'RD'}\n",
    "\n",
    "# anonymizeddata = 'data/users-table-pilot.csv'                      # pilot\n",
    "anonymizeddata = 'data/users-table-anonymized.csv'\n",
    "transformeddata = 'data/transformed_data.csv'   # file with appropriately transformed data ready for analysis\n",
    "\n",
    "BOOTSTRAPCONFIDENCE = 0.95      # confidence level used for bootstrap\n",
    "BOOTSTRAPMETHOD = 'BCa'         # method used for bootstrap, appears to be better than the textbook version for mean (but not for median), also available as 'percentage'\n",
    "BOOTSTRAPSAMPLES = 10000        # number of resamples\n",
    "VARIANT = 1                     # variant 1: all participants, variant 2: only for correctness = 1.0, variant 3: only for correctness = 0.9, variant 4: only for correctness >= 0.66"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "220858af",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "## Define subfolder where figures are stored\n",
    "\n",
    "By default, figures will not be saved. If you want to save figures, set savefig to `True`.\n",
    "Learned from: https://github.com/jorvlan/open-visualizations/blob/master/Python/tutorial_2/repeated_measures_python_2.ipynb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "80874ad0",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Directory already exists\n"
     ]
    }
   ],
   "source": [
    "savefig = True\n",
    "if savefig:\n",
    "    import os\n",
    "    from os.path import isdir\n",
    "    cwd = os.getcwd()   # Get current working directory, but you can specify your own directory of course.\n",
    "    if  os.path.exists(cwd + \"/figs\"):\n",
    "        print(\"Directory already exists\")\n",
    "        fig_dir = cwd + \"/figs\"     # Assign the existing directory to a variable\n",
    "    elif not os.path.exists(cwd + \"/figs\"):\n",
    "        print(\"Directory does not exist and will be created ......\")\n",
    "        os.makedirs(cwd + \"/figs\")\n",
    "        if isdir(cwd + \"/figs\"):\n",
    "            print('Directory was created succesfully')\n",
    "        fig_dir = cwd + \"/figs\"     # Assign the created directory to a variable\n",
    "    else:\n",
    "        print(\"Something went wrong\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce8b86e6-d818-4a15-a5f4-9ca691e84fba",
   "metadata": {},
   "source": [
    "## Loading full data, transforming it, and saving the transformed version\n",
    "\n",
    "Loading the full data, transforming it to make available for later analysis, and saving it\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "385495ea-f2c9-4273-9cd9-4e2378417641",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>worker_id</th>\n",
       "      <th>assignment_id</th>\n",
       "      <th>hit_id</th>\n",
       "      <th>qualification_score</th>\n",
       "      <th>current_section</th>\n",
       "      <th>current_page</th>\n",
       "      <th>sequence_num</th>\n",
       "      <th>pattern_order</th>\n",
       "      <th>start_datetime</th>\n",
       "      <th>end_datetime</th>\n",
       "      <th>...</th>\n",
       "      <th>q30_time</th>\n",
       "      <th>q31</th>\n",
       "      <th>q31_start</th>\n",
       "      <th>q31_end</th>\n",
       "      <th>q31_time</th>\n",
       "      <th>q32</th>\n",
       "      <th>q32_start</th>\n",
       "      <th>q32_end</th>\n",
       "      <th>q32_time</th>\n",
       "      <th>feedback</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>42</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RESULTS</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000</td>\n",
       "      <td>[1, 1, 2, 1, 1, 2, 3, 4, 2, 4, 4, 2, 4, 3, 3, ...</td>\n",
       "      <td>2023-07-06 15:33:53.819738</td>\n",
       "      <td>2023-07-06 15:56:48.16091</td>\n",
       "      <td>...</td>\n",
       "      <td>13,226.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>2023-07-06 15:54:44.186955</td>\n",
       "      <td>2023-07-06 15:54:57.575035</td>\n",
       "      <td>13,388.000</td>\n",
       "      <td>4.000</td>\n",
       "      <td>2023-07-06 15:55:00.530759</td>\n",
       "      <td>2023-07-06 15:55:13.292598</td>\n",
       "      <td>12,761.000</td>\n",
       "      <td>I found the tutorial to be very helpful in und...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>147</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RESULTS</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000</td>\n",
       "      <td>[3, 1, 3, 3, 2, 2, 4, 4, 2, 1, 1, 2, 1, 3, 4, ...</td>\n",
       "      <td>2023-07-06 15:35:09.753802</td>\n",
       "      <td>2023-07-06 15:47:48.277039</td>\n",
       "      <td>...</td>\n",
       "      <td>19,847.000</td>\n",
       "      <td>4.000</td>\n",
       "      <td>2023-07-06 15:47:19.25664</td>\n",
       "      <td>2023-07-06 15:47:24.331627</td>\n",
       "      <td>5,074.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-06 15:47:26.069748</td>\n",
       "      <td>2023-07-06 15:47:33.66475</td>\n",
       "      <td>7,595.000</td>\n",
       "      <td>good</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>148</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RESULTS</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000</td>\n",
       "      <td>[1, 3, 3, 2, 3, 2, 2, 4, 2, 4, 4, 3, 4, 1, 1, ...</td>\n",
       "      <td>2023-07-07 18:11:52.077358</td>\n",
       "      <td>2023-07-07 18:37:30.385374</td>\n",
       "      <td>...</td>\n",
       "      <td>16,843.000</td>\n",
       "      <td>3.000</td>\n",
       "      <td>2023-07-07 18:36:52.45584</td>\n",
       "      <td>2023-07-07 18:36:57.673103</td>\n",
       "      <td>5,217.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-07 18:37:00.669486</td>\n",
       "      <td>2023-07-07 18:37:12.637864</td>\n",
       "      <td>11,968.000</td>\n",
       "      <td>Good</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>19</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RESULTS</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000</td>\n",
       "      <td>[3, 2, 4, 4, 2, 1, 1, 3, 2, 3, 4, 2, 1, 1, 3, ...</td>\n",
       "      <td>2023-07-06 15:33:42.760353</td>\n",
       "      <td>2023-07-06 15:52:40.496031</td>\n",
       "      <td>...</td>\n",
       "      <td>5,311.000</td>\n",
       "      <td>3.000</td>\n",
       "      <td>2023-07-06 15:51:03.003979</td>\n",
       "      <td>2023-07-06 15:51:20.250417</td>\n",
       "      <td>17,246.000</td>\n",
       "      <td>4.000</td>\n",
       "      <td>2023-07-06 15:51:22.991053</td>\n",
       "      <td>2023-07-06 15:51:28.173166</td>\n",
       "      <td>5,182.000</td>\n",
       "      <td>Nothing particular. I enjoyed the task very mu...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>161</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RESULTS</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000</td>\n",
       "      <td>[2, 3, 4, 2, 2, 1, 1, 4, 3, 3, 1, 1, 3, 4, 4, ...</td>\n",
       "      <td>2023-07-07 18:11:09.874051</td>\n",
       "      <td>2023-07-07 18:58:04.501641</td>\n",
       "      <td>...</td>\n",
       "      <td>3,524.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-07 18:40:46.153988</td>\n",
       "      <td>2023-07-07 18:40:49.968978</td>\n",
       "      <td>3,814.000</td>\n",
       "      <td>3.000</td>\n",
       "      <td>2023-07-07 18:40:53.382182</td>\n",
       "      <td>2023-07-07 18:40:56.839762</td>\n",
       "      <td>3,457.000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>57</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RESULTS</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000</td>\n",
       "      <td>[3, 3, 4, 2, 1, 1, 3, 2, 1, 4, 4, 3, 2, 1, 2, ...</td>\n",
       "      <td>2023-07-07 18:11:49.878986</td>\n",
       "      <td>2023-07-07 18:22:30.45823</td>\n",
       "      <td>...</td>\n",
       "      <td>5,306.000</td>\n",
       "      <td>4.000</td>\n",
       "      <td>2023-07-07 18:22:06.119203</td>\n",
       "      <td>2023-07-07 18:22:11.094555</td>\n",
       "      <td>4,975.000</td>\n",
       "      <td>4.000</td>\n",
       "      <td>2023-07-07 18:22:12.775153</td>\n",
       "      <td>2023-07-07 18:22:18.426053</td>\n",
       "      <td>5,650.000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>166</th>\n",
       "      <td>62</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RESULTS</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000</td>\n",
       "      <td>[4, 4, 3, 3, 4, 4, 3, 2, 1, 1, 2, 2, 1, 1, 2, ...</td>\n",
       "      <td>2023-07-06 16:05:50.918429</td>\n",
       "      <td>2023-07-06 16:18:44.773885</td>\n",
       "      <td>...</td>\n",
       "      <td>5,546.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>2023-07-06 16:18:14.652581</td>\n",
       "      <td>2023-07-06 16:18:25.846863</td>\n",
       "      <td>11,194.000</td>\n",
       "      <td>4.000</td>\n",
       "      <td>2023-07-06 16:18:27.91728</td>\n",
       "      <td>2023-07-06 16:18:32.772318</td>\n",
       "      <td>4,855.000</td>\n",
       "      <td>none</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RESULTS</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000</td>\n",
       "      <td>[3, 3, 3, 4, 2, 2, 4, 3, 1, 4, 4, 1, 2, 1, 1, ...</td>\n",
       "      <td>2023-07-07 19:01:38.084371</td>\n",
       "      <td>2023-07-07 19:19:47.324967</td>\n",
       "      <td>...</td>\n",
       "      <td>7,472.000</td>\n",
       "      <td>3.000</td>\n",
       "      <td>2023-07-07 19:19:10.726193</td>\n",
       "      <td>2023-07-07 19:19:32.963664</td>\n",
       "      <td>22,237.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-07 19:19:34.721309</td>\n",
       "      <td>2023-07-07 19:19:39.613308</td>\n",
       "      <td>4,891.000</td>\n",
       "      <td>good</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>28</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RESULTS</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000</td>\n",
       "      <td>[4, 3, 1, 4, 1, 4, 2, 2, 4, 3, 3, 2, 2, 1, 3, ...</td>\n",
       "      <td>2023-07-06 15:34:02.551439</td>\n",
       "      <td>2023-07-06 16:19:00.534461</td>\n",
       "      <td>...</td>\n",
       "      <td>12,782.000</td>\n",
       "      <td>4.000</td>\n",
       "      <td>2023-07-06 16:18:21.318188</td>\n",
       "      <td>2023-07-06 16:18:31.224703</td>\n",
       "      <td>9,906.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>2023-07-06 16:18:31.963803</td>\n",
       "      <td>2023-07-06 16:18:39.588003</td>\n",
       "      <td>7,624.000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>170</th>\n",
       "      <td>163</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RESULTS</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000</td>\n",
       "      <td>[4, 2, 4, 4, 1, 1, 1, 4, 2, 1, 2, 2, 3, 3, 3, ...</td>\n",
       "      <td>2023-07-07 18:44:54.269601</td>\n",
       "      <td>2023-07-07 19:01:59.264466</td>\n",
       "      <td>...</td>\n",
       "      <td>3,067.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>2023-07-07 19:00:31.777433</td>\n",
       "      <td>2023-07-07 19:00:36.612726</td>\n",
       "      <td>4,835.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-07 19:00:38.045936</td>\n",
       "      <td>2023-07-07 19:00:40.864742</td>\n",
       "      <td>2,818.000</td>\n",
       "      <td>GOODNESS</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>133 rows × 152 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     worker_id  assignment_id  hit_id  qualification_score current_section  \\\n",
       "0           42            NaN     NaN                  NaN         RESULTS   \n",
       "1          147            NaN     NaN                  NaN         RESULTS   \n",
       "3          148            NaN     NaN                  NaN         RESULTS   \n",
       "5           19            NaN     NaN                  NaN         RESULTS   \n",
       "6          161            NaN     NaN                  NaN         RESULTS   \n",
       "..         ...            ...     ...                  ...             ...   \n",
       "165         57            NaN     NaN                  NaN         RESULTS   \n",
       "166         62            NaN     NaN                  NaN         RESULTS   \n",
       "168          6            NaN     NaN                  NaN         RESULTS   \n",
       "169         28            NaN     NaN                  NaN         RESULTS   \n",
       "170        163            NaN     NaN                  NaN         RESULTS   \n",
       "\n",
       "     current_page  sequence_num  \\\n",
       "0               1         1.000   \n",
       "1               1         1.000   \n",
       "3               1         1.000   \n",
       "5               1         0.000   \n",
       "6               1         1.000   \n",
       "..            ...           ...   \n",
       "165             1         1.000   \n",
       "166             1         0.000   \n",
       "168             1         0.000   \n",
       "169             1         1.000   \n",
       "170             1         1.000   \n",
       "\n",
       "                                         pattern_order  \\\n",
       "0    [1, 1, 2, 1, 1, 2, 3, 4, 2, 4, 4, 2, 4, 3, 3, ...   \n",
       "1    [3, 1, 3, 3, 2, 2, 4, 4, 2, 1, 1, 2, 1, 3, 4, ...   \n",
       "3    [1, 3, 3, 2, 3, 2, 2, 4, 2, 4, 4, 3, 4, 1, 1, ...   \n",
       "5    [3, 2, 4, 4, 2, 1, 1, 3, 2, 3, 4, 2, 1, 1, 3, ...   \n",
       "6    [2, 3, 4, 2, 2, 1, 1, 4, 3, 3, 1, 1, 3, 4, 4, ...   \n",
       "..                                                 ...   \n",
       "165  [3, 3, 4, 2, 1, 1, 3, 2, 1, 4, 4, 3, 2, 1, 2, ...   \n",
       "166  [4, 4, 3, 3, 4, 4, 3, 2, 1, 1, 2, 2, 1, 1, 2, ...   \n",
       "168  [3, 3, 3, 4, 2, 2, 4, 3, 1, 4, 4, 1, 2, 1, 1, ...   \n",
       "169  [4, 3, 1, 4, 1, 4, 2, 2, 4, 3, 3, 2, 2, 1, 3, ...   \n",
       "170  [4, 2, 4, 4, 1, 1, 1, 4, 2, 1, 2, 2, 3, 3, 3, ...   \n",
       "\n",
       "                 start_datetime                end_datetime  ...   q30_time  \\\n",
       "0    2023-07-06 15:33:53.819738   2023-07-06 15:56:48.16091  ... 13,226.000   \n",
       "1    2023-07-06 15:35:09.753802  2023-07-06 15:47:48.277039  ... 19,847.000   \n",
       "3    2023-07-07 18:11:52.077358  2023-07-07 18:37:30.385374  ... 16,843.000   \n",
       "5    2023-07-06 15:33:42.760353  2023-07-06 15:52:40.496031  ...  5,311.000   \n",
       "6    2023-07-07 18:11:09.874051  2023-07-07 18:58:04.501641  ...  3,524.000   \n",
       "..                          ...                         ...  ...        ...   \n",
       "165  2023-07-07 18:11:49.878986   2023-07-07 18:22:30.45823  ...  5,306.000   \n",
       "166  2023-07-06 16:05:50.918429  2023-07-06 16:18:44.773885  ...  5,546.000   \n",
       "168  2023-07-07 19:01:38.084371  2023-07-07 19:19:47.324967  ...  7,472.000   \n",
       "169  2023-07-06 15:34:02.551439  2023-07-06 16:19:00.534461  ... 12,782.000   \n",
       "170  2023-07-07 18:44:54.269601  2023-07-07 19:01:59.264466  ...  3,067.000   \n",
       "\n",
       "      q31                   q31_start                     q31_end   q31_time  \\\n",
       "0   2.000  2023-07-06 15:54:44.186955  2023-07-06 15:54:57.575035 13,388.000   \n",
       "1   4.000   2023-07-06 15:47:19.25664  2023-07-06 15:47:24.331627  5,074.000   \n",
       "3   3.000   2023-07-07 18:36:52.45584  2023-07-07 18:36:57.673103  5,217.000   \n",
       "5   3.000  2023-07-06 15:51:03.003979  2023-07-06 15:51:20.250417 17,246.000   \n",
       "6   1.000  2023-07-07 18:40:46.153988  2023-07-07 18:40:49.968978  3,814.000   \n",
       "..    ...                         ...                         ...        ...   \n",
       "165 4.000  2023-07-07 18:22:06.119203  2023-07-07 18:22:11.094555  4,975.000   \n",
       "166 2.000  2023-07-06 16:18:14.652581  2023-07-06 16:18:25.846863 11,194.000   \n",
       "168 3.000  2023-07-07 19:19:10.726193  2023-07-07 19:19:32.963664 22,237.000   \n",
       "169 4.000  2023-07-06 16:18:21.318188  2023-07-06 16:18:31.224703  9,906.000   \n",
       "170 2.000  2023-07-07 19:00:31.777433  2023-07-07 19:00:36.612726  4,835.000   \n",
       "\n",
       "      q32                   q32_start                     q32_end   q32_time  \\\n",
       "0   4.000  2023-07-06 15:55:00.530759  2023-07-06 15:55:13.292598 12,761.000   \n",
       "1   1.000  2023-07-06 15:47:26.069748   2023-07-06 15:47:33.66475  7,595.000   \n",
       "3   1.000  2023-07-07 18:37:00.669486  2023-07-07 18:37:12.637864 11,968.000   \n",
       "5   4.000  2023-07-06 15:51:22.991053  2023-07-06 15:51:28.173166  5,182.000   \n",
       "6   3.000  2023-07-07 18:40:53.382182  2023-07-07 18:40:56.839762  3,457.000   \n",
       "..    ...                         ...                         ...        ...   \n",
       "165 4.000  2023-07-07 18:22:12.775153  2023-07-07 18:22:18.426053  5,650.000   \n",
       "166 4.000   2023-07-06 16:18:27.91728  2023-07-06 16:18:32.772318  4,855.000   \n",
       "168 1.000  2023-07-07 19:19:34.721309  2023-07-07 19:19:39.613308  4,891.000   \n",
       "169 2.000  2023-07-06 16:18:31.963803  2023-07-06 16:18:39.588003  7,624.000   \n",
       "170 1.000  2023-07-07 19:00:38.045936  2023-07-07 19:00:40.864742  2,818.000   \n",
       "\n",
       "                                              feedback  \n",
       "0    I found the tutorial to be very helpful in und...  \n",
       "1                                                 good  \n",
       "3                                                 Good  \n",
       "5    Nothing particular. I enjoyed the task very mu...  \n",
       "6                                                  NaN  \n",
       "..                                                 ...  \n",
       "165                                                NaN  \n",
       "166                                               none  \n",
       "168                                               good  \n",
       "169                                                NaN  \n",
       "170                                           GOODNESS  \n",
       "\n",
       "[133 rows x 152 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# --- Load anonymized full study data\n",
    "df = pd.read_csv(anonymizeddata)\n",
    "\n",
    "# --- Filter on 'current_section=RESULTS'\n",
    "dfresults = df.loc[(df.current_section == \"RESULTS\")].copy()              # (7/6/2023: added filter to only focus on RESULTS)\n",
    "\n",
    "# --- Turn string to array\n",
    "from ast import literal_eval            # to turn string to array\n",
    "dfresults['pattern_order']= dfresults['pattern_order'].apply(literal_eval)\n",
    "\n",
    "# display(dfresults)\n",
    "# The \"current page\" is the section of the study the workers are doing to save their state & prevent them cheating\n",
    "\n",
    "# --- The following code block transforms the data frame to have one question per row. That simplifies the later analysis.\n",
    "# reshape dfresults (melt, pivot) to bring multiple question times (e.g. 'q7_time') per row into separate rows\n",
    "# https://towardsdatascience.com/wide-to-long-data-how-and-when-to-use-pandas-melt-stack-and-wide-to-long-7c1e0f462a98\n",
    "df2 = dfresults.melt(id_vars=['worker_id', 'sequence_num', 'pattern_order',\n",
    "                       'q1', 'q2','q3', 'q4','q5', 'q6','q7', 'q8', 'q9', 'q10',\n",
    "                       'q11', 'q12','q13', 'q14','q15', 'q16','q17', 'q18', 'q19', 'q20',\n",
    "                       'q21', 'q22','q23', 'q24','q25', 'q26','q27', 'q28', 'q29', 'q30',\n",
    "                       'q31', 'q32'], value_vars=['q1_time', 'q2_time', 'q3_time', 'q4_time','q5_time', 'q6_time', 'q7_time', 'q8_time', 'q9_time', 'q10_time',\n",
    "                                                  'q11_time', 'q12_time', 'q13_time', 'q14_time', 'q15_time', 'q16_time', 'q17_time', 'q18_time', 'q19_time', 'q20_time',\n",
    "                                                  'q21_time', 'q22_time', 'q23_time', 'q24_time', 'q25_time', 'q26_time', 'q27_time', 'q28_time', 'q29_time', 'q30_time',\n",
    "                                                  'q31_time', 'q32_time'], var_name='question', value_name='time')\n",
    "\n",
    "# replace time in msec with sec in column 'time'\n",
    "df2['time'] = df2['time'] / 1000\n",
    "\n",
    "# replace question string 'q7_time' with number '7' in column 'question'\n",
    "from re import search as re_search               # regular expression\n",
    "new_column = []\n",
    "for values in df2['question']:\n",
    "    new_column.append(int(re_search(r'\\d+', values).group()))\n",
    "df2['question'] = new_column\n",
    "\n",
    "# choose the right pattern from the list 'pattern_order' and add as column 'pattern'\n",
    "new_column = []\n",
    "for (pattern_order_list, ind) in zip(df2['pattern_order'], df2['question']):\n",
    "    new_column.append(pattern_order_list[ind-1])\n",
    "df2['pattern'] = new_column\n",
    "\n",
    "# determine the 'mode' (SQL or RD) from 'sequence_num' and 'question'\n",
    "#   sequence_num = 0 means that the first question is shown in SQL, 1 means we start instead with RD. Then alternate between the two.\n",
    "#   Thus (sequence_num + question_num) % 2 == 1 means SQL\n",
    "#   Thus (sequence_num + question_num) % 2 == 0 means RD\n",
    "new_column = []\n",
    "for (sequence, question) in zip(df2['sequence_num'], df2['question']):\n",
    "    mode = 'SQL' if (sequence + question) % 2 == 1 else 'RD'\n",
    "    new_column.append(mode)\n",
    "df2['mode'] = new_column\n",
    "\n",
    "# determine the 'choice' (among the 4 patterns) made by the user for this question. Requires all the 32 question choices (e.g. 'q7') and index of the question at hand ('question')\n",
    "questionarray = df2[['q1', 'q2','q3', 'q4','q5', 'q6','q7', 'q8', 'q9', 'q10',\n",
    "                     'q11', 'q12','q13', 'q14','q15', 'q16','q17', 'q18', 'q19', 'q20',\n",
    "                     'q21', 'q22','q23', 'q24','q25', 'q26','q27', 'q28', 'q29', 'q30',\n",
    "                     'q31', 'q32']].to_numpy()\n",
    "questionindex = df2[[\"question\"]].to_numpy()\n",
    "\n",
    "new_array = np.take_along_axis(questionarray,questionindex-1,1)     # take the 'questionindex'-th entry from each row of the questionarray (notice 1-index vs 0-indexin)\n",
    "df2['choice'] = new_array\n",
    "\n",
    "# determine whether the choice was correct by comparing the ground truth ('pattern') against the choice made ('choice'). Saved as 0/1 value in new column 'correct'\n",
    "new_column = []\n",
    "for (pattern, choice) in zip(df2['pattern'], df2['choice']):\n",
    "    correct = 1 if pattern == choice else 0\n",
    "    new_column.append(correct)\n",
    "df2['correct'] = new_column\n",
    "\n",
    "# sort by worker and question number, and reset the inde\n",
    "df2.sort_values(by=['worker_id', 'question'], inplace=True)\n",
    "df2.reset_index(drop=True, inplace=True)\n",
    "# display(df2)\n",
    "\n",
    "# select only the relevant subset of columns\n",
    "df_transformed_data = df2[['worker_id', 'question', 'time', 'pattern', 'mode', 'choice', 'correct']]\n",
    "# display(df3)\n",
    "\n",
    "# pd.write_csv(filename)\n",
    "df_transformed_data.to_csv(transformeddata,\n",
    "                           index=False,\n",
    "                           )\n",
    "display(dfresults)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eaeee9a1",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%% md\n"
    }
   },
   "source": [
    "## Loading transformed data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e29e9e0e",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>worker_id</th>\n",
       "      <th>question</th>\n",
       "      <th>time</th>\n",
       "      <th>pattern</th>\n",
       "      <th>mode</th>\n",
       "      <th>choice</th>\n",
       "      <th>correct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>139.259</td>\n",
       "      <td>2</td>\n",
       "      <td>SQL</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>10.359</td>\n",
       "      <td>2</td>\n",
       "      <td>RD</td>\n",
       "      <td>2.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>41.813</td>\n",
       "      <td>4</td>\n",
       "      <td>SQL</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>8.190</td>\n",
       "      <td>4</td>\n",
       "      <td>RD</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>80.589</td>\n",
       "      <td>3</td>\n",
       "      <td>SQL</td>\n",
       "      <td>3.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4251</th>\n",
       "      <td>169</td>\n",
       "      <td>28</td>\n",
       "      <td>4.797</td>\n",
       "      <td>2</td>\n",
       "      <td>RD</td>\n",
       "      <td>2.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4252</th>\n",
       "      <td>169</td>\n",
       "      <td>29</td>\n",
       "      <td>4.589</td>\n",
       "      <td>4</td>\n",
       "      <td>SQL</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4253</th>\n",
       "      <td>169</td>\n",
       "      <td>30</td>\n",
       "      <td>4.384</td>\n",
       "      <td>4</td>\n",
       "      <td>RD</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4254</th>\n",
       "      <td>169</td>\n",
       "      <td>31</td>\n",
       "      <td>5.703</td>\n",
       "      <td>4</td>\n",
       "      <td>SQL</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4255</th>\n",
       "      <td>169</td>\n",
       "      <td>32</td>\n",
       "      <td>4.874</td>\n",
       "      <td>4</td>\n",
       "      <td>RD</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4256 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      worker_id  question    time  pattern mode  choice  correct\n",
       "0             0         1 139.259        2  SQL   1.000        0\n",
       "1             0         2  10.359        2   RD   2.000        1\n",
       "2             0         3  41.813        4  SQL   4.000        1\n",
       "3             0         4   8.190        4   RD   4.000        1\n",
       "4             0         5  80.589        3  SQL   3.000        1\n",
       "...         ...       ...     ...      ...  ...     ...      ...\n",
       "4251        169        28   4.797        2   RD   2.000        1\n",
       "4252        169        29   4.589        4  SQL   4.000        1\n",
       "4253        169        30   4.384        4   RD   4.000        1\n",
       "4254        169        31   5.703        4  SQL   4.000        1\n",
       "4255        169        32   4.874        4   RD   4.000        1\n",
       "\n",
       "[4256 rows x 7 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_transformed_data = pd.read_csv(transformeddata)\n",
    "display(df_transformed_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4330c84c",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "## Filter users down to first 50 (VARIANT filters), and total time users took (in minutes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "56ad87ce",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dftemp:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>correct</th>\n",
       "      <th>start_datetime</th>\n",
       "      <th>sequence_num</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>worker_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>166</th>\n",
       "      <td>13.110</td>\n",
       "      <td>0.969</td>\n",
       "      <td>2023-07-06 15:33:42.37047</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>9.802</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-06 15:33:43.176727</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>21.080</td>\n",
       "      <td>0.656</td>\n",
       "      <td>2023-07-06 15:33:57.191982</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>3.517</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-06 15:34:39.261401</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8.065</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-06 15:34:43.567259</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>8.134</td>\n",
       "      <td>0.625</td>\n",
       "      <td>2023-07-06 15:37:12.663507</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>18.454</td>\n",
       "      <td>0.719</td>\n",
       "      <td>2023-07-06 15:42:20.453162</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>3.744</td>\n",
       "      <td>0.500</td>\n",
       "      <td>2023-07-06 16:05:20.144093</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>21.996</td>\n",
       "      <td>0.812</td>\n",
       "      <td>2023-07-06 16:34:33.028051</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>154</th>\n",
       "      <td>8.953</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-06 16:36:07.050799</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>8.125</td>\n",
       "      <td>0.969</td>\n",
       "      <td>2023-07-06 17:12:49.060194</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>7.874</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-06 17:18:50.168761</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>12.873</td>\n",
       "      <td>0.531</td>\n",
       "      <td>2023-07-06 18:15:32.990734</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>8.139</td>\n",
       "      <td>0.531</td>\n",
       "      <td>2023-07-07 18:10:46.043268</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>22.682</td>\n",
       "      <td>0.875</td>\n",
       "      <td>2023-07-07 18:11:03.035618</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>14.924</td>\n",
       "      <td>0.969</td>\n",
       "      <td>2023-07-07 18:11:10.045972</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>7.049</td>\n",
       "      <td>0.906</td>\n",
       "      <td>2023-07-07 18:11:10.432369</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>27.547</td>\n",
       "      <td>0.656</td>\n",
       "      <td>2023-07-07 18:11:30.090096</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>7.021</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-07 18:14:34.271669</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>16.834</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-07 18:28:15.617866</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>3.985</td>\n",
       "      <td>0.594</td>\n",
       "      <td>2023-07-07 18:43:12.941313</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>8.273</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-07 19:01:38.084371</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>6.495</td>\n",
       "      <td>0.969</td>\n",
       "      <td>2023-07-07 19:01:39.418515</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>19.664</td>\n",
       "      <td>0.844</td>\n",
       "      <td>2023-07-07 19:03:08.590044</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>16.758</td>\n",
       "      <td>0.625</td>\n",
       "      <td>2023-07-07 19:16:27.964457</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>13.759</td>\n",
       "      <td>0.656</td>\n",
       "      <td>2023-07-06 15:33:42.871209</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>11.329</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-06 15:33:43.726871</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>10.841</td>\n",
       "      <td>0.906</td>\n",
       "      <td>2023-07-06 15:33:53.523919</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>14.206</td>\n",
       "      <td>0.938</td>\n",
       "      <td>2023-07-06 15:33:53.819738</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>14.293</td>\n",
       "      <td>0.938</td>\n",
       "      <td>2023-07-06 15:34:02.551439</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>18.198</td>\n",
       "      <td>0.906</td>\n",
       "      <td>2023-07-06 15:35:42.063363</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>2.562</td>\n",
       "      <td>0.562</td>\n",
       "      <td>2023-07-06 16:06:22.118244</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>15.721</td>\n",
       "      <td>0.812</td>\n",
       "      <td>2023-07-06 16:24:41.817076</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>14.869</td>\n",
       "      <td>0.969</td>\n",
       "      <td>2023-07-06 16:24:45.985302</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>10.113</td>\n",
       "      <td>0.719</td>\n",
       "      <td>2023-07-06 16:46:59.265172</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>10.982</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-07 18:10:46.107235</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>8.165</td>\n",
       "      <td>0.906</td>\n",
       "      <td>2023-07-07 18:10:52.555864</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>14.199</td>\n",
       "      <td>0.594</td>\n",
       "      <td>2023-07-07 18:11:07.532996</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>17.812</td>\n",
       "      <td>0.562</td>\n",
       "      <td>2023-07-07 18:11:11.306257</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>14.009</td>\n",
       "      <td>0.500</td>\n",
       "      <td>2023-07-07 18:11:12.976612</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>3.266</td>\n",
       "      <td>0.969</td>\n",
       "      <td>2023-07-07 18:11:22.386833</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>7.527</td>\n",
       "      <td>0.875</td>\n",
       "      <td>2023-07-07 18:11:49.878986</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>13.752</td>\n",
       "      <td>0.531</td>\n",
       "      <td>2023-07-07 18:11:52.077358</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>6.050</td>\n",
       "      <td>0.969</td>\n",
       "      <td>2023-07-07 18:12:26.00741</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>19.763</td>\n",
       "      <td>0.531</td>\n",
       "      <td>2023-07-07 18:12:34.892116</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>12.004</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-07 18:13:35.80157</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>13.981</td>\n",
       "      <td>0.938</td>\n",
       "      <td>2023-07-07 18:14:47.210084</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>11.045</td>\n",
       "      <td>0.562</td>\n",
       "      <td>2023-07-07 18:16:33.378571</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>3.584</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2023-07-07 18:34:13.912492</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>14.587</td>\n",
       "      <td>0.938</td>\n",
       "      <td>2023-07-07 19:01:52.970403</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            time  correct              start_datetime  sequence_num\n",
       "worker_id                                                          \n",
       "166       13.110    0.969   2023-07-06 15:33:42.37047         0.000\n",
       "5          9.802    1.000  2023-07-06 15:33:43.176727         0.000\n",
       "110       21.080    0.656  2023-07-06 15:33:57.191982         0.000\n",
       "169        3.517    1.000  2023-07-06 15:34:39.261401         0.000\n",
       "2          8.065    1.000  2023-07-06 15:34:43.567259         0.000\n",
       "146        8.134    0.625  2023-07-06 15:37:12.663507         0.000\n",
       "96        18.454    0.719  2023-07-06 15:42:20.453162         0.000\n",
       "17         3.744    0.500  2023-07-06 16:05:20.144093         0.000\n",
       "87        21.996    0.812  2023-07-06 16:34:33.028051         0.000\n",
       "154        8.953    1.000  2023-07-06 16:36:07.050799         0.000\n",
       "159        8.125    0.969  2023-07-06 17:12:49.060194         0.000\n",
       "21         7.874    1.000  2023-07-06 17:18:50.168761         0.000\n",
       "92        12.873    0.531  2023-07-06 18:15:32.990734         0.000\n",
       "158        8.139    0.531  2023-07-07 18:10:46.043268         0.000\n",
       "43        22.682    0.875  2023-07-07 18:11:03.035618         0.000\n",
       "162       14.924    0.969  2023-07-07 18:11:10.045972         0.000\n",
       "125        7.049    0.906  2023-07-07 18:11:10.432369         0.000\n",
       "153       27.547    0.656  2023-07-07 18:11:30.090096         0.000\n",
       "136        7.021    1.000  2023-07-07 18:14:34.271669         0.000\n",
       "130       16.834    1.000  2023-07-07 18:28:15.617866         0.000\n",
       "119        3.985    0.594  2023-07-07 18:43:12.941313         0.000\n",
       "6          8.273    1.000  2023-07-07 19:01:38.084371         0.000\n",
       "121        6.495    0.969  2023-07-07 19:01:39.418515         0.000\n",
       "83        19.664    0.844  2023-07-07 19:03:08.590044         0.000\n",
       "0         16.758    0.625  2023-07-07 19:16:27.964457         0.000\n",
       "58        13.759    0.656  2023-07-06 15:33:42.871209         1.000\n",
       "66        11.329    1.000  2023-07-06 15:33:43.726871         1.000\n",
       "50        10.841    0.906  2023-07-06 15:33:53.523919         1.000\n",
       "42        14.206    0.938  2023-07-06 15:33:53.819738         1.000\n",
       "28        14.293    0.938  2023-07-06 15:34:02.551439         1.000\n",
       "80        18.198    0.906  2023-07-06 15:35:42.063363         1.000\n",
       "117        2.562    0.562  2023-07-06 16:06:22.118244         1.000\n",
       "52        15.721    0.812  2023-07-06 16:24:41.817076         1.000\n",
       "72        14.869    0.969  2023-07-06 16:24:45.985302         1.000\n",
       "75        10.113    0.719  2023-07-06 16:46:59.265172         1.000\n",
       "115       10.982    1.000  2023-07-07 18:10:46.107235         1.000\n",
       "39         8.165    0.906  2023-07-07 18:10:52.555864         1.000\n",
       "77        14.199    0.594  2023-07-07 18:11:07.532996         1.000\n",
       "168       17.812    0.562  2023-07-07 18:11:11.306257         1.000\n",
       "32        14.009    0.500  2023-07-07 18:11:12.976612         1.000\n",
       "143        3.266    0.969  2023-07-07 18:11:22.386833         1.000\n",
       "57         7.527    0.875  2023-07-07 18:11:49.878986         1.000\n",
       "148       13.752    0.531  2023-07-07 18:11:52.077358         1.000\n",
       "165        6.050    0.969   2023-07-07 18:12:26.00741         1.000\n",
       "89        19.763    0.531  2023-07-07 18:12:34.892116         1.000\n",
       "81        12.004    1.000   2023-07-07 18:13:35.80157         1.000\n",
       "60        13.981    0.938  2023-07-07 18:14:47.210084         1.000\n",
       "141       11.045    0.562  2023-07-07 18:16:33.378571         1.000\n",
       "91         3.584    1.000  2023-07-07 18:34:13.912492         1.000\n",
       "10        14.587    0.938  2023-07-07 19:01:52.970403         1.000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of participants who started with RD first: 25.000\n",
      "df_filtered_data:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>worker_id</th>\n",
       "      <th>question</th>\n",
       "      <th>time</th>\n",
       "      <th>pattern</th>\n",
       "      <th>mode</th>\n",
       "      <th>choice</th>\n",
       "      <th>correct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>139.259</td>\n",
       "      <td>2</td>\n",
       "      <td>SQL</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>10.359</td>\n",
       "      <td>2</td>\n",
       "      <td>RD</td>\n",
       "      <td>2.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>41.813</td>\n",
       "      <td>4</td>\n",
       "      <td>SQL</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>8.190</td>\n",
       "      <td>4</td>\n",
       "      <td>RD</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>80.589</td>\n",
       "      <td>3</td>\n",
       "      <td>SQL</td>\n",
       "      <td>3.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4251</th>\n",
       "      <td>169</td>\n",
       "      <td>28</td>\n",
       "      <td>4.797</td>\n",
       "      <td>2</td>\n",
       "      <td>RD</td>\n",
       "      <td>2.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4252</th>\n",
       "      <td>169</td>\n",
       "      <td>29</td>\n",
       "      <td>4.589</td>\n",
       "      <td>4</td>\n",
       "      <td>SQL</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4253</th>\n",
       "      <td>169</td>\n",
       "      <td>30</td>\n",
       "      <td>4.384</td>\n",
       "      <td>4</td>\n",
       "      <td>RD</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4254</th>\n",
       "      <td>169</td>\n",
       "      <td>31</td>\n",
       "      <td>5.703</td>\n",
       "      <td>4</td>\n",
       "      <td>SQL</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4255</th>\n",
       "      <td>169</td>\n",
       "      <td>32</td>\n",
       "      <td>4.874</td>\n",
       "      <td>4</td>\n",
       "      <td>RD</td>\n",
       "      <td>4.000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1600 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      worker_id  question    time  pattern mode  choice  correct\n",
       "0             0         1 139.259        2  SQL   1.000        0\n",
       "1             0         2  10.359        2   RD   2.000        1\n",
       "2             0         3  41.813        4  SQL   4.000        1\n",
       "3             0         4   8.190        4   RD   4.000        1\n",
       "4             0         5  80.589        3  SQL   3.000        1\n",
       "...         ...       ...     ...      ...  ...     ...      ...\n",
       "4251        169        28   4.797        2   RD   2.000        1\n",
       "4252        169        29   4.589        4  SQL   4.000        1\n",
       "4253        169        30   4.384        4   RD   4.000        1\n",
       "4254        169        31   5.703        4  SQL   4.000        1\n",
       "4255        169        32   4.874        4   RD   4.000        1\n",
       "\n",
       "[1600 rows x 7 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# New dataframe with worker id and when they started the HITS (allowing to sort by starting time)\n",
    "dfendtime = dfresults[[\"worker_id\", \"start_datetime\", \"sequence_num\"]]\n",
    "dfendtime.set_index(\"worker_id\", inplace=True)\n",
    "\n",
    "# New dataframe with worker ids and fraction correct (allowing to filter out those who did not pass the 0.5 correctness criterion)\n",
    "dftemp = df_transformed_data.groupby(['worker_id']).agg(\n",
    "    time=('time', np.sum),\n",
    "    correct=('correct', np.mean))\n",
    "dftemp['time'] = dftemp['time'] / 60\n",
    "dftemp.sort_values(by=['correct'], ascending=False, inplace=True)\n",
    "# display(dftemp)\n",
    "\n",
    "# Joining the dataframes\n",
    "dftemp = dftemp.join(dfendtime)\n",
    "\n",
    "# Filtering the dataframes for those who passed the 0.5 correctness criterion\n",
    "dftemp = dftemp.loc[dftemp.correct >= 0.5]\n",
    "\n",
    "# Keep only first 50 participants (creates imbalance: 26/24 between sequence numbers)\n",
    "# dftemp = dftemp.sort_values(by=\"start_datetime\", ascending=True)\n",
    "# dftemp = dftemp.head(50)                            # only keep the first 50 participants\n",
    "\n",
    "# Keep only first 50 balanced participants, thus first 25 from sequence 0, and first 25 from sequence 1\n",
    "dftemp0 = dftemp.loc[(dftemp.sequence_num == 0.0)].copy()\n",
    "dftemp0 = dftemp0.sort_values(by=\"start_datetime\", ascending=True)\n",
    "dftemp0 = dftemp0.head(25)                            # only keep the first 25 participants\n",
    "dftemp1 = dftemp.loc[(dftemp.sequence_num == 1.0)].copy()\n",
    "dftemp1 = dftemp1.sort_values(by=\"start_datetime\", ascending=True)\n",
    "dftemp1 = dftemp1.head(25)                            # only keep the first 25 participants\n",
    "dftemp = pd.concat([dftemp0, dftemp1])\n",
    "\n",
    "if VARIANT == 2:\n",
    "    dftemp = dftemp.loc[dftemp.correct == 1.0]                                  # 12/50\n",
    "if VARIANT == 3:\n",
    "    dftemp = dftemp.loc[dftemp.correct >= 0.9]                                  # up to 3 mistakes, 27/50\n",
    "if VARIANT == 4:\n",
    "    dftemp = dftemp.loc[dftemp.correct >= 0.66]                                 # up to 12 mistakes, thus 2/3 correct, 34/50\n",
    "print('dftemp:')\n",
    "display(dftemp)\n",
    "\n",
    "print('Number of participants who started with RD first:', np.sum(dftemp.sequence_num))\n",
    "\n",
    "df_filtered_data = df_transformed_data[df_transformed_data.worker_id.isin(dftemp.index)]     # only retain those that pass the 0.5 correctness criterium\n",
    "print('df_filtered_data:')\n",
    "display(df_filtered_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "524ecbe1",
   "metadata": {},
   "source": [
    "# Question 1. Figure 1a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "73a42469",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dfq1b:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>mode</th>\n",
       "      <th>RD</th>\n",
       "      <th>SQL</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>worker_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7.018</td>\n",
       "      <td>28.357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9.144</td>\n",
       "      <td>12.800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>9.378</td>\n",
       "      <td>16.846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>9.357</td>\n",
       "      <td>16.023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>9.960</td>\n",
       "      <td>13.986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>5.756</td>\n",
       "      <td>5.691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>8.377</td>\n",
       "      <td>12.603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>10.586</td>\n",
       "      <td>12.822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>17.987</td>\n",
       "      <td>25.491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>7.220</td>\n",
       "      <td>9.152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>13.366</td>\n",
       "      <td>23.652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>7.929</td>\n",
       "      <td>13.225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>5.414</td>\n",
       "      <td>8.546</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>13.305</td>\n",
       "      <td>17.604</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>5.921</td>\n",
       "      <td>8.445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>10.258</td>\n",
       "      <td>8.849</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>11.616</td>\n",
       "      <td>9.700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>8.540</td>\n",
       "      <td>12.380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>18.521</td>\n",
       "      <td>28.279</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>12.252</td>\n",
       "      <td>12.471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>18.601</td>\n",
       "      <td>31.567</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>15.697</td>\n",
       "      <td>30.716</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>14.380</td>\n",
       "      <td>18.619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>10.471</td>\n",
       "      <td>38.456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>19.988</td>\n",
       "      <td>42.377</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>20.476</td>\n",
       "      <td>54.861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>4.865</td>\n",
       "      <td>6.268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>17.733</td>\n",
       "      <td>18.866</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>10.912</td>\n",
       "      <td>30.937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>22.293</td>\n",
       "      <td>25.159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>11.087</td>\n",
       "      <td>18.552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>4.360</td>\n",
       "      <td>4.181</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>7.749</td>\n",
       "      <td>5.184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>8.207</td>\n",
       "      <td>10.006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>4.792</td>\n",
       "      <td>8.986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>7.131</td>\n",
       "      <td>6.258</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>10.635</td>\n",
       "      <td>14.470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>18.003</td>\n",
       "      <td>4.572</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>3.221</td>\n",
       "      <td>5.837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>4.337</td>\n",
       "      <td>16.021</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>17.566</td>\n",
       "      <td>14.229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>14.242</td>\n",
       "      <td>50.465</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>154</th>\n",
       "      <td>10.534</td>\n",
       "      <td>14.382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>6.451</td>\n",
       "      <td>12.857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>9.951</td>\n",
       "      <td>15.182</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>5.090</td>\n",
       "      <td>9.694</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>8.392</td>\n",
       "      <td>10.698</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>166</th>\n",
       "      <td>11.477</td>\n",
       "      <td>12.369</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>30.665</td>\n",
       "      <td>36.904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>4.778</td>\n",
       "      <td>7.235</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "mode          RD    SQL\n",
       "worker_id              \n",
       "0          7.018 28.357\n",
       "2          9.144 12.800\n",
       "5          9.378 16.846\n",
       "6          9.357 16.023\n",
       "10         9.960 13.986\n",
       "17         5.756  5.691\n",
       "21         8.377 12.603\n",
       "28        10.586 12.822\n",
       "32        17.987 25.491\n",
       "39         7.220  9.152\n",
       "42        13.366 23.652\n",
       "43         7.929 13.225\n",
       "50         5.414  8.546\n",
       "52        13.305 17.604\n",
       "57         5.921  8.445\n",
       "58        10.258  8.849\n",
       "60        11.616  9.700\n",
       "66         8.540 12.380\n",
       "72        18.521 28.279\n",
       "75        12.252 12.471\n",
       "77        18.601 31.567\n",
       "80        15.697 30.716\n",
       "81        14.380 18.619\n",
       "83        10.471 38.456\n",
       "87        19.988 42.377\n",
       "89        20.476 54.861\n",
       "91         4.865  6.268\n",
       "92        17.733 18.866\n",
       "96        10.912 30.937\n",
       "110       22.293 25.159\n",
       "115       11.087 18.552\n",
       "117        4.360  4.181\n",
       "119        7.749  5.184\n",
       "121        8.207 10.006\n",
       "125        4.792  8.986\n",
       "130        7.131  6.258\n",
       "136       10.635 14.470\n",
       "141       18.003  4.572\n",
       "143        3.221  5.837\n",
       "146        4.337 16.021\n",
       "148       17.566 14.229\n",
       "153       14.242 50.465\n",
       "154       10.534 14.382\n",
       "158        6.451 12.857\n",
       "159        9.951 15.182\n",
       "162        5.090  9.694\n",
       "165        8.392 10.698\n",
       "166       11.477 12.369\n",
       "168       30.665 36.904\n",
       "169        4.778  7.235"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Median time RD: 10.11, 95% CI [8.38, 11.26]\n",
      "Median time SQL: 13.61, 95% CI [12.37, 16.43]\n"
     ]
    }
   ],
   "source": [
    "# create two columns mode and median, with 2 rows per worker (used for Fig 1a violines)\n",
    "dfq1a = df_filtered_data.groupby(['worker_id', 'mode']).time.agg(['median'])\n",
    "dfq1a.reset_index(inplace=True)\n",
    "# print('dfq1a:')\n",
    "# display(dfq1a)\n",
    "\n",
    "# pivot to have one row per worker (used for Fig 1a individual points)\n",
    "dfq1b = pd.pivot_table(dfq1a, values=['median'], index=['worker_id'], columns=['mode'])\n",
    "dfq1b=dfq1b.droplevel(0, axis=1)\n",
    "print('dfq1b:')\n",
    "display(dfq1b)\n",
    "\n",
    "modes = ['RD', 'SQL']\n",
    "median_time = {}\n",
    "ci = {}\n",
    "ci_delta = {}\n",
    "for mode in modes:\n",
    "    median_time[mode] = np.median(dfq1b[mode])\n",
    "    ci[mode] = scipybootstrap((dfq1b[mode],), statistic=np.median, n_resamples=BOOTSTRAPSAMPLES, confidence_level=BOOTSTRAPCONFIDENCE, method='percentile', axis=0).confidence_interval        #convert array to sequence\n",
    "    ci_delta[mode] = [median_time[mode] - ci[mode].low, ci[mode].high - median_time[mode]]\n",
    "\n",
    "    print(f'Median time {mode}: {median_time[mode]:.2f}, 95% CI [{ci[mode].low:.2f}, {ci[mode].high:.2f}]')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "3c394044",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA24AAAIrCAYAAACXs8oMAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOz9d5wl13nfCf9uzqm7J/TkiJnBDAZhABAAQQokJYoUSYGUSVrB5lqrD3dpBUtraV9bMi2H9av0aqXXa5sixSDLIlfiUoxiABOISEQOgQEGmDw9oaenc/fNlfeP28/pU3Xr5tjdz/fzaczFDVWnqk6dOr/zJI9lWRYYhmEYhmEYhmGYocU76AYwDMMwDMMwDMMw9WHhxjAMwzAMwzAMM+SwcGMYhmEYhmEYhhlyWLgxDMMwDMMwDMMMOSzcGIZhGIZhGIZhhhwWbgzDMAzDMAzDMEMOCzeGYRiGYRiGYZghh4Vbm5w9exZnz54ddDMYhmEYhmEYhtkAsHBrE1VVUSqVBt2MDc309PSgm7Dh4WswePgaDBY+/4OHr8Fg4fM/ePgaDJ5+XQMWbh1gWdagm7ChKZfLg27ChoevweDhazBY+PwPHr4Gg4XP/+DhazB4+nUNWLgxDMMwDMMwDMMMOSzcGIZhGIZhGIZhhhwWbgzDMAzDMAzDMEMOCzeGYRiGYRiGYZghh4UbwzAMwzAMwzDMkMPCjWEYhmEYhmEYZshh4cYwDMMwDMMwDDPksHBjGIZhGIZhGIYZcli4MQzDMAzDMAzDDDks3BiGYRiGYRiGYYYcFm4MwzAMwzAMwzBDDgs3hmEYhmEYhmGYIYeFG8MwDMMwDMMwzJDDwo1hGIZhGIZhGGbIYeHGMAzDMAzDMAwz5LBwYxiGYRiGYRiGGXJYuDEMwzAMwzAMwww5LNwYhmEYhmEYhmGGHBZuDMMwDMMwDMMwQw4LN4ZhGIZhGIZhmCGHhRvDMAzDMAzDMMyQw8KNYRiGYRiGYRhmyGHhxjAMwzAMwzAMM+SwcGMYhmEYhmEYhhlyWLgxDMMwDMMwDMMMOSzcGIZhGIZhGIZhhhwWbgzDMAzDMAzDMEMOCzeGYRiGYRiGYZghh4UbwzAMwzAMwzDMkMPCjWEYhmEYhmEYZshh4cYwDMMwDMMwDDPk+AfdAIZhGBnLsqBpGkqlEsrlMsrlMjRNg2VZsCwLpmna/pVf+3w+BAIBBAIBBINB27/05/P5Bn2IDMMwDMMwLcPCjWGYnqDrOrLZLJaXl7G8vIylpSVks1nbv/l8Xogz+iuVSjBNs2ft8vv9iEQiiMfjdf8SiQTS6TRGR0eRyWQQCAR61iaGYRiGYZhGsHBjGKZlLMvC0tISJiYmcPXqVczOzmJmZgYzMzOYnZ3F3Nwc8vl8x/vxBwIIBMPw+f3werzweDyVPy+9XvnX6wHggWWa0HUNuq7B0DXomgZdU2FZltimruvI5XLI5XIttSWVSmFkZASjo6NV/27ZsgVbtmxBKBTq+JgZhmEYhmHcYOHGMIwrlmVhYWEB165dw/Xr13Ht2jXcuHFDiDNN0xpuw+PxIBJLIBpPIBZPIppIIhpLIpZIIhpLIByNIxQOIxAMIxgMIRAKIRgKIxgKIxAMwevtThiuaRhC0OmaBqVcQrlYQLlUQLlYQKlYQLmYX/n/IkrFPMrFAvK5JeSXl2AYurAcXr58ueZ+RkdHsXXrVoyPj1f9m0gkunIsDMMwDMNsTFi4McwGxzAM3LhxQwg0EmmTk5MolUo1f+fxeBBPZZDKjCGVGUVqZAzJzBhSmTEk0yOIJVIIR2NdE1+d4PX5EPT5EAyFW/6tZVkoFfLIZxeRW6785bNLyC0tIp9dRHZpAUvzs1DKRczPz+PkyZMAgN27d9u2E4/HsXPnTuzatcv2l8lk4PF4unKcDMMwDMOsX1i4McwGQlEUXLlyBRcvXsSlS5dw6dIlTExMQFVV1+97vF5kxrZgbPM4Rrdsw8imcaRHNyGVGUUiPQKfb/0PIR6PB9F4xWq4edsu1+9YloVSMY+luRn84Ud/C5qi4LZ7HsTi3AyW5meQz1bi+V5//XW8/vrrtt/G43GbkNu9ezf27duHeDzej8NjGIZhGGaNsP5nXQyzQSmVSrhw4QIuXrwohNr169ddE38EQ2GMbtmG0c3jGN28DWNbtmF0yzgyo1vg8/Mw0QiPx4NoLIFoLIFUZgwA8J5f/F/F55qqYHFuGrM3JzF78zrmbk5i7uYkFuemkc/n8dprr+G1116zbXPr1q3Yv3+/7S+VSvX1uBiGYRiGGR54RsYw6wDLsjA1NYUzZ86IvytXrriKtGg8ia07dmPL9t0Y3bwdX//bT0BVyvilX/3dtlwJmcYEgiFs3rarymKnayrmZ6aEmJu9OYmZG1exvDCHmzdv4ubNm3j66afF9zdt2oT9+/fjwIED2L9/Pw4dOtTvQ2EYhmEYZkCwcGOYNUi5XMb58+eFSDt79iyWl5ervpfMjGLrjr1CqG3dsQfxZNoWU3X83gf72XRGwh8IYsv2yrWRKRXyuDl5BdPXJzB1/TKmr1/BwuxNzM7OYnZ2Fs8++6z47pYtW3D06FEcPnwYhw8fxq5du7hWHcMwDMOsQ1i4McwaoFwu48yZMzh16hReeeUVnD9/HoZh2L7j8/mxdecebN9zEDv2HMT2PQeQSGUG1GKmEyKxOPbechR7bzkq3lPKJdy8PrEi5iYwdfUSFmZvYnp6GtPT03j00Ucrv41EcPDgQSHkDh06hGQyOahDYRiGYRimS7BwY5ghRFVVIdReffVVnD17Frqu274TT2WwY88BIdS27NgNv5+LRK9XQuEIdh84gt0Hjoj3ioUcbly5iMmJC5icuIAbVy+iVCrh1KlTOHXqlPje7t27cfToURw7dgxHjx7FyMjIIA6BYRiGYZgOYOHGMEOAYRi4cOECfvzjH+PUqVM4c+ZMVZ20RHpkZeJ+K3YfOIzUyCZOI7/BicYSOHDrHThw6x0AANM0MXdzEpMT5zF55QKuT1zAwswUrly5gitXruCb3/wmAGB8fBzHjh0TQm7z5s3clxiGYRhmyGHhxjADYn5+Hj/+8Y9x8uRJvPTSS8jlcrbP48l0RagdvBW79h9BZqz7k2vTMHD+tR8DAA7eeie8HBu1pvF6vdi8bSc2b9uJOx94KwCgkM/i+qVzuHrxDK5ePIPpG1cxNTWFqakpfPe73wUAjI2N4dixY7j99ttx/PhxbN68eZCHwTAMwzCMCyzcGKZPaJqG1157DSdPnsTJkycxMTFh+zwUjmLPSlzT7oO3YmTT1p5bQXRdwxc/858BAL/zR59EkIXbuiMWT+LQ8btx6PjdAIByqYDrl8/j2sWzuHrxDKauXcbc3Bwee+wxPPbYYwAqFrnjx48LIcdlCBiGYRhm8LBwY5gesrCwgBdeeAHPP/88Xn75ZSiKsvqhx4PxnXux7/Bx7Dt0G7bv3s8WL6bnhCMxm3ulqpQxOXEBVy6+jivnXsONa5eERe7b3/42AGDPnj1CxB07dgzRaHSAR8AwDMMwGxMWbgzTRSzLwuXLl/H888/jueeew4ULF2yfxxIp7Dt8G/YdPo69txxDNJ4YUEsZpkIwFMbeQ8ew99Ax4Gcq2SuvXngdE+dfw8T51zA7dQ0TExOYmJjAV7/6Vfh8Phw+fBh33nkn7rzzTuzfv5/LDzAMwzBMH2DhxjAdomkaTp06heeffx7PP/885ubmbJ9v270fB4/eiQO33oHN23ZxEghmqAmFIzh47C4cPHYXAKCQW8aVC69j4txpTJx/DUvzMzh9+jROnz6Nz372s0gkErj99ttx55134o477uD4OIZhGIbpESzcGKYNisUinn/+eTzzzDP48Y9/jFKpJD7zB4LYd+gYDqyItXgyPbiGNsDn9+Pdv/Bh8ZphnMQSKdx653249c77AACLc9O4fPZVXDr7Kq6cP41cLoennnoKTz31FABgx44duPPOO3HXXXfh2LFjCIfDg2w+wzAMw6wbeKbGME2yvLyM5557Ds888wxeeuklW121eCqDg7feiYNH78Tug7ciEAwOsKXN4/P5cfzeNw+6GcwaIjO2BZmxLbjrjW+DaRi4cfUiLp19FZfPvoIbVy7i+vXruH79Ov7hH/4BgUAAR48exYkTJ3DixAns2LGDLc4MwzAM0yYs3BimDvPz83jmmWfwzDPP4NVXX4VpmuKz0c3bcOj2u3Ho2Als3bmXJ6TMhsPr82HH3luwY+8tePM7fg7lUgET51/DpTOv4NKZU8guzuOll17CSy+9hE9/+tPYvHkz7rrrLtx11124/fbbOckJwzAMw7QACzeGcTAzM4Onn34aTz/9NM6ePWv7bMv23Th0/G4cvv0ejG3ZPqAWMsxwEo7EcPj4PTh8/B5YloX5mRu49PopXDzzCq5ePIOZmRk88sgjeOSRR+Dz+XDkyBFhjduzZw8vfjAMwzBMHVi4MQyAubk5PP3003jyySerxNqOPQdx6PjduOW2u5EZ48QLDNMMHo8HY1u2Y2zLdtz70DuhKmVcvXgGl86cwsXXT2FxbhqvvvoqXn31Vfz1X/81RkdHceLECdx9991sjWMYhmEYF1i4MRuWhYUFIdZef/311Q88HuzadwiH77gXh267G4lUZnCN7DGqUsZ/+fe/CQD4jX//nxEMcSIJpjcEQ2Fb/biF2ekVEfcyrlx4HfPz8/jOd76D73znO/D5fLbYuF27OBsrwzAMw7BwYzYUS0tLePrpp/HUU0/h9OnTsCxLfLZj7y04cscbcOT2exFPpQfXyD6jlIuDbgKzARnZtAUjm34Kd7/pp6CpKq5ePIOLr7+Mi6+/jMW5aZw6dQqnTp3CX/3VX2HTpk3CGnf8+HFEIpFBN59hGIZh+g4LN2bdUygU8Mwzz+CJJ57Ayy+/bEswsn3PARy54w04fPu9SKZHBthKhtm4BIJB7D9yHPuPHAfwT7Ewe3NFxJ3ClQuvY3Z2VsTG+f1+HDt2DHfffTdOnDiB7du3szWOYRiG2RCwcGPWJYqi4MUXX8Tjjz+OF198EZqmic/Gd+3DrXe8AYfvuBepzNgAW8kwjBsjm7ZiZNNW3PPmn4amKrhy/nVcPPMyLr72MpYWZkWmyk996lPYunWrEHG33XYbQqHQoJvPMAzDMD2BhRuzbjAMAy+99BKeeOIJPPPMM7ai2KNbtuHoXffj1jvvx8imLQNs5XDh8Xixa/9h8Zphho1AMIQDR+/AgaN3wPo5q2KNe+0lXHj9ZVy7eBY3b97E17/+dXz9619HMBjEsWPHhFvltm3bBt18hmEYhukaLNyYNY1lWTh79iwee+wxPPXUU1heXhafJTOjOHrn/bj1rvuweRsnN3AjEAzin/z6vxl0MximKTweD0Y3j2N087jIVDlx/jVcfO1lXDzzMrKL8zh58iROnjyJT37ykxgfHxcJTo4dO4ZwmJPvMAzDMGsXFm7MmmRychLf+MY38NJLL2Fqakq8H40ncPj2e3H0rgewY88BeLxsRWKY9UowFMYtx+7CLcfugmVZmLs5WXGpfP0Url06i6mpKWGNCwQCuO2223DXXXdxbBzDMAyzJmHhxqwZFhcX8eSTT+Kxxx7D+fPnxfuBYAiHbrsbR0/cjz23HIXPx92aYTYaHo8Hm8Z3YNP4Dtz3lndBKZdw5fxruPj6qSpr3Kc+9Sls3rwZd911F+666y6uG8cwDMOsCXiGyww1pVIJzz77LB577DG89NJLIiOkx+vFvkPHcOzEG3Hw2F1cf4xhGBuhcAS33HYCt9x2omKNm76BS6+vWuNmZmZEpkqv14sjR47gzjvvxF133YX9+/fDy9Z6hmEYZshg4cYMHYZh4OWXX8YPfvADPPPMM1AURXy2bfd+HL3rAdx65xsQS6QG2Mr1ga5r+ME/fB4A8Jb3/GP4/YEBt4hhuo/H48Gmrduxaet2vOEtPwNVKePqhTO4dPYULp15BQuzN3H69GmcPn0an/3sZ5FMJoWIu/322zE6OjroQ2AYhmEYFm7McGBZFi5evIjHHnsMTzzxBBYXF8VnmbEtOHbiARw98QBGNm0dYCvXH6Zh4IUnvg0A+ImfeT/Awo3ZAARDYZGpEgCW5mdw6cwruHT2FUycO41sNovHH38cjz/+OABg586duPPOO3HHHXfg2LFjXACcYRiGGQgs3JiBMjMzg8cffxw/+MEPcO3aNfF+JBbHrXfeh2N3vxHbdu3nJAIMw/SM9Ohm3PXGt+GuN74NhqFjcuIiLp05hctnX8HU9Qlcu3YN165dw9e+9jX4fD4cPnwYd9xxB+644w4Eg8FBN59hGIbZILBwYwbCjRs38N/+23/DqVOnxHs+fwC3HLsLx048gH2Hj8Pn5+7JMEx/8fn82LX/EHbtP4SH3vUBlAp5TJx/DZfPvYqJs69iaWFWuFV+7nOfQyQSwe23347jx4/j+PHj2LWLS48wDMMwvYFnxsxAePTRR4Vo23XgCG67+404dPwehCOc2Y1hmOEhEovjyB334sgd9wIAFudmcPncq7h89lVcufAaSsUCnn32WTz77LMAgHQ6jWPHjgkht23bNhZyDMMwTFdg4cYMBF3XAQD3vPmn8VPv+ycDbs3GJRAI4lf/7Z+J1wzD1CczthmZsbfirgfeCtM0cfPaZUycfw1Xzr+Ga5fPYWlpCU899RSeeuopAMDo6Chuu+02HD9+HLfddhu2bNnCQo5hGIZpCxZuzEDhAtmDxeP1Ij2yadDNYJg1idfrxbbd+7Ft93488JPvga5ruHHlEq5cqAi5yYkLmJ+fx2OPPYbHHnsMADA2NoajR4/i6NGjOHbsGHbs2MFCjmEYhmkKFm4MwzAM0wX8/oCIj3vTT78Pmqri+sR5IeSmrl7G3NycLWNlMpkUQu7o0aPYu3cvfD7fgI+EYRiGGUZYuDEMwzBMDwgEg9h7y1HsveUoAEBVyrhx9SKuXjyLaxfPYPLKBWSzWTzzzDN45plnAACRSARHjhzB4cOHcfjwYRw6dAjRKMf+MgzDMCzcGGZDY5omZqcqZRg2je+El11XGaZnBENh7Dl4FHsOVoScoeuYun4Z1y6exdWLZ3D98nmUSkWcPHkSJ0+eBFApHr57924cOnRICDpOeMIwDLMxYeHGMBsYXVPx6T/9KADgd/7okwiGwgNuEcNsHHx+P3bsOYgdew7i/re9WyykXL98HtcnzmPy8nksLcxiYmICExMT+Pa3vw2g4l556NAhHDp0CAcPHsTBgweRSCQGfDQMwzBMr2HhxjAMwzBDgNfrxZbtu7Fl+26cePAnAQD57BImJy5UhNzEBUxdu4xsNosXXngBL7zwgvjt+Pi4EHEHDx7E/v37EQ7zQgzDMMx6goUbwzAMwwwp8WQah47fjUPH7wYA6LqG6ckrmLx8HjeuXsLUtUtYnJvB1NQUpqam8MQTTwCoiMBdu3YJEbd//37s3r0bkUhkkIfDMAzDdAALN4ZhGIZZI/j9AWzffQDbdx8Q75UKeUxdu4wbVy9i6tplTF29hHx2SbhYfve73wVQiZfbtm0b9u3bh71792L//v3Yu3cvMpnMoA6HYRiGaQEWbgyzgfH5/Hjwp98nXjMMs/aIxOLYd/g27Dt8m3gvt7SAG9cuYerqZUxPXsH05JWK2+XkJCYnJ/Hkk0+K746MjGDv3r3Ys2cPdu3ahd27d2PHjh0IhUKDOByGYRimBjxTY5gNjM/vx5vf8XODbgbDMF0mkR7BofQIDt12t3gvn1vGzOQVTE9exc3JCUxPXsXC7E0sLCxgYWEBP/rRj8R3vV4vxsfHhZCjf7dt2wa/n6cODMMwg4BHX4ZhGIbZAMQTKcQPH8e+w8fFe6pSxszUNUxPXsHs1HXM3ryOualJlIp5YZ2jGnMA4Pf7MT4+jh07dmD79u22f+Px+CAOi2EYZsPAwo1hGIZhNijBUFiUJCAsy0Ihu4zZmxUhR4Juduo6NFXBtWvXcO3ataptpdNpbN++XfyNj49jfHwcW7du5QyXDMMwXYCFG8MwDMMwAo/Hg3gqjXgqjb2Hjon3LdPE8tI8FmamMC/9LcxMIbe8iKWlJSwtLeH06dNV28xkMti6dasQcvTv1q1bkUqluKA4wzBME7BwY5gNjKaq+Ov//B8AAP/Tb/47BILBAbeIYZh+0M697/F6kR7ZhPTIJpu7JQAo5VIlXo4E3ewUluZmsDg/g3KxgMXFRSwuLuL111+v2m4wGMTmzZuxefNmbNq0qer1yMgIfD5fdw6cYRhmDcPCjWE2MJZlYubGVfGaYZiNQbfv/VA4gvGdezG+c2/VZ6ViAUtz01icn8Hi3AyW5qexuCLqcsuLUFUV169fx/Xr11237fP5MDo6itHRUYyNjYnX9KcoCrZt24ZAINDxcTAMwwwzLNwYhmEYhukZkWgMkV37ML5rX9Vnuq4hu7iA5cU5ZBfnsLwwh+XFOSwvziO7MIfs0gIMw8DMzAxmZmbq7iedTmNkZAQjIyPIZDLi/52vOd6OYZi1Cgs3hmEYhmEGgt8fwMimLRjZtMX1c9M0kc8uIbs4h9zyInJLC8gtLyK7vIDc0iJyywvILy/BMHQRY3fp0qW6+4xEIkin00ilUjX/0uk0kskkkskkW/IYhhkaWLgxzAbG4/EglRkTrxmG2RislXvf6/UimR5BMj1S8zuWaaJYyFWE3fIiCrllFHLLyGeXkM8ur/z/EvLZJWiqilKphFKphKmpqabaEIlEkEwmkUgkhJij1/K/8Xgc8XgciUQC0Wh0qM8rwzBrExZuDLOBCQRD+LXf//NBN4NhmD6znu59j9eLWCKFWCKFrTv21P2uUi4hn11CMZ9FMZ9DIZ9deZ1deZ2rvM5lUSrkYFmWEHrT09NNt8nr9SIWi9kEHf3FYjHbv873otEoJ2NhGMYVFm4MwzAMw2wIQuEIQuEIRjePN/yuZZool4ooFnIoFfMo5Vf+LeQr7xXyKBbyKBVzKBcLKBcLKBUL0DUVpmkil8shl8u11c5IJIJYLCaEHL2W/+h9t38jkQhb/BhmHcLCjWEYhmEYxoHH60UkFkckFm/pd7qmolwsolTMo1yqiLlyMV8RdqUilFIB5VKxIvZKRSjl1deaqgCAsPLNzc211Xav1yvEHwk65/9Ho9Gar+n7bPljmOGChRvDMAzDMEyX8AeCiKeCiKfSLf/W0HWUSwUo5RKUUhHllX+VcrEi8kpFIfbsn5dW3ivBNA2YpolCoYBCodDRsUQiEVfBJ/9bLpexc+fOqvfZ7ZNhug8LN4bZwBiGjh899T0AwIkHfxI+Hw8JDLMR4Ht/OPH5/SJerx0sy4KuqVBKJZTLRSHmlHJRiMGKyCutiMES1PLq5+WVzw1dA7Bq+VtYWGj7mMLhsBBx8Xi8Stg54/voNf0Fg0F2+2SYFXikZpgNjKHr+N5XPgcAuOO+h3jyxjAbBL731ycejweBYAiBYKgtix+h69qK0CNRtyoAy5IQVCWxZxeJRWiqCgAol8sol8uYn59vqy1+v79KzNGfM7GL/D69x+UcmPUEj9QMwzAMwzCMwO8PwB8PIBZPtr0Nw9AlsSe7e5YqcX4ON1ByBa24gFb+tSwLur5ao68dQqFQVRZPt/93y/QZCoXY2scMFSzcGIZhGIZhmK7i8/kRjScQjSfa+r1lWVCV8kpcX6Fi6aPELivvKSVZ9BXE++ViRRACgKIoUBSlLYsfWfuorINT7MnlHpzlH4LBYFvHzTD1YOHGMAzDMAzDDBUej0eUb0hmRlv+vWmalbi9YmFF+BVXXxcLUMrFSsbPFaFH79N3LdPsyNoXDAarRB4JP7f6fvJ7nNCFqQULN4bZwPgDQfzSr/2eeM0wzMaA731mveP1ehGJxhCJxlr+LVn7SMyVJMFHQo/KPdhE38rnlmVBVVUsLCy0ldglEolUCbt6f9lsFplMBrFYjEXfOoeFG8NsYLxeL3YfODLoZjAM02f43meY2sjWvlRmrKXfWqYJRSmv1O4roiREXV5Y/WTRt/q6UgYCWM3mOTs723LbKTNnrVi+erF9fj/LgmGHrxDDMAzDMAzDdAGP14twJIpwJAq06OFpGgbKVJB9xXpXKhZsLp6V98jKlxcWP1UpAwCKxSKKxSJmZmZabjuVbnBm62z0LxVs93q9Le+TaQ0WbgzDMAzDMAwzYLw+H6KxBKKx1hO6GIbuiONzxO1Jok9xxPyRpa+T0g0ej8dWo6+VP/odW/waw2eIYTYwlmWJzFuhcJTTHjPMBoHvfYZZX/h8fsTiybZKOJimKWXvJEHnzNy5+pn8nlIuQtc0WJaFQqGAQqHQ9jEEg0GbkKv3r/M1/X8gEFjX4xkLN4bZwGiqgj/7vY8AAH7njz6JYCg84BYxDNMP+N5nGIbwer2IxOKIxOJt/V7XVFGXj8ReaSVmr1GtvnKpCE1VAACqqkJVVSwuLrZ9LH6/3ybmIpGI7f/dPnP7NxwOD6XrJws3hmEYhmEYhmHawh8IIh4IIp5ItfV70zAqIq9crJRwIIFHr1feV0pFKEoJSqkEtVxCuVyEWi5XPlfKwErB9mw2i2w22/FxRSIR2x+JOre/YrGIq1ev4k1velPH+60HCzeGYRiGYRiGYQaC1+fryOIHVLJ5qqqyIu7KUCUhqJRLK/9fgrJS1F0pl6DS95RSRQAqle9YpglgNbtnK7BwYxiGYRiGYRiGqYHH6xUlHDrBsizomibEnBB2SlkIPVUpV8Sh9P+V91oTee3Awo1hNjBenw+33fOgeM0wzMaA732GYZhqPB4PAsEgAsEg0KbrZy9pW7jpuo5vfvObePTRR/HKK6+IyvCZTAYHDhzAm970Jrzvfe9DPN7Y7GlZFh5//HE88sgjePnll3Hz5k2oqopUKoX9+/fjTW96Ex5++GFs2bKl7nbe+ta3YnJyEvfeey/+5m/+pt1DY5gNg98fwHt+8X8ddDMYhukzfO8zDMOsPdoSbmfPnsW//Jf/EhcuXKj6rFgsYnJyEo8//jg+9rGP4fd///fxzne+s+a2XnrpJfzbf/tvce7cuarP5ufnMT8/j+effx7/5b/8F/zyL/8yfv3Xfx3BYLCdZjMMw/QUy7Jg+sOwfAHkFR2xoG9dpyVmGIZhGKZ/tCzcpqen8Su/8iuYnZ3F2NgYPvzhD+P+++/H5s2b4fF4MD09jaeffhp/+Zd/iYWFBfz2b/82IpEIHnrooaptfe1rX8Pv/u7vQtd1+Hw+PPzww3jHO96BgwcPIhwO4/r163j++efx2c9+FlNTU/jEJz6BF198EZ/4xCeQSLRenJBhGKYXFFUdz11ewGPnZrB8z4cBAP/qS6cwFg/ioVs24w17RxANsmc6wzAMwzDt0/JM4pOf/CRmZ2eRTqfxhS98Adu2bbN9nk6ncejQIbz1rW/F+9//fuRyOfzhH/4h3vzmN9vqITz//PNCtI2Pj+Mv/uIvcOTIEdu2RkZGcPz4cXzoQx/CRz/6UXz1q1/Fj370I/zWb/0WPvWpT/FKNsMwA+e1qSw++dQlqLpZ9dlcXsXfn7yOr526gQ8/uA+3jrdeGJVhGIZhGAYAWq4s94Mf/AAA8J73vKdKtMns2bMHH/lIpbjnxMQETp8+LT5TVRUf/ehHoes64vE4/sf/+B9Vok0mGAziT/7kT/Dud78bAPDUU0/hi1/8YqtNZxjGga6p+PtP/zn+/tN/Dl1TB92cNcdrU1l87PEL0FxEm4ymm/jY4xfw2lTndWUYphvwvc8wDLP2aFm4zc7OAgDK5XLD7z700EM4dOgQ7rvvPui6Lt7/5je/iStXrgAAfuM3fgO7du1qat//5t/8G6RSlQwvH//4x2Ga9SdLDMPUxzRNnHv1JM69epLvpxYpqjo++dQlwAKsBt+1Vv7zyacuoajqDb7NML2H732GYZi1R8uukjt27MDFixfxjW98A7/4i7+IW2+9teZ3Dxw4gK997WtV79N7oVAI73//+5ve98jICN7xjnfg85//PK5du4aTJ0/i7rvvbvUQGIbZ4Hzr1Sk8cvpmR9swLQtmI8UmYQFQdRP/6kun4F2jbt7vOLoV7zw2PuhmMAzDMMyGpGXh9r73vQ9/+qd/imKxiA9+8IP4yZ/8Sfz0T/807rvvPmQymYa/13UdL7zwAgDg2LFjTZULkHnb296Gz3/+8wCA5557joXbGmf6+gRefOq7g27GhkXXNfH65DOPwu8PDLA1/eP6cgS6GR3Ivk2rIvrWItevXMSLS68OuhlMF9io9z7DMEwvuf/AB3q6/ZaF2y//8i/j+eefxxNPPAFN0/Ctb30L3/rWt+DxeHDw4EGcOHECb3jDG/Dggw+6Zn68du0aVLXiT3/gwIGWG7x3717x+tKlSy3/nhkOAoHKJOHKhddx5cLrA24NAwCPfvVvB92EvpHdfh+w/f5BN2PNcfH1lzE7+eygm8F0mY107zMMw/SS3/hnQybc/H4/Pv7xj+Mzn/kM/vIv/xLZbCXY3rIsnDt3DufOncPf/u3fIhAI4K1vfSt++7d/G7t37xa/X1paEq+TydYzrI2Njblui1lb/NRP/RTm5uZQKpXa3kaxWEQ0OhirCVNhrV6DV43teI3Delpm585dOLbHN+hmDBVr9R5YT/A1GCx8/gcPX4PB069r0FZhIZ/Phw9/+MP4p//0n+KJJ57AY489hmeffRaTk5PiO5qm4dvf/jZ+8IMf4A/+4A/wnve8pysNlksA+Hw8gVirbN68Gb/5m7/Z0TauXLliWxRg+s9avQYff/wSXntiYtDNWHM8+OAb8ZGf2DfoZgwVa/UeWE/wNRgsfP4HD1+DwdOva9BRRdhwOIy3v/3tePvb3w4AmJqawnPPPYcnn3wSjz76KIrFIlRVxb/+1/8au3btwu233450Oi1+T9a6VlheXhavR0ZGOmk+wzAMwzAMwzDMmqAj4eZkfHwc733ve/He974XS0tL+OM//mN86Utfgq7r+PjHP46/+Iu/wM6dOxEMBqGqKs6dO9fyPuTf7Nixo5vNZxhmg+DzeBDwdZbZ0TBbyypJeD2Az9u9rJKWZdk8EXqJb41mw2QYhmGY9UBLwu2b3/wmXn31VUSjUfz6r/963e+m02n84R/+Ic6fP49XXnkFp06dquzQ78fdd9+NH/7whzh9+jSy2WxLsW7PPPOMeP3AAw+00nyGYRgAwIffvBcffvPexl+sQ66s4R3//x+irBkN67gBFcEW8vvwyG89gES4exn82EWGYRiGYTYGLRXgfuSRR/DpT38an/70p0VmyEbcc889AOwFux9++GEAgKqqIrW/k7m5OfzVX/0VFhYWxHu5XA5///d/DwDYtm0bbr/99laazzAM0zUS4QD+fx84Bo8HaGSHos//9APHuiraGIZhGIbZOLQk3E6cOAGgkjnlS1/6UlO/uXr1KgDg4MGD4r13vetd2LNnDwDgYx/7mGta/29/+9v4oz/6I7z5zW/Gn/7pn0JVVfyn//SfRFzcr/7qr3JyEoZhBsoD+0fxf/3C7Qh7LHhQ+ZPxrPyFAz78l1+4HffvHx1IOxmGYRiGWfu0JNwefvhhpFIpAMAf/dEf4amnnqr7/ccffxyPPvooAODnf/7nxfuBQAB/8Ad/gEAggGKxiH/2z/4ZTp8+bfvtvn378Ja3vAW6ruOTn/wk3vKWt+ArX/kKAOBNb3oTfu7nfq6VpjMMw/SEB7aE8MjYZfxOYg7bg/YaA9szEfzOTx/EI7/1RhZtDMMwDMN0REsxbul0Gn/2Z3+Gj3zkIyiVSviVX/kVvO1tb8O73/1u3HrrrchkMigUCjh//jy+8Y1v4Otf/zpM08Rb3vKWqnIAJ06cwJ/+6Z/id37ndzA9PY0PfOADePjhh/HOd74TBw8exJEjR/C//W//G+LxOP7hH/4Bc3NzAIB4PI4//uM/rmttm5mZwRe+8IWGx3Pw4EHccccdrZwChmEYO9euIeE18QvRZfz8g/uxfOvtKKo6okE/UhF/3xKHMAzDMAyzvmk5q+SDDz6IT37yk/h3/+7f4cqVK/j+97+P73//+67f9Xg8+MAHPoCPfvSjrkLrHe94B7Zv346PfvSjOHPmDL70pS/VdMH0er3wer3I5/P44Ac/iN/93d/F2972NtdJ0cTEBD760Y82PJYPfehDLNwYhumMFXdwAPDs3o10NIB0lOPYGIZhGIbpLm2VA7j//vvx9a9/Hd/97nfx5JNP4pVXXsHCwgJyuRxisRi2bt2K+++/H+95z3tw9OjRutu67bbb8JWvfAWPP/44vvnNb+Lll1/GzZs3oes6kskk9u/fj/vuuw8PP/wwCoUCfu/3fg+nT5/Gr/3ar+GWW27Bl770JQQCPEliGGYAGAYwOVl5HYkAY2ODbQ/DMAzDMOuWtuu4BYNBvOtd78K73vWujhvh8Xjw0EMP4aGHHmr43S984Qv44he/iM985jPYu3evTbRRPB3DMExfmJoCdL3yetcugN0iGYZhGIbpEV0twN0PfD4fPvjBD+IDH/iAyDDJMAwzECQ3SezaNbh2MAzDMAyz7llzwo3weDwiwyXDMEyrWJaF//m/n8TL15fx73/2CH729vGGv1ksqvjij27gqQtzuDxXRFnRMOrZjf0BDT+14MdP7zQQ8ne3TMml2QK+8tINPHtxAdM5BWXNwEgsiNu2J/HwHduwgz3FGYZhGGZDsGaFG8MwTCd85ukrePn6ctPff+LcHP79117HUkmT3vXgphXATSWAp795Hv/9uUn8H++9FUe3JTtun2VZ+K+PXsLfPHsVummvDzedVTCdncX3Xp/FG3bF8Oe/uAPhANe1ZBiGYZj1TEt13BiGYdYDX33pBj72g0tNf//ZSwv4nS+8IkTb5kQI790Vwkdi8/jZcBaZYCW2bWK+iF/93Eu4NFvouI3/8etn8Fc/vCJE2337MviVB3fjw2/ag3v3ZsT3nrtawO996XStzTAMwzAMs05gixvDMBsGy7LwqScn8PHHL8Nq/HUAgKIb+Pdfe10IqPcc34p//c5DiHznW4CyCADIPfwg/uMTU/j+mVnkyjr+8Ftn8ckP3dV2O799ehpffWkKAJAM+/F/fvA2nNidsX3nhxfn8S8//wpUw8Rj5+bw+Lk5/MQtnNWSYRiGYdYrbHFjGGZDMJdX8C/+7hT+ogXRBgDfOT2DmZwCADi0JY6PvvswIjCAmzcrX0gmkdgyhv/43lsxEqsEnP3oyhIm5tqzummGif/zO+cBAH6vB3/xT+6sEm0A8MD+UfzyG3eL///aitBjGIZhGGZ9wsKNYZh1TVkz8KknL+O9/+1ZPH1hHgAQC/pwYne6qd8/c3FBvH7vndsQ8HmB69cB06y8uZJNMhLw4a5dq9u82Ka75FPn5zGXVwEAP3/vDhwZT9T87nvvHMd4IoBj25OIhzjGjWEYhmHWM+wqyTDMuuY7r83gY49dFv9/eGsC/8fDR/Dd12fwoytLDX//n953K/75Q3txaa6I4ztWko7UKAOg6KZ47fe2V9Pt26enAQBeD/BP76tfYmBLMoyPvW8Xdu/eXfd7DMMwDMOsfVi4MQyzIUhF/PiVB/fg5+/dAb/Xi+++PtPU77weD3aORLFzJFp5w7KAa9cqr/1+YLxSRmChoAoh6Pd6cGS8vcySpyYrmS73bYphUyLU1jYYhmEYhll/sHBjGGZdMxoL4jfeug/vP7EdiXAXip7NzQGlUuX19u2Az4ezN3P4D/9wBkXVAAD87B3j2JxsXXTlyjpuLlfi6fZvigGoxLx9/dRNfPv0NC7NFpAt6RiNB3H3njT+8T07EOv8iBiGYRiGWQOwcGMYZl3zxgOjeOOB0e5tcMVN8jOFDGbmU3j9My/ilcms+PihW8bwr95xS1ubvjJfFK/H4iFcnMnjd790Ghcc8XJTy2X8w8s38fWXb+K9R9P46K5d8Hjac81kGIZhGGZtwMKNYRimFa5exYLpw3/NjwL5ou2jdx/fio++63AlgUkb5Mqrxb2Xihr++edewlxeRTzkw0OHNmHXSBR5RccPL87jwkwBFoAvn15CMHq+bbHIMAzDMMzagIUbwzBMs5RKwOwsbhrubpBfP3UTz11ewO++8xAeOrSp5c2TqyUAfOOVSrmBe/dm8IfvO4pMLCg++8237cffPn8df/bd8zAt4PMvXMcb94/gwYNcx41hGIZh1itcDoBhGKZZVpKSjPs0fPGNUTz7uw/h8f/9TfjPP39cZJyczan4nS+8gu83mfxEpqyZtv/fOxbFn//j4zbRBgAejwe/+Iad+ND9q1kn//LJiZb3xzAMwzDM2oGFG8MwTLOsxLdlvCb23roHQb8XiXAAbzo4hk/9T3fhp27dDAAwLeD/+82zKEkWtGYI+e1D8v/ypr2IBGrXZ/vlN+5G0FeJbXt1Mou5vNLS/hiGYRiGWTuwcGMYhmkG06wU3gaAUAjYssX2sd/rxb9992GkIhUP9KWihsfPzba0i6ijiPYDDZKqJMIB7B9ddds8ezPf0v4YhmEYhlk7sHBjGIZphps3AVWtvN65E3DJ4hgP+W0ZLF+fyrW0i3RktVxBOOBFItw4DHk0uvqdpaJW55sMwzAMw6xlWLgxDMM0QenyldX/2bWr5ve2JMPidV5pzVVy71gMJAcVzYRmmHW/DwCGZYnXsVBtt0qGYRiGYdY2nFWSYRimBrpp4n/+q5O4tlhEsazh8U0ehL2oWNxqkC2tWr3S0dYKfkeCPuwcieDqQgkWgAszBRwZT9T9zUxeF69l0cgwDMMwzPqCLW4MwzA18Hu9yJV1LJd0aJYHTyrRSmxbyL0cgGFaeO7ygvj/WxuILjfefMtqSv9HTk/X/e5MVsHlhUpCklTEjwObYy3vj2EYhmGYtQELN4ZhmDq87chqPbbPFEag76htbfu7F67j+mIZQMXa9sD++slF3Hj38XHx+v954TquLhRrfvfjj1+CueIp+ZNHNrdd+JthGIZhmOGHn/IMwzB1+IV7dyLhq6ijs3oIv3/OQlmzx65ZloXPv3Ad//l7F8R7v/aWfYgEW485u2VLHD97e0W8KbqJf/7Zl3B+2p4t0rQsfPKJy/jKS1MAgGjQh19+4+6W98UwDMMwzNqBY9wYhmHqMBr24j8kZ/C/L26GAQ8eOb+EF//rM3jr4U3YnAhhqaThqfPzmJhftYy9/8R2/KO7trtu713/1w8xtVyxyv0vb96Dj/zEvqrv/PbbD+LszRzOTucxtVzGL3zyebzxwCgOb01ANyw8cX4OF2cL4vu/8/aD2JaOdPnIGYZhGIYZJli4MQzD1GNyEg8Fc/jztIHfL2zDkgbM5VX8Py9OVn016PPiIw/txT97oDPrVyLsx19+6E78/ldfx+Pn5mBawJPn5/Hk+Xnb90J+L37t/jG8985tHe2PYRiGYZjhh4UbwzBMPa5eBQA8GCriq+/Ygy/P+fHEuTlcmi0gr+iIh/zYMRLB/ftG8P4T27Ep4Z64pFUS4QD+/B8fx/OXF/CNUzfx42tLmMursCxgezqCBw6M4Ofv2QFteaYr+2MYhmEYZrhh4cYwzIbkIz+xz9VNsYoV4QafD4m9u/Chg3586P7addwa8Y1/8UBL37937wju3TtS8/Mry203hWEYhmGYNQQnJ2EYhqnFwgJQWIklGx8H/LzWxTAMwzDMYGDhxjAMUwuytgHArvatbAzDMAzDMJ3Cwo1hGKYWLNwYhmEYhhkSWLgxDMO4oSjA9HTldToNJJMDbQ7DMAzDMBsbFm4MwzBuXLsGWJXC22xtYxiGYRhm0LBwYxiGcYPdJBmGYRiGGSJYuDEMwzixLOD69crrQADYunWw7WEYhmEYZsPDwo1hGMbJzAxQLlde79gBeHmoZBiGYRhmsPBshGEYxgm7STIMwzAMM2SwcGMYhnEiC7edOwfXDoZhGIZhmBVYuDEMw8gUCsD8fOX1pk1ANDrY9jAMwzAMwwDwD7oBzMblxYlFAMDdezItfdbqtty+R9T7/ueerVhdfum+zl3lWj2edrZNOPdR6/Nm21TrPLw4sQhNj2B3G23txXloFjqeQ1sTtvfFefnRBXxveQw7/Rp+6UT1te/kGNx+K7/34sQizt7M4dDWRMvb7/Tcdtrfu31eOmkD0WpfZ7rHIM55rX2u1+vfyvNs2Gg0Fra7DWY4GNS12Qh9goUbMzA+8cRlAO43WL3PWt2W2/eIet+n73ZDuLV6PO1sm3Duo9bnrZ4z53mg9+8/sLnltg5yUKU21BJunzg5hx+XU4h6TPySi5tkJ8fg9lv5vU88cblt4dbpue20v3f7vHTSBqLVvs50j0Gc81r7XK/Xv5Xn2bDRaCxsdxvMcDCoa7MR+gQLN2YgvDixiB9dWRKvnatutT5rdVu1vie/5/b9zz17FXnFEK87EW+tHk+725bfky0Nbp8DaKpNtc5DO8fUy/PQLPLxuJ4X08SPch4AQN7y4XMXS/ilzfbvtHsMbr+V3/vcs1fF6x9dWWpp+52e2077e7fPSzt02teZ7jGIe73WPodh3OkFrTzPho1GY2E3n/tM/xnUtdkofYJj3JiBIK8UOlcN633W6rZqfa/R91ttQz26ua162252f5944nJb56yZ1822tdvnoVkaHesnvn/O/t6TEzV/3+oxuP222fugnW132rZ+/b5b/aLTvs50j0Gc826OVWuBVp5nw0arY2Gz22CGg0Fdm43SJ1i4MX3HuVJI1oVGn7W6rXrfq/d92foAAHnFEPE/rdLq8XSybec+6n3eTJtqnYd2jqmX56FZnMfj5EdXlvCjmyXbe/K17+QY3H4rW9hoX872NLP9cwtGR+e20/7e7fPSTr/otK8z3WMQ93qte2AYxp1e0MrzbNhoZizs1nOf6T+DujYbqU+wcGP6Tr2VwlZXEZv9fivb6OZKZi9XRRtaj9q02DR6r51jGobV4U6vYSfH0O5vm/nO37280Ha7an23X7/vVr/otL1M9xjEvV7rHhiGcacXtPpMHCZ68cxu9H2mfwzq2mykPsHCjekr9VYKnatu8mduKyfOVdZa36+1T7fv17LKtGN16+WqaDPHVO/zRm2qdx5aPaZhWB1uZG2rR14x8EffPNP2MdQ6/mba08zK84+vZdtqF9B5f+/k2narXzS6FzrdPtM8g7jX690Dgx53ekErz7Nho5WxsBveM0x/GdS12Wh9goUb01faXSl0+8xtldXt+61YNrq5ktnLVdFerCR1M86pmffb3U87dLqfvz95o+1t9/Jad3puB/n7bvWLXlvomOYZxL2+0a5/tyz1g6Abz9BheJ4w7gzq2my0PsHCjekbjVYKG8UfOa1obquszu83uxr/oytL+KNvnqnbhlasbr1cFW3VwtAs1KZ2rVNuxzQMq8OdWNsI06r9WTvWxlZodeW5mXYBjc9Lo/7eyf671S/aPb/rdSV2kAziXt9o17+V59mwHV8716oT7xmmvwzq2jxzYWbD9QkWbkzf6KblodlVx1b2Wc+q0sp+m/1eN+PmukWr58zt9/X+v5nfdJtB+dd3c9+9iJEb5O+71S+62VeZzhjEvb7Rrv9ajuXsxvNuGJ4njDv9vDaWZaFYLGJ2dhafeqrxYvp66xMs3Ji+0E3LQyurjq3ss55VhWjG6tbLVdFeWduIH11Z6sg6JR/TMKwOd8Pa1gztWBs72X6n57bZ81Krv3ey/271i07P73pciR0Ug7jXN9r1f+7i3JqN5Xzh8kLb16qd5/6wHPdGoV/XxjAMZLNZTE9PY2lpCa9MFet6XnVrv8MGCzemL3TN8vD4ZXzi8cGunnQzpqmbcXPDQjPxgrV+06u29IN2rI3tbn+tZI10+263+kU3zu9auJ/WAq2Myb20Qg9iG73GMAwsLS3hLx01JZthGI7PNM3OvW1afO4Pw3FvJHr9rFcUBQsLC5ienkY+n4dpmgCAz52c7el+hxUWbkzP6arl4eoSfnS1O9tql3pWt15muOu1ta1b1MsQWu833V4R65e1jWjH2tjO9jvtY5995kpL58XZ3zvZf7fuj26d3/W2EjsIXpxYbGlM7kdW3X62pVdYloVcLoeZmRmcvLrclGXByaCPT9M0PH76Ok5eXe5oO60+9wd93BuJXs15LMtCoVDA7Ows5ufnUS6XxWcejwcXl9HSPbGe+gQLN6bnrKeVDqKbMU3djJsbFoYh09sgzlc71sZWt9/pue105b6T/XerX3Tz/K6l+2oY+fjjl1r+zTBl1R3G618sFjEzM4NcLgfLslqyLDgZ1PHRMXz2xZmB7H8Yr+t6pNvPel3XhTvk8vIyNE0Tn/l8PiQSCWzZsgX/44WbfWnrMMLCjekpa8VK1CpuVrdeZjhba+exW1kp26Xf1jaiHWtjq9vvpI/99VOX2jov1N876ePtnpd24/va3T7TPM9cmGnLmjJMWXWH6forioLZ2VksLS3BMCr36Zk5rS1rG9Hv47MsC0tLS5iamqpYCq+33/ZOGKbrul7p5pyH3CFnZmZs7pAAEAwGkclksHnzZiQSCZy8uryhssk6YeHG9JT1ssLhRjdjmroZN7eWGab4l7W473p84vHL+PQPr7X/+y5nHG33t704v8N6zYYZTdOayuhWi2HKqjvo66/rOhYWFjA/P2+zMITDYfzfP57rePv9Oj7DMDA1NYXJyUnk83l8+UyhL/utxaCv63qn0zGd3CFnZmZc3SGj0Sg2bdqEsbExRCIReDyerux3reOxLKuJXHqMk1deeQW6ruPOO+8cdFMYhmmXUgn4m7+pvB4ZAd7//sG2p8dYliUefkx3uHLlCnbv3j3oZvQVwzAwNzeHsbEx+Hy+QTdnzWKaJnK5HIrFIuSpWCAQQDKZhMfjgc/nG/pzfOXKFWzevBnXr19HsVgEAHi9Xhw5cgR+v3/ArdsYrMVxyDAMTE9P297z+XyIxWKIRqPweteWbalf14DvKGYgkCuI1+tFOp1ueTJpmiampqYQDAarPvN4PEilUggEArb3C4UC5ufnYVkWEokERkZGXLddKpWgKAry+TxCoRDC4TASiURL7QOAxcVF6LoOABgdHe3qIFQqlVAqlUQbQ6EQ0um07TuapkFRFJimCcMwEAgE4Pf7oaoqgIr7QTgctv2GAuKJeDwOr9drc13w+XzCjccwDGQymYbtXVpaEivJIyMjfZ2IKIoCRVFgWZY4Bl3XEQqF4L90CX5VhWma0EdH4SkUxEQpEAhA0zRomiYETzQabWnf+XwepVIJXq8XsVis6nyXSiXouo5SqQSgMtmJRqMN91Mul8V1KpfL2LlzZ8M2AJV7IxwOi2vh9XrFBLEZDMPAwsKC+P90Ol11nzXDwsKC6EOt9gdd17G4uCgmtZFIBKVSCR6PB36/H/F43Pb9bLbiquXxeNq6jxk7pmmK6zc7OyvGnkEtCCwvL2NxcRGxWAxer1eMs/F4vK2+2Q/I0uB0CaMYnkgkAtM0MTc3B8uyRD939u1hwLIsKIqCCxcu2J4RIyMjtmNrBk3TkM/nxf/TPTvswnUjQ/MQoLLg4JyHyKiqilKpJOYg9AwAKvMRekby4mJ9WLgxA8EwDCiKAgB1b/RaeL1e6LqOWCwmVvjK5TIMw0AsFoOiKFUP7WKxiMXFin9zJBKpue1isYhyuYz5+XmMjY21NeGzLAvlchmWZcHv93d95Ygm7ktLS8hkMq4ClibsxWIRfr8fwWAQwWBQDJpugqtcLotBOBKJiHbrui7OM4kYj8cDVVWbEm7lchmmacLr9fb9IZzL5cQDIxgMIpfLoVQqwe/3I/HCC4gvLUFRFCzpOqI3byISiWB8fBxAZdJP/bSW0K9HoVCAYRjweDxV58k0TSwuLkJRFBQKBViWhVAoBF3XbW4hbmSzWbEoQP/WOnZZiKdSKRQKBfGbkZGRlh6ScrB4LBZra2KsqqpwiQmFQi33B7kN8XhcXFugcnwypmmK/hwKhVpuK2PHsiwsLi7ahH8qlRroRIsW2lRVFcImGo0OrWgrlUrI5XK2+9bj8SAejyMej8Pj8YjzTBNbv9+PWCw2qCbXpFwuY3JyUiwQApX7bNu2bUgkEi31i3K5jMXFRWF5DAaDGBkZWXNWl42EZVnIZrOinzrHX6AyBpdKJdtzh6DF0HafJRsVFm7MwOnEfSudTguLUCAQQC6Xg2EYCIVCVauTcuxAvUGCJto+nw8ej6flVUOgMpmmB1AvBiRVVaFpmminc1JqmqaYHPv9ftEGRVHEuXZzYSFxBsBm9aHvWpYFVVUr1irJelcPwzDEOez34GwYhmgjTQDovJm6jsD0NHRdhycchrVpkzguj8cjLJZA5fhbnfirqioeaMFgsGoCQtdHURSEQiHRb03ThKZprmKcvk8PwGAwaOvXMm6iDYDtt04LYD1KpZJoM1kG2oFEFlB/AaXWb+maUBtmZipZ6zweT9X25HuXJ4CdIy9keL3egU+sNU2DaZowTVOMa2RFHjZUVUU2m60aM6PRaJVVaWlpSXyPrFfDZIUwDAOLi4uYmZmBpmmiD4yMjGDbtm0tL8aUSiUsLS2JZ2YoFBq6Y2aqoYVJoBKPKT+zdF1HoVBAqVSqmkOtZXfIYYCFGzMQujkgJ5NJ+Hw+LC8vw+v1QlEUzM3NIZPJ2AYFebWn1qQYgG2VE0Bbwq1ZkdgO5PqoaZoQGc7jIWsfUJkY0LGT6AKqhZtsBXUKFfquYRhV56cRvTwXjaDjASrXXHYZ9M7OAooCA4Bv/374AgF4vV7RxkJhNbA+Fou13GflQGs3gVIqlYQQTqVSwjpL7a7VR2VXong8bmun/B2naItEIkLkAGhpcmuapnA5pN+288C1LKvKbbOVNiwvr2YwTKVS0DRN9MdQKFTVJtkVh92tOiOfz4u+5vF4MDIyMvD4JVVVYVkWDMMQfandvtkrDMNANpu1LVgAlf6aTCarxkSna/OgxbEMuXguLCzYLC0ejwd79+5tSzAXCgXbfR2JRAbqess0h+zNAEAs5JXLZRQKBduzlwiFQojFYgiFQnx9O4CFGzNwupEfJxaLwefziRXhcrmM2dlZW/C8LNxqCQjDMER7ZCtTq8irqvVEYjvQZEXTNITDYfj9/qoJlDxJiMfjWFpaEr+VrUoysrXNKTRk4UbncS0IN/k8BINBcR4URcHIzAxM06xYrbZtE0I1EAgI9w6gsoLfqmVI3rebQDFNE6qqQlVVIRZDoZB42JXLZVeLlq7rNouTmxUwn89XiaxYLIZsNisWISKRSEv9kizZQGVltZ3zAcDmUhUOh1uakDrbHw6HqyZ8Ttji1h1KpZKtT6XT6a6Pa+2gqqpI7EH3UKtxqL2CJrbkBk34/X4kk0nXRQtFUarO87C4kCmKguXlZeRyOXFMXq8XiUQClmW1JdqcXgGxWKylmFtmcBQKBZt7rKqqtrh+gjwh2B2ye7BwYwZCLwbmcDiMsbExTE1NwTRNFItFzM3NYXR0FD6fzyY4ak3i5BV6WbhRfFaz9FKsaJomXDEDgUDVBIpEAQAR1L60tCQsdUBrbpK0HY/Hs6Ysbs7zYJqmOG+maSJ48yYM00QwFEJ+82aEVs5jIBCwTbbaceeQ3STdLEGULEVVVXH94vG4cJMkFzDn7xpZAQuFQpVoi8fjMAzDZi1pZZKlqqrtt25xDM3SrpskTdCBVVe4Zqx3zsQPTOuoqioWPACI5BnDQLFYFIlpAoFAR32zW1iWhWKxiFwuV7VwkEgkEI1GXZ9/lHSHGJbzTBbDYrGIfD4vPAkikQgymQxGR0dx7VprZUYoNkoezxKJBCcPWiOYpincJCm3gNPCxu6QvYOFG7OuiMfjSKVSInkDpa2Ox+NVljQ33IQbgJaEm2VZQiQGAoGui1Sy1FAbnVYXcsEDIJJc+Hw+m5hwngNFUWxCwznJpYx9JNxkt75GkHCjbfQL2V00HA6LSZGiKAipKnzZLEwAni1bYAaDom1+v982UW0nKYAsUNwEhewmSa5dwWAQoVBInC9FUWwTN1qMAFaDumWcLkeJRELEeWazWXEuyDrdDJZlVW2zXQEkx116vd6mYwapoK+zDZTwBkDNTGTy/cyTh9ah+mLyIsawTK51Xcfy8rLN8jNo102ymMmLVR6PB7FYTGTodYMydcr9edDnmdwiydq+vLwMXdcRCAQQj8eRTqdbTkBC211aWrKNkalUaiiTrzDVWJaF+fl5LC4uQlVVhMNh2zOB3SF7Dws3ZuB0s5QgpbxPp9NikmuaJqanp8XDNBAI1NynPNELBALi/1tpI6WPp210G03ThIudm7ucm1WDUvibpukquupZ2wgSbmSxamYCL1v5+u0mIceYhcNhe0zZ/HzF6hYIQN++XUyofD6fTcRGIpG2hArt280SROmzqT+S+PZ4PAiFQiJuwCnc5FpPzlXMYrFYJbBo4idnXfR6vS2lFC8UCrb7ppPJlSykG2XNdLZBTqhC/bMZ6x27SraPU0yEQqGhsGgRcokRn8830FT5mqbZErcQkUikoaAkISMv9rWTabmbkFukruvi2IBVK2A6nW7LGmhZFhYWFsR58ng8bW+L6S+0cJjL5TA7O2srkUP/DnM21/UECzdmIPRqJUYuVhqLxUT6e03TkMvl4PP56qavd1rc6P9bSVDSS9dAsiLquo5gMIjASkINQs6iGAgExP5lV1HTNKusibIlpFbCCPl80Cp3IwblJkniCKgcO1lBy+Vyxb10ZgbweOD1eqFt3SrcFclNkmhHqDRykyQBQ9kk5WyIwWBQpAOXJ4G0+u3WLsMwbBapeDxuW62XXScTiUTTAsYwjKoEJ53ct+24Seq6LtpAbpqU6bUZ6x0nJ2kPSkcvu5dnMpmhWUGnrIYE3TeDaAcV0JYJBoNIJpNNxQHmcjlbX3Ym1eonZFmj9lA5GYodDAaDyGQybY3ltBBAzycqkdJKgiKm/1B2SFo4zOfzYgEuHo8jk8mwO2SfYeHGDJxuWtyA1eyBFMtDhbcBezpxN5zCjb47LMKN4p9o281Y24DqotmycHNzrXTD6/W2VVCV6Kdwc7pJUqFrRVGQisUQmJ4GPB5YkQi0ZBIBKYMotZnq3rVKI4FCbVNVVbga0XWk1xQ3oGkaAoGA+H86Hrp+pVJJJKkBVoP75bbIbrWtJG4gNzTabifJKJwZS5vdlrMN1Ieatd5Rf/WsiHSmOZaXl4cq7b8T2boViUT67iJJE1h5EgtUxtlkMtn0woRcvJiEzCDcPZ3HQ/9vGAZSqZTwCmhXVBqGgYWFBVv9v5GRkaFIcMNUQwuHzuyQFNNGng/bt28fuHvyRoTPOLPuCAQCYvKs67rI4FUoFOD3+1EsFrG0tCSKa8vI6Y2dVqlmkVcUuy1WnPFtzgdfrdgqWbg5J7HNuEk6aVZsD1K4EaFQCAsLC0LMR5eW4Pd4oAPQxsdtGTblbKDtWNuo8DoAmyCTP5frsJH4lvshCTcAopC8m7WNah/J78uubJZl2SxmrWRr61bNNnl7RCuTWmfNtla3R/ftMImOYSefz9tiKYch7b8MpRs3DANerxfpdLoq1X4voQybzvjJeDzeUtkQTdNs928ymRxIkfhyuSxiwoHKPZPL5RAKhcQ9R1b8dqyauq5jfn5enC+qS8cudcMHuUPK9dkI8gTJZDLCNXmYxoWNBJ91ZiD00q1FFjOqqorgcHL7ACoP34WFhaoVRBqsKH6MaFa4UeZCAK4p9zvFaXGTj1X+LCgl25DbBtjPvfM39R6m7RzLIBKTyG6GXq9XiCkqah2+fl0cizY+jsDKa13XRSIXn8/XlgtPM9kkTdMUddrcikbLkzf6nuz+SsJOLlgbjUar4o/k2LBQKNT08XSrZptMq8LNrWYbXTPZeufz+Wqu2lN9L/oe05hhTftPULIcGlfi8biwqPcaVVVt+wZWkwS14oIMrFqg5Pu338k5dF1HNpu1LXLR84COp9MYNE3TML8STwxUnonDthDAVK4TFct2Lsr6/X7h7TA3Nweg9Vhpprvw3cMMnG67SlImR6p1Rg/UcDhsq5WlKArm5+cxMjIi4qDoAePz+WwP4maFm/xQ7/aER7bWUFISuY31JsfyOZYFWL3abU7IxdJZp6Vee3spYmvhrBUmuyMFg0EkFhdRMk3A64WxdasoFaAoirA4tlNwG2hcdJs+V1VV7MO5yk51+XRdh6qqVXWOyuUyFhcXxTX1+XxVyQycxVFbSf/frZptBCU4ANwXFNyQa7aFw2Gb6Gx0jglOTNIazrT/rbj89Qvqm7QIEwqFei4s3QQOUOmXyWSyZRFC8YN0jwWDwb4mfanl5kkLVzQGUlxju5YxRVGwuLgo7sNAICCetczgqeUOSVB2SBp75ZjSehlSmd7Dwo0ZCL2cxJN1h1YP5YcTDUQ06GiaJmq9yXRDuPUiMYkcJ1Qrvq1WJkOCjstZB6vRJI0EIz3gna4UTobBTTIYDGJ2dlZk4YyUSggoCvKGAW3LFvgjEZimKSylHo/HNdV+MzSqKyZb/qjwt5tVDqg8NHVdF9c8GAzC6/XC6/XaRFskErGdZ0KuIdVKpq9u1mwjWlkcoDbINducbZAXKOpdJxZuzeOW9n/YVtQ1TROLEbquCze+Xgk3chmUs7kClbGsE7fG5eVlW33JfiZ9KZfLWF5ernLzdGYaDofDSKfTbd83zsWlYDA4dHGSG5V67pBer1cUy5b7g6ZptszEXLrBHcoQ2+ussCzcmHVJMBgUoq1UKiEQCAgxRnERtHpLtd7kwcgp3IYhpouyY9K2nS6h8gquc1XTNE1hhSScdbAaPVRJuJE7YSPL2yCEmzPGjLIxUrHy5Nwc9JXzpI2PI7hynsrlsjhn7WbIUlW1bl0x+pyKbtcqGg1UhFuhUBBJOMhS5RRt6XTa5toGVK6THKPUbHxaN2u2yTQqkt1KG8gKCayW/qgFZ5RsjmFP+w/Ya/mZpilqTTqz6nZrX24FtCnOspVSFk4oOx+wGj/Yj75JNe9kywpZ+2mBiOi0EDaVJpGTQw1TRtKNSjPukJFIxPV+kr0+2o11XM+Qh4uiKCgWiyzcmPVPt10lgVWhQA8lOTU+UBmoxsbGRKYr0zSF/zZNCuTBqVWLWy9iuurFtzWKIdJ1HV6v1zaZbTUpCQk3in0YRuHmFE8Uk2KaJiKRCCKzs+IcqFu3Iryy2lwqlYRwb3c1sVHRbdlNkq5RLSFD17ZcLsPr9ULXdVE3h36XTqddH6C5XM6WrrnZiWE3a7YRzpi/Rm1xtsHZL1uJlWOLW2Oortawpv0nisWibTyh+6bb1jZnog6gMpbH43HE4/GOzgsV5ybS6XTPx0VKUESLV0QoFBI1I+UEPul0uqP0/Pl83naMtLg0bP1po0ALmYVCwZZ4iwiHw6JYdi1UVbUlqmrHG2W9QgvD8n3UD1i4MQOh1wM5PdApviYSiVQV3vb5fBgdHcXi4qIoulwoFBCPxzEyMiKyL5IrXSPkxCQUZ9dNSqUSDMOwFb0EGrvoAauiiyb/mqbVdbt0QoW35Tp5bi56MoMQbrKbpN/vx8LCAoDKtU6Fw/DOzkIzDJiJBLwrq2IUE0cJSdoR3E5LXy3hRnX46rlJArD1O3LzJCtIvRVsudi2z+drWnx1u2Yb0YqbpFyzDYDrhK9d4cYWN3eWlpZsbnujo6NDJ3INw6gSAzR2dUu4aZpmc2EkKPFIp/1H13WbtTwej/c8ftAt+yWVK6C4PaIbSUOy2awtrtaZ5ZbpH+24Q9bCWQeURfiqVZ5KZsj0w8WchRszcHphcfP7/ZXiyppmE1POBzPVk1laWhKTRkqHTe6DzQq3XgoVGogBVGU8dFqZnBMv0zRF6myagMgT5GYmEPKKPG2jnsVtUIlJZPFULBbFoBqJRBCZn4euafB6PFBXim5TUW5qZ7uDbjNuklQcndy7Gq1sk7WYFhNo2/UsIu0W2+5mzTaiGTFbqw3xeLzqHpLv5Wasd844HsZOLpezLfhQmu9hw9k3u5kAikShMytlKBRCMpnsyjjudEUNh8Mdl9eoRy23yHg8jmg0aiuwTe3pJJ6NkoDJoq1Td0umPZpxh5QXfRuhKIqt/NCwJSsaBKVSCblcrmr+E41GoapqS4nA2oWFGzMQ+jGRDwQCwsJBMU5u7gI0aZEzh1GNLGqnaZo2VzU35G1324WnXuHtZtwkAdjcPwuFgi2mqxG0DY/HY7Nm1vs+PTj66SYpx/nNzMyIz9LpNHyvvQbNMBDw+6Fu3Yp4MCjiHMnq2O51a9ZNUlGUhm6S9D3TNFEoFMTiARXArdUH5WLbgUCg6YdsuVzuas025zEA9YtkA/VrthGtJjlhi1ttKIaLyGQyQ5X2n3D2zXg8Lu5reRGpVSgmxelC6Pf7Rd3PbrG0tGRbxOqV62CtY6LslwAwPz/f1Xg2ij2UF0lSqRQnr+gjtECmqipmZ2erPm/GHbIWbG1bhVydnZ5GcnZZOT67l7BwY9Ytfr9fPFAaiS6gMhmkxA5ksSmXy6JAcqNt9CsxiZxSXXaT9Hq9roMzPagpNo0ETiQSQTgcbmryIz/sQ6GQKJ1Qy1o6aDdJuVB5LBZDOBSC58YNGLqOQCQCc8sWeDwelMtlkeSi3clGs26SlAgnGAy6JpCRWVpaEtkdqf/VE22WZVXVXmvmIeusl9aNmm1Es0LLWTfOzU2zVesdwBa3WiiKUnXNuylUuoVb35QXhNoRmjReyuUmgEr/SCQSLVkjmiGXy4l+S94dveiLtdwiU6kUwuGwqPvYzXg20zRFmAGAjmu+Ma1BC3vkWeLsz9FoFNFotG33V8qADLS2ELjeoJI8zpIJVKB+EAteLNyYgdMLV0knzTyMScxQxj9yASmVSkilUjBNs+5DlwY5r9fb9cQkpVJJDMxyjTHKOgi4u+gB1cKtVCoJ0dDsYOwUbjQZGUbhRgHDxNjYGPQbN2CtvGeOjyMYDkNRFGiaJhK9tDuJaeQmSS5+zbpJlkolzMzMCCsbxdlomlZz1VSOZaDEA83Q7ZptBNXFA+oXyXZrg9u5aVTYvFYbgNV+z1THWsVisaFL+0+49U3ZStjqhMltxdzj8Yhz0G1BRS5VRCaT6UnCqmw26+oWSclUcrmcrR2BQKDjtpD7p+xlMjIy0naJBKZ56rlDUkKnThcgnAuBG9HtlWKunW7UgUAAiURioItdLNyYgdCPiZQ8qDUSh+ROCVQmM7FYDIuLiyJ74tLSUt3gbbKmAL0RKnJtLXly3UyyBtlV0rIsqKoqJr/NDj50bGTVGzbhJsc/ySnjKZ5k6do1aLoOeDzQxscRCoWqig232ycbXQPZTZImNrXOu67rmJycFKJjZGREPITl38tQQV35WJqhFzXbCHlBoZ6bZLNtaCUpCSH3WcY91qof8RjtUKtftOOO7iZugEo/SiQSXRdTtE/n+NJNUdPILdLv9wuLmLNgfaeumoZh2FwuvV6vSLbE9IZmskMqioJNmzZ1ZX+lUklc304WNdcilKjLKYz9fr8oBzJoWLgxA6dXFjfDMODz+UQmxnpCzuliEgqFMDo6Kty9qNbb1pWkFk56KVQMw7Al0KBBtFmrBrXN7/eLYrKGYTS9KicnGvH5fLaJjlvSFrJU0j77MXGm82MYBhRFEcW0KeGC9/r1SoIWjwfGjh3iXJA1pt0JrNOFz23yQtZSTdNEwgO3yaKu65idnRV9LhAIYPfu3aJMhXPiKf+u1WLbvarZRjTjJulsQzKZdG2DfI6bXWywLEvc4xzfVp32PxAIDG2a9lp9kxadgOpxyA2agMl9EahMRJPJZE8Ld/eymLmbW6QzNo8KqsueEslksuN26LqO+fl5sW+fz4eRkRHcuHGjo+0y7hiGgWKxaEu0RZA7ZCwWg8/ns1lVO8G5ELhRrG21FkN65UbdCSzcmHUJPeRpEtuo7phbsd5AIICxsTGx+kQPrUwmUzV57Fd8WygUEhOWZtwkSaRRu2g7VNesGZwTBHnC5Ca6ZetlL1az3aCJPQlToHJO4vE4rEIB5orrIcbGEEilUCgUoOs6gsEgYrFY29dMTsDhdg2o3yiKUtdNkvoWPTT8fj/Gx8dFPBz1AVqMkH9H16eVYtu9qNlGUPZMoHL9a53bRjXbCGeh+Gbdngm2uFWn/e9VrFU3qNU3NU1rKr6NXKXleobAair8Xq6Yk0CWkyR1y5LtVrLArcZcuVy2ucN6vV5kMpmOLWKapmF+fl7ci36/H6Ojo7ww0gPI4iw/4wm6JzopBF+PYrFoy9673i2ptWqxeb1e4UY9LIKNYOHGDIRe3wgUxO73+8XAV6/umJtwAyqDZCqVQjabFav4CwsLSKfTtolmLzNKyitA8gS7FTdJACKhCNBaHJ68DcrkRpkO3YRbv90kqVYfiSTLssSgGw6HYb7+ujgGY/t2xIJBTE9Pi7Z3sqLodEOq9bmqqqJfOIWb7HpULpdF5jlqVygUEv1LURRbv5PjEJottt2rmm2E3C9ribFmara5ba/ZSTcX315FTnfv8XgwMjIytJNtZ7+Q+2YzY6ybNcrr9SIej9tig3uFLKx8Pl9XipmbpimKaMuEw2GkUilxLanYtmwt6UY8G1AZd2QrYiAQGOp+tBZp5A4ZiUQQjUZ7KqQ2krWNEhXJsbRAb+NeuwULN2bg9MJVUnbVo+3rul7zRqwl3Lxeryh+LE8GKQUyDWxyYpJuP8xkVx8SbuQSCFSOsdZERhZdmqbB5/NB0zSEw2GRBr8RTuFG/6qqKsSsPDnpt3CTrW0+nw+KoogVSa/XC/3KFXF9je3bba6foVCobf99pwtfLTdJch2Nx+NVFigSbWSlkjOz0bkOhUJiMisLN0VRqlKlN0MvarbJNOMm2ahmG9FKkhOZWvfzRoOKxBKZTKZvyYLaoV7frCfcTNPE7Oxs1eJcLBZrqZ5hJ1CGP6B7dfGKxWJVBky/349UKmUbb5wZHoHuxLMB1Ra8YDA41BbbtQa5QxYKharQA6c7ZK+Rk1yFw+GhLBHSDerVYut22EAvYOHGDIRer3zKrlq0r3pZ+eoJNwAiDsrj8YiJEK3UxONxMeD2MjEJDeJAY0sPQRMZEisU0xUMBrsi3Ohz+bgHIdw0TbO5I0Yikcp5MQwYV69W3DdDIfi2brUFXofD4Z65SRqGIdrl8/mqCqc7g/xVVRUWBtmySsdEIoaEcjs1dnpVs42Qk8TUKnnQTM02ub3NJDlxwha36rT/lBp+WGnUL2i88Xq94p7VdR3ZbNbmTg7Yk3T0A8pYSaRSqY4mvbXcIhOJRJXlUNM0LC4u2mptJpPJrrg/F4tFW5KVcDjcFSsiM1h3SDcoxotYj9a2erXYEonEUC9qybBwY9Ylzoc8xQLVSulfKyZGfm2aJjKZDLxer3hIU+AwTaa7vUIlT0jkmmvNuo/Rw1wWNZRR0hns3GgbgF24yZ+7CTcSK72ELFWFQkHU7aPVyXA4DNy4AXVlFVzduhXhlTg/0zTh8/kQCATaHqwbFd2mzymLp/w9p2izLAuxWEz0V3mBgZKeyElOyD2UPm+miHova7YRjdwkm6nZVmt7rcQmbXSLG03mnZlyhxVn30ylUra+SfcsUFkQIPdBOaYVqEx4u53BsRHOEgvxeLyp+9GNWm6RkUjENXlPqVTC0tJS1+PZACCfz9vu1Wg02nW36o0GuefJcZwE1afshRdEM8gWv0gksmZETDPUyizb60RFvYKFGzNwuu0qaZqmmBCTG6GcKt7tJpWzZMkPJvk1DWoUS0QPTBqEU6lU1wc7t/g2Od19rQyFhCzcQqGQLRtbq8JNPjfyPjVNE5NquRBov6xtJG7Jukrujx6PB7h6FdrKudK2bkVoxbXTMAwhYNuZ2FuWJR4CtdwkaSVVVVVEo1Hh6kdZ5+Tz6vP5bAXDnZBwo+3KgqbZ89yrmm2EXAy+VpHsZmq2EU534Fb600a2uBmGsWbS/hOyO6Bbv6B7gzwHqM4hIbsY91NYWJaFxcVF0RYqytsOzbpF0n7d4tm6FXeWzWZt247H40Pfh4aZRu6QsVhMPCMGARXzBlpLcjXsDHMttk5g4cYMhF4+XOWVrGAwaJvwkXBzlgagwdQ5cJJrofwdACJ+anFxEZqmibo9Y2NjXT0WeeWVJvTNWiFIoJCY8ng8CAQC4tw3I9xM07RlESOcFjdCPvf9cFOi1UuC3IjovJgTE9ANA/B4oI+Pi4yXpmmKvtFOX2zGTVJVVaiqKhLBhMNhmKaJ+fl5m1Uyk8lgfn4eAISbpxN54ra4uCj6dDgcrpt0h+hlzTZ5H/WKZLfahnatbYBduG0ki5szqyElpxhmK4mqqiI2jOKJ3b5D/UeOiaSsipqmtW3l6gQa/4HKeNfOuVZVVbh7ErXcIgH3eLZuWcOoFIMcp9qNMgIblWFzh6yFnFFRjq9eqxiGgXw+X2WR70dm2X6wtq8Osy7otsXNGcTuFG5OGrlVkVuhc6WMar3RpJseqCMjI10zvbslJml2QksihTIVAhW/dVpJbUa4ublJAnbrWy3h1muLG7ndUeweWb5EEovlZejz87BME/rYGMwVi5yu60IIdcNNslE2SdpXMBisEm1UK1C2qro9xMlSqqoqlpeXxcp6MplsWL+n1zXbiHpuks3WbKu1vVYftHKJhI1icSPrj9y/RkZGBj4prIdlWbYYKre+qWkapqenoSgKPB6PGIfkRAILCwv9bDaAivVYTk7UasIOGr+cdeZquUUClXMhC/NuxrNR/5Hjp53Zk5nGDLM7pBuGYawba9taqsXWCSzcmIHQL4sbuRLSA9XNOtFIuFFb3YpNk+VgeXkZPp9PWFTcar21CqWHByqWFb/f31TyB4LqHimKItoSi8XEoNaqcJP3JU+gSCB6PJ6+CrdSqWTL4kaTe7GCefVqJTbGsqBs2SKKcpObJNCeVdCZTdLtAUwrrIqiIJ1OA6isajpFm8/nsx1DvUlSOBwWsTSapiGRSDTV/l7WbCNkN0k319Fma7YRZMUGKv281eskl73YKLgJiWG3Nubzedt4JvcLwzCEy57sMkuun4OMwSmXy1XlLJrto5ZloVgsIpfL2Z4pjeLzisWiLesmWeu7IQCcVjzZ9ZRpDhJAxWJxKN0ha5HP520Lh8PWvmaoVYuNLPIUP75eYOHGrDvkxCT0MCWXONM0qwRLMxY3ADVT33u9XqTTaVuiicXFRaRSqY5WK2VLDG2nFSsEJbCgos2U3MTn89kKNzfaBuGcmND/kwgkYQm0VieuXZwuYTSBofNiXblSab9loTg2JgS83M522kiZHQF3N0nTNEXcHVkCZbcmr9eL0dFR+P3+qoDweg9Nr9crJua6rje1Mtrrmm1EvWLwrdRsIzqxtjnrFW4EaNICrKaiH/bkArqu29pMfZPibWhCKd+rmzdvxujo6CCbLRK/EMlksulFOrKYywtcjawBlmUhm83aXMKDwWBXyg0AqHLfplp/673wcrdQFAXFYrEqjgqoXCeqJzqMlh5d120Lh2vNJbZeLTayyK/HZwALN2bgdNNVUtd1W/YxQn7tdJdsVrgBENkICRKJHo8HW7ZsQalUErW7nLXeWkUODne6ScoWplrIFju/3y/EHwk3muDWG9jkc1NLuAGrAlYWUr1E13XheiefC5HEQtOgX7tWSWYQDkOLx+GD3VIItFcsvRk3SbK2BYNBZLNZce5l0Qa4xzDWglzFqLB8Mw+kXtdsI2R3L+dihbMNzfSNRklO6kELLMDGiG8rl8tVGRnXwqTb2S/8fr+wRMnjDpVcoZTdg4QSC8klKpqZ7NZyi2xUN6qX8WxAdXZbstQOixvfsLLW3CFrIVvbhrngtBvlclmESsislVpsncDCjRkIvVp9kuPb5AlirTpjQGfCzZkIJRwOw+v1igk5TULaedDKfuexWKxh8gcnFNBPYoUmc3L7DcOoux25NpDz3DgTlMjb6bVwk61tsiudEFKTkyKbZHFsDD6fD16vF+Fw2CZmW70mrbpJAhBJYUi00blRFMVWCLzeOVMURZR0MAwDwWAQmqbV/U2va7YRcvZHZ5HsVmq2EXI/b+QOXKs9xFqaiLSD0/rTSSr6flIsFqv6zNzcnG08pXGPahj2otxKK5Anhbw4RW7Q9X5Tyy2yUa03VVVt+yOrZLeur67rmJ+ft2VUlheVmGrquUP6fD5Eo9GhdId0Q7a2kSvnWkBRFORyOdtcD1h7tdg6ge9QZl3hFFJurzu1uBFONx76XiqVgs/ns9V6oxpwzQoFXdfFxIbcAOWaOo2sbeTzTdYZOemFU7jVG+jcSgEQTuEmf97LwdMwDDFZdRarFufl6lXoK9cmt+I2RqUCaIWxHeFWzyUQgCiSrWkaCoUCotGoSJgiizbA3aLqBrlKAZV+TJlOSci50Y+abUStYvBuNduaaUMnbpK0X2ItTKDahdL+y9aftZCyXe4XNL44k4tEIhFRVJ7uk3YzwHYLuQ5UM4lf2nGLJHoZz0Ztk0tG+P1+EXPLVKMoisgO6WTY3SFrIbuvrwVr23qrxdYJLNyYgdNNV8laFje5XpfsTgmsCjeKR3Iivye3ldwNnfsCqmu9lctlzM/PN515rFwui3Y549uacR8zDEN8n1YCCadwq7cNWeQ4qVUSoNb3u0UulxODt2x5lC1v+qVLlXp+lgVldBQBVFtv2mljLZFCUJkAyjRKbRwZGanKbionXKjn2lYqlcTkLxqNimuiKEpNN61e12yTqeUm2UrNNqKZWnCN2AgWN2fa/2Aw2ND6MyyQexMtaMli0zkJk0X8ICdmVIMLWI0hrCVyKK7UzS2y0QJKr+PZgMq4IQv+QCCA0dHRdXuvtAtZS4vFoqs7ZCQSQTQaXZOCQdM02/xgmK1t67UWWyewcGMGQi9Wpqg4KwDhmiZDE2f5e4C9+LYbtSxujTIoyrXeKGHF3NwcRkZGGooG+cEdjUYb1g1zUiqVxDFSRkqiWeFWLzEJULmGckkAmgg448i6ia7rQgxTADK1QYiTuTnoKyv6hXQagZX3Y7GYyLQJuFsR69GMm2SpVBKxLHT9x8fHq77rVnuu1j7lldHR0VHhPqWqquuiRz9qthGUAAewF4Nvtw3Oft7OZHK9F99ei2n/CVrAIlGTyWQA1K6v5CztMgjIckbUcnHsxC0SWPUkkI85FoshmUx27dqWSiUxfgKVRaVMJrMu75N2oUWFeu6Qaz1LoewJEY/Hh3LsqFeLLZFIDEX9u0HBwo0ZON2yuMmTcjch5RbnJluVui3cgNVab7Q6rus65ubmqtzmnNCk1+v1IhqNtuw+Jk/2nXFF3RJuwKoAp0LfXq+3py5NsrXNOXCLlTcqA2CaKIyOCktbNBrtqEC47Cbp9tAwTROzs7PiQRMIBLBp06Yqa5ppmrbYgnrXM5/P26xWwWAQoVBI7MPpNtKvmm2EW79sp2Zbve21SiPX57VONpu1LSCsFWtJsVjElStXhDCJx+Pw+/0iXbfbmDFo4eZ0R6WU7k4URUE2m23LLRLofTwbUHmmyPdlOBwe+uLs/WQ9ukO6oaqqzeV32GJinVllifVWi60TWLgxA6EXN16jh3wt4UbUmuTJbZWFWy23TLf9jo2Niexdpmlibm6uZq03TdPEtp3xbW41spxQfBu13elO103hRhNGEsDOgufdRNM0FItFqKoqLF50bSiODACMiQkYhlEphbB9u3C58/l8IpU/ucW20g9lUeG8bpZlYWZmBuVyGbqui1V2t5gjeQWRrHJu0IojsFpol46VhJ9TuPWjZpuM7A5GQqvVmm2E06LZbmbE9WxxKxQKVS57w55MQlVVZLNZmzWJxsR66bpN07T1o35fS3JHpf4UCoWq7meqNed042rGLZJwi2dzulZ3Si6Xsy3mdTMz5VqGrKSFQqHK3Z/cIZvNhLtWkK1tFEc6DGy0WmydMNwjPsO0QCMhRZYgOalIM8LNzeImu1s2k5rd5/NhbGwMCwsLwsWtVq03Sm4BVCYLshtiMyt+5XLZ9nuniKUMkWQlq0UrFjf5u716yNGKtmmawr2jyk2yXIY+OVkRAdEofKkULF1HJBKx1fdqp6BzrcyJZGGi1WxVVbFlyxaEQiFXgddsCYBcLmer40dtDoVCoh/Lwq1fNdsIZ5ZTKjPRas02opFFs1nWa3KStZb2X45N0XVdiJtQKIRdu3Y1FPSDtrYtLS3Zkk/J1im6j+V7lNrZbKIEGjfkxY9euC5SEXMiHo+viSQ2vUTXdRQKBZRKJVd3SLKsrjehoCiKuK/8fv9QFFjfqLXYOoGFGzNwuukqCUC46zkhwaLrurB8tSvcZLfMZicV5Na0tLRUt9ab/DCJRCItu48Vi0VxXLVckGTh5iwqTsi1fWoNnP0SbpQOX1EU+Hw+hEIhm1uREEjXrkHTtIrVcvt2+P1+EcvmVpOuWXFQr+j28vIyCoUCVFWFaZqIxWKiho/zvMlJZ5yxhzJkXQRWXUQI6t+qqtpEfb9qthFyv6RJeDs12wg36107yP1iWFaTO8WZ9p9choYR0zSRz+dFVlug4vLr8/kQj8eRyWSaavsghVs+n7clyZHFlKIoWF5eto15Xq8XyWSy6WvSj3g2N2GYTCbXXJHlbrJR3CFrMWzWtlq12Cir7LB7EwwKPivMQOj2gCFbjxq5LdIgoWlaR8JN3mazeDwepNPpqlpvcnY1et/n8yEcDtuyPzVaYaeVbao3V2si4XSXdA6QlmWJc1Nv8JSFG02UezHY0sq2qqqIxWK2NstJLMyJCZEww9i+HX7J5U62lrYqLmuJZ5oYkZCneLpaGRGbtbY5g8edAjAUColJn2mafavZRrhlf2ynZhtByXvot51M1Ok6rxdrm1va/0EXonajVnIOVVURDodFrcJm2z4o4UaTSSKzUk6klltkLBZrySrgTMVPz4RuWj/Io0MWKOl0emjFfi8xTVMUy94o7pBuyNmJA4HAQK1tXIutM1i4MQOnGxa3Zh/yNNknEdBsjBu5pnUq3Gh7zlpvhUIBhmGIzIe0XdM0bRO2RpRKpaYsOo2Em/yAqzcB9ng88Hq9MAxDiLZui/JSqSSsS1REW9O0ajdJ04R++TJ0XYfh88E7Po5AILAa+9amxc3pJknXmyxtQKX/JRIJFItFIRSdwo0KogOVa1tLhJfLZVupADeBFwqFhEuiYRh9q9lGOLM/yrXmgOZrthFkgQY6s7bJ9+h6cK8xTXNNpP13WzknVyeZZt13nTUy+yXCKWstkUgkEAqFkM/nXd0iU6lUS+N/oVBANpvtaTwb9Rkaa8hiuNHSpm9Ud0g3nNmJB7Xww7XYugMLN2Zd0KyQkgWK0+JWbwAncUIPXNqfx+Np+6HrVutNzpooWweB5opuF4tFYf1qRbg5aSa+zblvy7K6PsGSHziKoiAWi1VNzoUAmpmBtuKyqI2PIxSJIBKJiGNpt76XW9FtudaSZVminpxhGAgGg1U144DmrG1OAVTLnYXcMCmBg3zN+7GS6nSTbKdmW63tdcNNElj7Fjdn2n+/3z90af9rTcQoNkXuy5FIpOmYPLnURb8mcyR4ZDf1YDCI2dnZjtwigf7Fs5F1Vg4byGQyQx0L2W3K5TIKhUJVnwQq5zwWi4k44Y2CXB4oGAz2XcTXqsXm9/uRTCY33KJCp7BwYwZCtwfNVixu9KCUf9OophdNyskCJk+mOjkWZ623QqGApaUlRKNRm3Dz+/0NBaKiKMJllCb27Qo3N+vUICAhCkC4IiqKIs6FHI9gXbkizoG+bRsSkQj8fr+IA6NskvI1a+baOYtuO4P9Y7EYisUiSqWSyHTpfBAZhmHLmFhLnMjHGwwG64qYUCiEbDYrykyQBaDXkGsmsCqAO6kbp+u6zRLZifVhPWWUXF5eFpNPr9eLkZGRoTmmWkWmKfNiIBCAoihioub1elvqF/12kySRLMf1mqaJ+fl52/dadYsEqsUUUFm063aMka7rWFhYsB1Do7Iz6wV2h6zNIK1tXIutN7BwYwZOp66STiHV6KFKQoSEmM/na7g6T9t0ZvLrxoOAar3Nz88L4UEuj9SuZt0kgcoDnL7fD4ubfP26OQhbliUEkqZpQgyRAAPs50W/dAkKuQft3o14PC4mlrqui+NoxRLjdJNUFMUm2tLptOh75XJZtMd5veREDbUSxpimaXvANsr8FgwGRVtUVcXo6GhfrExOC6TTQthqG7plbQPWT0ZJmuwAlXtqZGRkKAL16Z501lhyrpy71fJrRez0W7iR1ZA8H8LhsK0vteMWCVQW02hRDuhNPBtQGR9ll1qfz4fR0dGh6DO9hNwhncIA2HjukLWQk5WFQqG+WF/r1WKrV7eRaY71fVczQ0s3b9pGhbed+6SHmWma0HW9JeEG2GtndWtSEQgEkEwmcePGDbG/hYUFjIyMNLS8AHYriJwko9YDq1fCrZtQ3B8AET8nZ8C0JbHI56HevAlD16GPjCCUSiESiQhhYxiG6Bvy8TRTWoGOzy3dfjQaxfT0tMhQSnXs5PNLLqy0v1ouVnL9GnLTqgdlRyXrQK9rthGy0JIXTdqtG9dN4dauO+wwUSqVbGI4nU4PPPajVsruWkVxc7mcGENCoVDLboVyoppeiw+q46WqKgqFgig1QvtPJpNt9UtnXByVFOi21ceZ7MTv9/dtEWdQsDtkc8iLn0DvrW1ci60/sHBj1jyyC0ozExxK+kHxQXLx5lrIg02zhbdbRdd1RKNRFItFeL1eEe/UzGo7rThSxsRG2R2p+LScPdLZFqCxCymwKtx8Pl9VEHi7UEpxgo6lXC6LFUObm8XVq+Ic6Nu2YfNK/TBqj9wuv9/fdDtJDBeLRZsbXyqVEolkDMMQ7puUPEWmWCzaBJlbXzMMw+Zu2OgBaxgGisWiOBZKENLriQodK7XTmbWu1f1TfB4A17jAVlnrrpKqqlYlxxh0rSVFUUQNRcLj8SAWi7lmPNU0rWPX2X7Ft5HooQx38XhcjJ+1jq8RVOZFXpAIh8Mim3A3KZfLWFxctJ2vYXKp7SamaaJYLNrcyQlaEKMQA6aCvPgZDod7dj81qsVG8fxMd2DhxgycTi02rbrV+P1+kVFSFij1cFrcSPh0czWY3CTj8bjI9Eiumfl8vm79HdlNkoRDo7ZRTTuncCNR28w2TNMUmSUpnqwbyCt2wWBQXONabpLaxYvCTdLauVMIH+pbThe6ZoQbuUmWSiWUy2VhNUgmk8KyROddURRx3t3cJIlaFik501wsFmt43qleGp0biinq9SRfnoxqmiba2W78iFstuE5Yy8lJKEZJLro+yLT/mqYhl8tV1bxqVGNJruUXj8dbHiN74dHghq7ruH79unDniqwkM5Lj9NrZppxQBqiIb9mK1y1KpZJIbAVULE3DlrymG9Rzh/T7/UKwrUex2gnOxc9ejSWUVE3u8wDXYuslfEaZgdBtV0naZjODhNfrhc/nE6v9zWRDpPbKVjpame0GVMeKXPqCwSD8fj9KpZItCYVbgVZVVcU5kN0jmxVulKWRfteOmyS5NFHcYCcPUaf1iYQ2iURql5hY6TpKFy/CMk2YkQjie/YI90GyQspCVG5bvetXLpdRLBaRz+dtok0W0OVyWfSJZDIJv99vO2eKothcxtwmg5qm2ZI4NCqQK9dsk2Nx5Bi7XiELVTlmsJ1JgVstuE5ZqxY3Z0bDUCjUl0QztdqSy+VsCw5Acym7yeUQqNxr7RR77kd8W6lUwpWVZEa0n2QyidSKi3U7uMWz9SoNf6FQsMUQRiKRtizewwotWLI7ZPsUCgWbp0cvXHSz2WxVLbZOFj6Y5mDhxgwMctXrxOImW81aEVJ+v1/EL8lJQGohixoaDLs5MNEEnyxKNIGXi0uT20Mmk7Edp5zZTXYDbEa4EYZhdEW40e87mXDJcSGxWEw8uBVFsblJijZMTqK4EhOkbd2KzZkMALi6SVIh3WZYWFgQK5bBYFCsnBMk/Ou5STqzT7ohT8AaZawzTdP2/dHRUXH93SY43UTTNGiaJiyRdC1ardlGyHUUqZxCp6xF4ebMaEixUP2ekNaKT2k2zsstBrSdY6CJoNfr7frkT9d1ZLNZzM7O2pIOjY+Pd1T/MJ/P2+ISqXRDL6wNuVzOdp5jsZjrgt5ahNwhZRc/gtzumvFI2OjI1rZmXO9boZYlnp6RG6n0xKDg3s+saVqNbyMozo220Ypwk7fRLajANFAZdOlYxsbGAKy6H5XLZczPz4s4BnoPsFungNaFGx1PK8JNFondEG6apgkhQkJITjDi5iZZPn9+1eK4d6/4jNpGCTyAauFWa7KTz+exsLBQ2abPh0wmU/Xwo/OuqqqroCRRB1TOj9sDrVwu265XI3dBZ7002UWILH+9Wukk6xjF+1Eb2rUodDMpCSFn1VsrE9lhSPtPCVGc90YikWg6A1w2mxX3XDQabWsCJy+MddPaRinRyeWO7t1gMIg9e/a03f/6Gc8GVPqKbAlNJBIDdaftFhQXWSqVXN0hY7GYKJ3DNEbO5hiNRrsidLkW2/DAwo1Z07SbKEQWbrIgqIUs3Oh1NycWiqII8UEiitzHyLWT4l9UVcXc3BxGR0ehKIrNHUKeADYSo7UyS8rCrdE25KB4N3HbKvJKcjweF9fXed7lB1HxzJnKC68XicOHxfuycJNdLKnNtSajpVIJMzMz4nuZTMY1NT9ZbDVNE8HXch90Wtuc+3MW2260al4r6YN8jeQad92E3Bo1TUO5XMbo6GhbiSfk7cm14Lr10JcLs68F5Fpog0j7X8vdiRJzNBsnWC6Xbe6+jUpZ1KIX8W2yKFVVFfl8XqSK37p1a9uirVY8Wy+ElJtAdLptrzWadYdkQdAazlCDTvsIuU5zLbbhgYUbMzC64SrZisVNHlxk4dZMogpZlJCo6lbyA3L3dAom2X0sFAphbGxM1OrRdR1zc3O2cxcOh8UksJnJXy3hJqfgr3eMcva3UCgk2t+ucFNVVUzmaWJFBXBrJf/QZmagrljGjC1bkBwdFZ9R22QLYCNRUyqVsLi4KCYSkUgEmzdvrvqeruvQNE3Eevl8Plu7qCAsULvgtrPYdr0JCk3cCLlemixSyuVyTyZzZBHO5/OiyHg7NdsIig0E7EXUO4FiGoG1IdwoCxvRz7T/tVbP24lPcdZsa9d1FuhufJuu6zZrJrlyUjKLdtP8A5X+u7S0ZFsoSKfTPREZ5Eore1ZQKZK1SD13SBor2R2yfWRrWywWa3uM5lpswwvfGcyaRq7308oARQlJaLLXKKEGJbugbI/dTExCVjPDMGzp/J2TikAggNHRUSwsLEDXdZE6nCYgcvubeejJ36EHqGVZTZcCkEUz1UyTf98qzkLOlLAFgKi350xiUTxzBgbFOO7bZ5vsyRY3EsE+n6+mxY0mY2TVpIB/t0ms7CZJ+7S1S1qddHPxcRbbbmS5KhaLNeulkYssiclOk8O4USqVUCqVRMmKdmu2ydsjuu0mCQx/Rkln2v9OREQrUOyLXBAeWK0j2Y57o+y+GwqFOjoOut89Hk/blmPZLZKOkSzGqVQKfr8f4XC4bcuYM8asl/FslLRGPi+9SnjSa9gdsvfoum6z4LeziEd1R3O5HNdia4N+lOVh4cYMnHYtbvIKe6ursyTAKL6NMkXWwuPxuMagdQM5MYlpmvD7/TWz7Pn9fmF5owxmy8vLiEajLcWmAXarBE28SDw2sw1ZuAUCAXE+6VhaGbycsV6RSEQ8gOS4LbnWl2EYKJ87V9mX14vokSO2bVJZAxIy9SaCcj2kcrmMcDgsksPU+j4JPFrVpO1TkgfCTeA4i23Xa5thGDZR65b0gSye1KZuTuyoiCvVFwwGg20nngBWyywAjiLqHbJWEpO4pf3vtctbrckYuTW26+6kaZotCUInmTANwxDjEFl1W8UtVo8W3egcBwIBZFYSGLWCaZpYWlqyJWXoZTybYRhYWFgQ4yzFPw66GHsr0HgqZxuVCYfDIjsk0zlyYq9WaxByLbb2oeeupmmYnp7G1q1be7o/Fm7MwOh0VcK5YtwKJNyAyuRRTjDhhsfjsVmlepGYhNwwA4FA3YkUPcApeQZgTzsPNCfcyBVSnjDJA3a7wg1YdSltBlohJyjWiyZItdwkC4uLsKamAAC+dBqRLVts2zVNU5wTeQXfaXFzFrH1er1iNd7NekCxMqqqCkuv7O5XLpdtCUSc55HqElEbGsUDyXWxYrGY68QtFAqJbcrnqxtQnR7LshAKhRCPxzuaPJLoBbrnJglU1+obRgaR9r9cLiObzdrGh26snru573Zideokvs2yLMzPz9u2QcdILl9A5d5uJ2MniW35HPYyxsy5P6/Xi9HR0TWTYp3dIfuP0z2/FY8IrsXWOiTWqM4reU01E3rTKXw1mDWLfIO0Y3Gjh6DX660asOrtzzTNrj1AqQC2HC9GFqd6kKWH2uP1erG8vCwG7GbbJws3p5tjo8FanlTI5QDos2bbQEkvgNVYL7LKUJucVkjLslC+cEEkHvHv21clvGkgpTa6tYdqL8mWW/k6tOMm2cja5ix3UE9kyDXb6tVLo3pGtMLdTTGwuLgoVsu7URS6F26SAKqsLMOGZVm2yThZfnrlVqNpGrLZbFXih0gkgmQy2bG4ree+2w7txLeR66eiKLbfhMNhJJNJaJqGxcVFAKtuhq1OQt3i2TKZTM+sRJqmiVhmoHLfj46OronJczPukNFolGOjegCFBADNW9u4FltrUEIdei67ibR+9O3hHwmYdQt18E5cJWk77VjcqFi11+t1deNw+w3RrRV9mlQ5axc1mrgUi0WxohwKhaAoCgzDEGKn2YmrM0FJs8JNttLRuXcKt2ZwWttIFJBYkd0k5Zp2xWIR5sQErBUR7d+3r2rAdFrcqH3U32iCJMeiBQIBcS1qiQrZTZISMdD1IkscnRfn5I5W6IDGxbadNdvq1ZnyeDwIBoOiH5DY7RRa+QcqfWVsbKzj4upyiYRuun0Nu8VtaWnJFpPbq7T/lIRDru8IQLi4dmMy1oz7bqvIcVzN9As3t0g5NblbHGGrYssZz0Ziu1ciSlVVm0U2EAhgZGRkKPszQbVQ5+bm2B1yQKiqKvoMJfaqB9dia55mxVo4HEa5XO65myTAwo1Zo8jxbWSNaQUqAUAZAWmSX+uB7JwcdCsAVRZcZLVqFG/ijBEaGRlBsVjE3NwcgMpkW671Vo92hZvTTVL+l9rQDLIrTTgcFg+Mhm6S+Txw9Wplv5EIQnv2VG2brim5hMrHQ7E5FOtCsSqy+6mbuyGde03ThBiU3f0aWducCVjqXR9nzbZG1ikS8ABEtstOmZ2dFW3oRgKNXlnbgOG2uMkZHCntf7cn4xSL6MwC14s6S82477aCvMjSKPGTpmlYXl6uEgmJRALxeFy4tcuWdCrc3Ep7nPFslKyoVyvqTpftYDA4kJp+zUIuqGR5la+H1+sV2TvXgqVwreMso1Orj9Za1OFabHZIrJVKJVvJJRkSa5FIRHi8ZLNZtrgxTC3aKbwt31CycKP3NU2r+ZChiTpQGeS6kbmPrDYU3xYMBkV8Wz2cGQtphTqZTCKbzYoi3FTrrd4EsZZwI0tkLdyEG2V9bDazJLk5EWRtc7pJBgIBeL1em6gzp6dhFouV9u/YgbDLpIyOx3mdVVXF8vKyOPZwOIxMJmMTxI3cJBVFEe2hh51hGLbaZM7rWCqVmi627azZ1kxdLHmVVFGUrriuyW5mbmURWqWXwm1Yk5NQUhAik8l01QWpVlIBitXstmtas+67rdCMm2StelJkXaN2UPp8OdFJK67D5F7Zr3g2oNJHZDEcCoUwMjIylC6F9dwhA4GAEGzD2Pb1iKIotoVct+cK12JrjCzW5DhsGZqHyGJtELBwYwZGJ66Sndb7kZMDyBP6WpNJVVWrhFunUOp2+VhCoVBTbpIEDdIk/NLptK2e2tzcHEZGRmpOFGXhRvF2QOuJSYDVZCckRBtZJQuFgmtmRRJtqqra3CRtVq2rV2EYRiV99J49rhN1t9TiJNrkCRLFGTlX192Q3SQpiQoJJjn9uLPGTa0ELG60m/QhEAiImEU6h+0+WCjZA13nVCrV8WoslSsAqouodwO5/uCwCDdFUapqnHVzVVtRFGSzWdv9SFngGll026EV991WaJSYpFgsIpvN2sZdv9+PVCqFUChku7dka5zP52spjrBUKomSIEDv49mAyrghn9NeW/baoVF2SEqewi52/ccZaiD3G67FVh/q1/RXS6yFw2HhETQM54uFG7MmcRMOzUJp94GKICChUy9BiaZpYoLSLeFGD8BCoSAEVKNVXbLQAfbJL71HNYWouLNhGJifn685+ZCFm/xAbkW4yd+lemIU91DP9VROIy6v2pNVRnaTJIFKRa+9V68KS5p/376q7VuWZUujHQgEREwbXTtyRaKBWLYG1XOTJPfLQCAgBCWlW6fjca56FgoFcY1CoVDdyXsnSR9CoZBYVW2UKbUehUJBWPyoT3VKL61tAGwxHsOArus217dYLNaxFVTedjabrYpRoaQcvXJPa9V9t1lqLcS5uUXSeOE26STXPfpesy6ptLAiewD0I77MGUMXi8V6nmW0FQzDQLFYRLFYdM0OSS6o169fZ9E2AOQyOnLt13q12GKxWMulAtYTa1WsybBwYwZGNyxuNHlvBXkgI9FH1oBaVgpylaQMit0QbmQVKZVKiEaj8Pl8DYWbm7WN2k+Ew2FEo1FRuJVSkKfT6aqJljwpURTFZlWshTNbo4wzQUmt7cgrgHIcBLkryG6Scq2vQqEAT6kEc2YGgWAQZiaD8OhoU22cn5+3Bf3Lq9qU4p+OoVYGSmqfM5tksVi0WQ/l8+p0Ca3n9thp0gcSbtTediZTFAdBVpB4PF7XrbNZGgnjTpAXY4ZhQuJM+0+CqhvbdXN5CgQCSKVSPa3xpaqqzX23WwLDNE3bQgXVXaMi2jL1MmI6rZvpdLqpRT3TNLG4uGiz+vXa6mVZFrLZrO34EolEV9xOuwFda7eJLS0msXvd4HEWgvd4PK5JezZ6LTZadKWYtXpiLRKJtF1Hsl+wcGPWHLqu26wmrSIXmaZBjKxEbmnsaX+ySOxUuJFFhIQVDRr1Jhok8oDqGCo5eyLFmo2OjmJxcVE8fCnuQxaHXq9XWIxkC029wd3pliXTTGZJXddtq+LyZIUG1XK5LJIU0HEahoFSqQTv5CQ0y0LA74e1a5erOJSTHcjuXVSDzymImnGTpHOvqqoIAJfdJAmnVUVe9YxGo3WvcadJH5xxbu1AFg5d1xGJRLrysJctxaFQqOuTh2GKb+tF2n8q6i4Xbgcq92k3ksY0s39ZFCUSia5dQ2e8ciO3yFrto3hMoLLY0Mw5ccazUTxptyyjtdq6tLRkW8hIpVI93WczNHKHjEQiiEajbFkbEuQyOoFAAOVyGbOzs1yLbYX1JtZkNtaVZIaWVuJxuhXfBkAUUJYn5c6JtdPljtrbCaqqCncGGlAbrbbKweByzJecDIREG7Bat2h5eVkIJVqJk2OsqHC2LNyazSjpnCQ3k1lSrmPmdNmQ3STJykMTMLIymBMTCKysLvr373fdByUmoX/T6TQAiCQuHo/H1t8aufGRpY2sspTq3+v1olwu20SJ8xzUEqlOupH0gUoTqKoKTdNE2YtmocBsRVGEK1Q3REGv3SR7UaqjXdzS/ncyIai1gl7LXbAXFAqFrtZsk6EFBl3XkcvlbONBM8dJccKyFbwZ6+Yg4tlI1NMxezweV0+IfkLukHLMMSG7Qw76vmJWkWOm6b7UNM32bN6ItdhM0xTP0XpiLRKJIBwOrymxJsPCjRkY7d4wncS3AdXCTZ4ouMW50XtyLbBOLW704C4Wi+IY2nWTlC2IzvNBEwOfzycGenpAkyuQz+cTk3wS0M0Kt1Ytbpqm2ayG8gSQxBG5OQYCAfFHFgeYJnD9eqVQdiiE0M6dNdtIcXQ0iSPLx8zMjO27sptkIBBwPXY5YQoN9uTuV8/als1mbSK11uSnm0kfQqGQOB5ZADfCNE3RXkVRRHs7dWuULcXyeesmw2Jxy2azXUv7X6s4LiUe6ddEmgQV0Y2abTLlchn5fB6lUgmjkttzs4XCZfHl9/vFIk0tBhXPRu6zckxSJpMZWAp2dodcu9ACG7noy30+EAi0VbNwrdKMWKPn2FoWazIs3Jg1hzyRaUW4yfFM9ICm+l6yu2Ct/XVTuJHrGE2QKY1yLXRdt8VgyZbGZmqv0USPXPFKpRIMw7BNVqiodqNgXFnIOifJJASdNeGIenXMyG1UUZQqN0l6QHlu3oRH1+Hx+2Hu2IFADYtrsVgUcXQUszbqiIVrNimJ/B1VVYU4C4fDIlkKUDn38sNSVVWbBa2eMO9m0gc5y14rwo3aQGUxKIlKp0JIroPTje25MQwWN+pzRLtp/0koyf0SGNwKerdrthG0GDM9PQ3TNMUiWiO3SBk5QYvX621Y98wtni0ajXZdjDqhJFFyuZWRkZGexiS6QWO/bEElaFGlm9eY6T66ruPGjRtiwZBEGy0EDNJ62y9IrFGJnXpijTJWr3WxJsPCjRkKmnWVlN0C2037TZMEYNW1MBAICDEl12iTsxNS4W3aRruQaw9NODweT92imUBtaxvQnHCj33m9XpHpjmq9hUIhmKYp/uptQz7/VC/Oid/vh2EYwnJG51quN+NWx4wmquVyWXxWZdW6elWINf++fa7nzDAMzM3N2ZKFjI6Owuv1VmVGo/0R9dwkqW/IbpKyEHW6czlFaq3r207NtnrICR6aLQsgJ56gxQSgO26NvXaTBAZvcetG2n9KYiOXlQAGu4JO8SFA92q2AatlOeSkPiRMm3X/LJVKVUK5kafAwsKCrWxEr+PZgMr4PD8/L/ZL7rP9FODsDrn2ofFhbm5OjNXBYBDhcBiJRAKapq1r0daqWFvPiw8s3JiB0c4KCMUYAe1P0MgNTxZ+JNxoHzRJoqQlQEVEkFtJJ8KN9kPxbSTcauF0NWtXuAGVYxgdHRUZ73RdR6lUEtuol8IfaM5N1e/32+JWaDJQT8hQYDzFpQWDQVHuQI4h8924AZ/XC3g8CLqUAaCVbXkVfuvWra59xePxNO0mSWKb+k04HIZpmjWTxcjFtusVVW+3Zls9KGlKqVQSGfvqPcTkxBNkoSTrR6diga4rgK5srxZO9+d+QlYc2SW2FTFQK3W31+sViUcGsVpMrrNEKpXqWBTTNmkhisaTcDiMrVu3Nl3kWtM0233jtHY7cRa4pvpuvZ7caZpmy2ZLpTX6lSiC3SHXPpZliQUdwzCEaKNn27DV/OsmhmHY3CDd8Pl8tpi1jQALN2YoaDbZh2zhaXcS4VbvKRgMigFRTtLhFCpkwehEuFGSCxICZMGp931arSVLj0ytmmq1CAaDGBsbs2W+o4xs3RJuhK7rQkTIyQ2cQkZ2kyQfdPqOsEblcvDl84DXC3PTJoQclinZHUnXdTHxlc+ts581Yw0i4SG7HYbDYVtK9kgkYrPSyhPeesW2O6nZVg8659Tueg802W3KMAxhKerGhE6eMPZygihbUvtpcSPXUupjtPrdLOVyGdls1rb4Miy1lpzuu53EYtUSp5ZlibT9zVoLDMPAwsKCrZxIrRqcbmn3g8EgMplMzwW+oihYXFy0lSDpdRwdwO6Q6wW6Z/L5vLgPKUFZNBrFyMgIMpnMgFvZfZoRa36/3xazttFg4cYMjHYmcZ0KN9M0YVkWLMuyPUBlESI/7JxChVzQOhVu5HoXCoUQCoXqDj713CQBe/HtZs+p3+8X4o0sbxRb1qlwc2ZVlDNgAe51zJziiISbHEPmvX4d3pXj8+3daztW0zSFaKNrHI/H6wYiU80bwm1iShYjiv8jS6DX662ZlIRWRgGI6+tGpzXb6uEsC1BLTFDNNiIYDNpcTDulH26SwGBcJSlDoJwYqNm0/5qmIZvNVk1Mmk3I0Wu6WbON3CKd2Whl6xrFGjdCLmsCVPprKpWyuakShmFgcXHR9syIxWJ1F1K6Rblctllhg8Fgw/i7TiFrjOx+SlAiKKoXygw3pVIJuVzOtqBD8xbqR8NUqL1TSKzJnipOSKxRzNpGhoUbs6aQU/O38/Ct5VLl9/uFKJMHDjfhBqwOoq22gVboqdg1ubXVmrQYhmGLMXGKADmjZKvuN16v1+Y2CVRW2cvlck2x0Y7FrVgs2tLlu227XC5D0zSYpolgMOgaQxa8eXO1hIHkJimLNqBybWKxmIjXkydLssWNrgUdi9v5I0ugM5skiTmgIvjkpDXNFtvuVdIHoNJXAoGAsOzKcZu12kDHBVQnwGkHsqBSe3q5MioXW++H2xAJCOo/zab9J6EsL8YAq2UqhmH12Om+266QrHWslBVTHoubPW6qMQisuju6nXNVVW0Cj8RnNwrJN6JYLNrOXzgc7riOXz0URUGxWHR1hwwGg6Kkx3p1p1tPlMtl5HI511pswOpidzQaXfN12agua7lcZrHWImv7yjPrhmZcJcnqAay6LbZKvexzwWAQ5XJZWKAoTT6wKuzkya+c5KRZ5PpaVHOrXhZHuXab28O3lfg2NyhleTAYFPGDtNLuXNFzJiap1WY6TxRfJbfRzfKjqqoQqCSOI5GIGNgBwGsY8E1PwwRgxWIIb9sGYFW0yRPoaDQqVuDrtVPuP824SZJFLhwO2yZmsrWt2WLb3ajZ1ohQKCTOi6IoVcfobIPcf7plbZP7bi9xc3/uJbTAATSX9p/iVCjTKeH3+5FMJgeWEt6NfD5vKx7eqtip5RZJhe9JpMlJgZoRbmRNAlYz6Lmd80HFswGVcycvNkUikZ7EILE75PpCVVXkcrmqOU0oFBJxz9PT0wBWk5mtRZoVaxSzxmLNHRZuzMBo9WHWaeFtoH4Sg0AgICYTJGKc9dE6FW7kJkmxZM6CzU4auUm2Gt9Wi1gsJixQJN7kWm+APVFLowGVMk7m83khOGtleqKVYkVRREa5cDhsm+RGFhZgUjbR3bvh8/lcRdvo6KitwLdzf/KkuVwu28SYG7KQDwaDCAQCNqusHJ/YbLHtbtZsq0coFBLWP6dwc2uDbClca26SnSYsaoVCoSDOFWWkrXU/0ARbjhejdiYSCeEWPCzoum47tlZFRy23SLdjlSepjcZzRVFsgiidTrve24OKZwMqyZfkeygWi3Xdna2eOyQtWrE75NpB13VbSQvCmUlWrge61qxtuq6LRcJaYi0QCIiYNRZrjVk7V59Z1zRjceu08DZlESTcLG6EM92sm3BrNqGKDAXcUhIIiplyg8oTAKjpTtmpxQ2AsP7RwEmClLISZjIZeL3els4/ZYMslUoIBAIIBoM1hYzTTdJZAsDj8cB/4wZo7/59+0QxW6doI8FomiY8Hk/NcyvXmaPslU7cLIHhcLhmbFuzxbblSXwoFOqZqCHXThLFtdpArp50Lmudj1aQ6w7WEzbdoJ8ZJcvlclXa/1qTERIbzoL15Co4yMQjtXC67zZ73She01l7LhqN1lyYoPNGtdtqoet6VdZO5z1jWRbm5+cHEs9GWVnlRbZEItFVK7qiKCI7pJNgMIhYLIZwODxUiwBMbWq5Efv9fiQSCVv/ljNJrhVrG4k1OSmZExJrkUhkTQnRYYDPFjMwhtHiRshlB+T9OS1urUCp92m75CJZa3IkD+q1JvfdEG6UgRFYnWhRxkRFUTA/P4+RkZGWhRutChuGUXNwJldKp5skiUagIizMK1cqP/D5ENyzBwsLC7aJH4k2ue4epbSXoXOvqqrof/WsbfRdWvkMBAJiVV0uAaAoSlPFtp0123oZYE7xkxSPp2maKHvhbIMsRteStQ3oX0ZJTdNE9lWgIiBqJQtyW0UPh8NIJpNDO0kpFost12wj67xs5Qaq3SKdUBIhYHWBodb25cyM5Domo6qq+AP6G89G7ZOvdSqV6kp2WHI5lTPPEuTBEI1G2R1yDVGrViM9M9ws8PK9Ncy19mh+QwuxblD2WDkunGkdPnPMmsBZCLvTUgBAtXCjlV9d14UFiKjlKtkKcgFqmlTXWm2uVyNMRtQ3cxEpzSILN0pqISct0TQNc3NzrhbIWni9XjGZMU2zrrVNdpOkWl+zs7PiO5FiEeWVzIfm1q3ISZmnZNEGwHbd6vWTcrksJjz14tuo31GcgWyJlQsFN1Nsuxc12xpBwg2o9D+/32+zGCUSCWFdBVYnhJ3ST+HWj4ySzhT0lP3R2Q5aRW9FxAwDbjXbGi2sUdFxefGo2dpzzS7CyQlg/H5/VZKPQqFgs3T3s7g11e+Tx/R0Ot1xfyeXa3aHXD/QAkc+n68ar6juo9v9QmJI/u4wwWJtMPBZZIaCRm6HsgWskwmQnGXMbaAMBAJiNVjTNPj9flF0mX5HtCrcyMebsh3WcyFrpv6VXJagkwHRMAwxCfD5fDAMA9Fo1FbrjSauiUQC4XC44QRZtiwGg8GakwxyHbUsC8FgEJFIxOYiGgwGgfPngZV4w9KmTfBQeYAV0eZWfoDOiVs7yU2S3FTd2iZbAmk7oVDIFsNGK/rOWmy1Vvp7VbOtHs6yAACq2tCoTmCryAlp6l37blHP/bkbkFuunII+nU7bvkOJR5wLQyRihp1sNmsrA1EvWUott8hYLNa0C2gz8W1yAhiv12tLp+/mnhgKhYRbd69xxtdSgppOCsyzO+T6w60WG9B8rUantW0Y3Ks1TRNukPKijQyFPLBY6w18RpmB0coDqNP4NgC2wtm1JnjBYFAMSLquC4FFdGJxo5Vh0zQRj8frxmA14ybZrcQkTosbPWD8fr+wvJVKJRiGgeXl5Ybn31nuoJ67FH03EAjA5/MhEonY6orF43Holy6t1lPbvh1+rE7knG2RBb6bcJPjvSgJihtubpJy/4lEIvD5fE3VqAN6W7OtHn6/X1zTUqlkmzBTG7ptHeuntQ3orcWNrKSy1UdO+0/nVL62FIdC9/iwQ+nkgVWLmRu13CKpllor47Ls1uj2O0roQqTTaTHGudVnc16XXmIYhq38CI1F7Swo0sS+UChUTYLJ+t1KrCEzPLjVYgNWy2E0WmTSNM1mbevHQl+9tjQr1ujZyPQOFm7MmqAb8W3NrMzTA5KEm/we0L5wkwdhWTC4PZCdiR3qCR+iU+FGrpZer7fqPI2NjWFqagqAfQWxltsGTbio7hxl/XNOqmRXRHL98Xg8Qlz4/X4ETRPlqamKK0YsBt9KlrtaE6VmXCWbLQNgWRZUVRWJRpyJDwDYVlLD4XDNFfde1mxrBCVUyefzIiEOtYEEMQCRfKUTKIsi0D23y0b0MjmJHKsmW31UVUU2m61KYtTspGxYIMsVUavt9dwiW40lk8u6uMW3aZpWVUeO+qWqqra6k+SeqGlaX0SbruuYn58X7W/XNVPXdRQKBVssL+Hz+USx7GGwsDCt4ZaUCGg9xtW5gNnvvkBzlnK5XFeskYV+rYx56wEWbsxQ0IyrJLCSYbCD7IlErUGQ3CK7LdxkNznZiuc2gW9UAoCQB9NOrJCGYdhWvuXzBKy6dZDrkt/vt7lWyaiqKia6NKhTW51tpBTBlmWJ7IrOjI3a5csorRTwtnbuFO6RbueNRCC5ono8nqrr3IwbH1kCVVUV4o/6BABRwsEwDFvq9FqWin7UbKtHKBQSsTgUL0ltoHIHALrihkWZOGm//Zhs9Co5CSURAFbrhgGVuCunm2AoFEIymVxzlhG5ZhsJehmysjvd91pxi3RSbxGO3FJlN3FaIBpkPBu1WxaN5JHQyoSV3SHXL41qsbWyWCc/R0nI9wMWa2sDFm7MwGj24SSnbpfjzVqlmZV5EjBUJNuyrK4IN9mViqwytbIeynFUbhYLmkxNTExgdnYWy8vLIjECJUCh2DFyWfT5fOKPLCuJREI8EMgCs23bNoyNjVXtU9d1sRpPwjmfzwvXRLckHalUSlw3p3AzDAOqqooYMrLOLSwsAFi1/sydOgV9RfB69+yp65JE18uyLHF9nedXnnA34yZJ+5LFAZ0z2WWsVm2dftVsq4ff7xcCU1VVWxvWupskAJv1pVvntlwuVyWcoUm3M/FIMBjE6OhoV/bbT5w12+QMp7UKhrfjFumkVnwbZWikey0QCCCdTg88no3aLAtKSuDUzP4p0RS7Q65PdF1HLperWsxx1mJrBae1rZdCnkQihUO4EQqFRMwai7XBw8KNGQrqWdxkl4NOXMyajYWR47ycLn4UtyUX524ETYIImtC6HYuiKKKdpmnitddew+TkJG7cuIHJyUlMTk7i5s2bNVfDugG5Im7ZsgVbt27Frl27kMlksH37dqTTaSQSCeHKRElLMpmMLSU31aOh9OnO9sop6mnSIrudRSIRZJeWoF26VPlBKISxY8fqPgSpn5DFDai+zvJKdy3hRolVFEVBJpMRNeHI2hsKhaBpmi0uqJYVrV812+pRKBSEC6xs7TVN05b+vVP3Tdntsl78YLeh89utSbyqqqLfWpYlhK9z/KCi0s5aTGuFpaUl15ptJFrle5Ysxd1IsS/Ht8l9LpvN2vrjyMhIVb1GoDKRrZW5tReUy2VbHblmRSO7Q65vqBabnIgLcK/F1gpy9mm/39+TshYs1tY2LNyYgdHsg7cb8W1A89nnnELNCU2Cm7W4laT09X6/X4gKeYW1WCzi/PnzeOWVV3Dx4kVcvXoVMzMzNbfp9XiQioWQSYQrf/EIkrEQIkE/QkEfQgE/QgEfgoHKcRqmBdO0YJgmDMNESdWRL6nIFVXkVv5dypcxvVhAWa3EcczPz+O1116z7TedTuPAgQPYu3cvdu7ciU2bNkFRFMzNzdnOh9OX30240cOJgpplN0ld11G6fBlWuVwRAQcOINLAXYQmdzTZBuyTedlNMhAIuFrISExqmiYslJSkBlgtASBbY2rFH/SzZlstqA3BYBC6rosskn6/3zbhaJS+vRnkRYdubK8Zmkk41AqU+ILiG3Vdt02cms0GN+wUi8WqRRa3+nPdLhhumqar9wQl6KB9ZjIZUXjbGc/Wz8WPYrFoi7cLh8NVJQmclMtlFAqFKpc5oDIZjsViooYns/ZopxZbK/TK2qaqqnCDdBNrtJBCbpBreXxb77BwY4aebmSUBLqXNrxV4Sa7ypEVolQqYWJiAmfOnMGrr76Kixcvum5vLBXFrs1JbBtLYNtoHNvHEtg2msBoMgKfr/sDq2VZyBYVTC8WcHOhgJvzeUxML+Hy1BJuzOextLSEF198ES+++CKAynncs2cP9u3bh507d+L48eNIpVIIh8PCWmlZlk24kaVHURSb6yUdvxBgV64Iy0308OGGbad91LK4ySUWalnu5JpncmIV2lYkErGtiNaKPxhEzbZ6baAkED6fT9TM67ZbYzOZULtNNzNKknVHdomUxTbVblvrq8/Omm3JZLJnbpFO3BbhVFW1jZGpVEokfyGohls/3Qnz+bytDdFotGY2WHaHXP/Uq8VGizmdiiwqGQRU+nwn4yjFfLNYW3+wcGOGglpuh7TyDUBYQNqlUQ03eZ9er9c2aZeRawm5ZUt0Qok8Jicnce3aNZw+fRpXr16t2vbmdBT7t4/gwLYM9q/8JWPt1wVqB4/Hg1QsjFQsjFt22ON2ioqGKzeXcWlqEWeuzuPVy7NYyJXw6KOP4tFHH8Xu3bsRDAZx/Phx3H///bjnnnuExYrqq3k8HhH4TNaMYDAoBFM+nxeiynPtmqgDEzpwoGHb5QQ2lFBEvjayUKm14k1WKFVVxQSS2kMrqc5Jr9t2BlGzrV4bIpEIAoGAcAGl5CvUvk4nlLLbpdfr7VvWzG5llLQsC/Pz81hYWEC5XIbP50N6JYNpMBhEMpkc6gLarUAxsUDlWrm5Rfaq/pxTuDkLm0ejUSiKUhWLmk6n+zqpzGazNvf2eDzumnyI3CGdRdcBdodcL3Rai60VnKVlWhWC9OyihFi1xJrsBsl9c+3Bwo0ZGM0MSnJB5U4nTs3GwmiaJmKuKJugbC1xJiipNWEsFot47rnn8IMf/ADnzp2zuQICwLbROI7u2YRjezfj6J5N2JTuvi97N4mGAjiyewxHdo/hXfcdhGVZuLmQxy/8q/+GbEHBSCKChVzJZpHbt28fbr/9dtxzzz3YtGkT/H5/lZukz+dDqVQSGe7i8TisXA7hYrGSHXLzZvhqlB4gTNOsur5ubpJU6LtWYW5yj5OtU9RPo9EoSqWSEENy1kzndgZRs61RGyjexrIs22fdmKA3UzC+F3Qjo6RlWbhx4wbm5+fFok0qlUIwGOwoVmUYIVFkGAaKxSLi8bjNFbHXbqCy+2AgEKjK0kj3H5FIJPqahdUtEUoymawqfcLukBuDTmuxtbovebGv2RhhEmuapmF6etrVc4fF2vqChRszFNSyuHXLTVK2ntUbbCkGw+/3i6QUmqY1LdxKpRJeeOEFPPXUU3jxxRdt7Y+GArjjwBbcfWgcdxzYgtHkcAu1Rng8HoyPJrAlE8eWTByf+f+8G5emlvDCmRt44ewNXJhcxKVLl3Dp0iV8+ctfxqFDh/ATP/ETOHjwoKgdR6vu+XwepVJJrGrHFhaAlXPu37evYVvk8+wm3OQVfFoAcKsrB6y6SVIsGAARoN1Mse1B1myr1wZd18V5WF5eFhODtZpNEujMVZJqzk1PTwshS3FUmUxGxDOuF8h1lgS87NoVCoWQSqV66s5rmqZtYirXupJj34DKtUyn031LcAOsZrWUY/zS6bSIcTRNE8ViEcWV8iQyFAsYjUbZHXId0K1abM1iWVZTzxb5+3LMGi1cyuMhibVIJNK30iwbGVocnpmZwebNm3u6LxZuzFDTzcQkNEmpN4DJEwsaBFVVtU1G5ckcTUaef/55PPHEE3jxxRdtbd42Gse9R7bj7lvGcWT3GPw9iEsbFjwej3Dx/Pm3HsV8toQXztzAk69cxemJWZw9exZnz56F1+vFgQMHcP/99+OBBx4QdbFIyKXTaRg//CHIlhJswU2yVikAeTJWL76NXAlTqRTy+byIcYrH41XFtt3646BrttVrAx03xT1QwfBOV40Nw7AXTe+jUG03bpUmZnIck8fjwZYtW7Bp06Z1OcmZnZ3F7OwsDMMQK/q9dIt0Ik+CyUIAVER/IBAQ59zv92NkZKSvMaGmaYpah4A9EUo9d0jK+sfukOsDTdNs2U2JXrtLl0olWz1Ft2eUm1hzwmKtv9BiaLlcFsKtlxm/CRZuzMBoZjVbTh3dj8QkcmAw/ca56ub1emFZFq5du4Yvf/nLePrpp22rZVtH4njwtp148NhO7Nnafze5YWE0GcE77t2Pd9y7H/PZIp565RqeOHUVFyYXce7cOZw7dw5f/OIXceedd+LEiRPYs2cPMpkMIsEglq9eBQB4olEExscb7qteKQBnNkl63ynAZTdJcpmjfufz+RoW2x6Gmm312uDz+RAIBIRLqmmaa9raBrRucZMzJ2qaJu7bUCiELVu2iCLb6wldr2SJnZ6eFnGmiUSi72n1aWxVVRWmaSIYDCKfz4s4VmAw8WymaWJ+ft4WI0v9YH5+nt0hNwD1arElEomeWn7rWdtoIZEW42qJtXA4jHK5jK1bt3J/7DGGYYhSCvIifT9h4cYMBW6ukrVSR7dDs7Ew8sObBmu5yHQ+n8e3vvUtPProo7h+/br43Ugigofu2IU33bYLe8fTPHg6GE1G8fAbD+HhNx7C5FwO3z95Gd8/OYGlfB5PPvkknnzySdx666342Z/9Wdw5NgasXHff3r1AE+dSFtckzOkaOJOSuD38ZGtbMBhEqVQSq6uxWKypYtuyRW5QNdsa1Y0LBoNiIqppWlcmJMMi3Bq5QOdyOWE1oRhASuGdSqWQTqf70OL+QRPCQqFgq0OWTqcxPj7e9yyniqII4ZxMJrG0tGSzXPc7ng2oPBfm5+dtq+ThcBjLy8tViR3IHTIWi/X93DG9oVe12FqhWCzaPDkCgYAQao3EGv1R0iyed/QG0zSFWHNbyAEqz9ZAIIAtW7b0vD08+jADo9Eg0634NqB14eb1ehEOh8VE7+zZs/jOd76Dxx9/fNUq5/PiDUe24yfv2oPbD2yBj90SmmL7WAIfevtx/OLbjuHFs1P4zouXcPL8FF577TW89tpreINh4C2bNuHAgQOI7N/fcHtyuQGyktFroLroNiUekPsffUdVVUSjUeTzeRF87vf7RVr9WsW2NU2zWeQGWbOtURvkZBSdWjZ0Xbe5F/d7QtvovnZL4U11mGKxmHAXHcT16iWlUgnZbBaGYYikBz6fD6lUCjt27Oj7BI8WRWhhIZfLIRgMChfDfsezAZX7ZWFhAYZhQNd1qKqKcDhcZXXx+/2IxWKIRCLserZOqFWLjcb3TmuxNQstrlDqfr/fXzfBSDgcFm6QLNJ6i2VZNrHmZlwIBAKIRCKIRCLw+XwoFAp9KRfDwo0ZWroV3wZUu0q63YSGYYjvkVB89tln8dhjj+Hy5cvie3u2pvDT9+zHm27biUS0drp+wzDxwtkpAMA9h8Z7UndtLeP3eXHfrdtx363bMbtUxLdfuIhvv3AJofOX8OPr1/HyK6/Ak0zi3bEYdu7cWXM7TmubnOpcdpOs9bCjSSVlndR1XVh4o9FoVUFU5+Rt2Gq21WuDYRii1EU3JqHyJFcuVN0v5GvtvLayeHH+hlJtU32w9TIJ0nUdy8vLYlWYEmpQHNbY2NhAjpXqspFVOpVKIR6PDySejdozPz8vJmZUs0++J8gdst+Ckukd/ajF1kpb5ubmxP0aDAar5jler9cWs7ZexqlhheYCFLfmNk+k+nqRSGRglncWbsxQ4HaD9NviRvsrFAp49NFH8Z3vfAeLi4sAKiLj/lu342fecABHdjc3+dEME3/4fz8NAPj87/8cC7c6bEpH8U9+6jZ88MQeXPzvKl65NINziokXvvc9fPN738N9992Hf/SP/hEOuxTidtagkifzTjc+uU/RNXS6SVKBarkkALC68u5k2Gq21WoD1VoLBALCAkOlL9rFWW+r37iV+CCR4Iw/iEajVePAyMjIurCi1LIgaJqGdDoNn8+HaDRaMzFPr5mdncXi4iI0TUMsFkMqlRIFrft9/kulEm7cuCHKIvj9flGzz+v1imLZ7A65fqAMsrIrOdCfEhjOdpALZLFYtJXDoIUv8vYhbwAWa72lmaQvPp9PiLVhyBrLIxMzMBoNSDTx8nq9HT9E5Qlerf3euHEDX/va1/DUU0+JyfpIIoJ33LsPb797HzKJ9VPPaVgJzi3iyK4xHNk1igujI/DeyOO512/g2WefxbPPPoujR4/i/e9/P06cOCGuo1spAHpNwoLcTJyJZgC7myQJv2AwiHA4bCvC65bIYRhqtjnrstVqAwlUStBCxdDbfRDJNbe6kZ2yVZwlPmolGAiFQkgmk2KyBFT6wyAsPb3AzbLo8/lsIs3r9TZMMd4rstksrl27Ju69TCaDdDpdVRut12iahvn5eczOzorJGWULpMUOdodcf/SzFpsbsliTrTjFYtEm2pLJJIu1PiKLNbdC5bSIQ9dkmFj7Ty1mXeC0uFHWO6BzN0lKRkAWFCdXrlzBF77wBTz55JNin7u3pPDeBw/hTbftRMDf3wnphubm/MoLDw6cOIjfe0sM12ez+PKTZ/HYy1dw+vRpnD59Gnv27MEHP/hBPPDAAzWFm67rYkCmotvOfkYPVaqB4xQ2copmt0D1YajZpmmaLZFKrTaQoAkGg6KfK4rSdkKIQSYlAVbdJC3LQrFYhKqqVQkGaDJEkzcinU4P5Fp1E7fU5R6PB/F4HLFYDLOzs+L9QWQ4BSqibWJiQizCJRIJ7Nixo28TIbq/C4UClpeXbQsxoVAImzZtQjweH7qJGdM5FFPptLyHw2EkEomeWk7Iu6FWfBR9Tpa17du3r4tFpGGHSuGQtd3JWokj5J7CDIx6N0Wv4tvkycvly5fx+c9/Hk8//bR47/b9W/DeBw/hzgNbhvamXbdoOjC3kso+HgGSFXe/HZuS+I2fuwe/+JNH8bWnz+GRFy5hYmICf/Inf4IdO3bg7W9/O+6+++6qfiL3ITdh4fF4xEOVJr+GYQh3CDmpiZu1YlhqtpGAqdcGudYaTVIpsUg78W7kegTYM7D2E0q8USwWEQqFhAXHmWBAVVVb/F+/6pb1CmeGTEIuDixb4EKhUN/jDy3LwvLyMq5fvy4WVqLRKHbu3NkXkUSxfYVCAYZhiNfAanHvrVu3DoXbE9NdBlWLjTIPlsvlmsksyIqjqqqIrWW33N5CtdbkWnkycu07ytA57HBvYYaSXsa3XblyBY888ghOnjwp3r//1u34wEO3Yv+27tVx8vu8+Bc/d494zTRgegEwVx52W0erPh5NRvHL77wDH3joVnz9mfP42g/P4fr16/jMZz6Dr3/963jPe96D+++/X3xfLqZLwsL5MCXhJT9oQ6EQLMsSgigSiVQ97Ie9ZpsTp3WMSm2Qf3+rwktRFHF+wuFw34+9XC5jdnZWWFDIQuqMV9F1HQsLCzaraL9d9LpJsVisitORLYtA89lFewUVs75586ZI+BSLxTA+Pt5zAUnHLqd3z+fzKJVKIk5lbGxs3WURZQZTi60Zsebz+YRljbwdZmZm4PF4hIWc6S60qFcul11rrXk8HuFFM4jnV6ewcGOGAueAJ99s3RJuk5OT+MpXvoJXX30VQKU82IPHduIDD92K3Vu6/yD3+7x42117u77ddcvU/Orr8WrhRsQjQfz8W4/iZx+4Bd987gK+8vRZzMzM4NOf/jS+9rWv4Wd+5mdw1113CVeHUChUc2AmN0lVVUUsJcVLARDFip0MQ822VtrgFG66rovJfblcbnlSMyg3SXk1XR4jYrEYNm/ebHOFNk3TFvxPsW5rEU3TsLy8bDtmmvTJmfAouyiNp5S5sZ/tXFxcxOLioriHwuEwRkdHxWSp25Ar/NzcnO38WJYlsgemUilhceGJ8vrCMAzk8/kqC7TP5xMLGt20orQj1uT9yzVBY7FY32OD1yvN1lqjJCNrTazJsHBjBkatwVSuy0XxRp0wOTmJv/u7v8MLL7wAy7Lg9Xjw5tt34QM/cQQ7Nq3Nidy6w7IqFjcA8HuBsXTDn0TDAbz/J47gXfcdwCMvXMSXnzyL2dlZ/PVf/zW+8Y1v4N3vfjfuvfdemyiRH7KqqsI0TfEvZfMyTVP0TTc3lmGp2SbH69Rrg6ZpwoIdDAaFOPV4PDY30WahuCFgNV11r6HaX5RcBKg8qCmpxMjIiG0CZFkWFhYWbOPIWkz7T26RJLKJcDiMVCpVNelzZhftp0gplUpYWlpCsVhEuVwWCVGor3UjyZSMaZooFArimJ2iVtd1RKNRcY7S6fRASlYwvaGftdhkUeCMpSVIrLl5aBDktkft5EWEzmin1tp6gIUbM3RomiZuwE6sbfPz8/jbv/1bfO973xOWiTce24FffNsxFmzDxmIOKK9MvDZngBZcSyOhAN734GG8894D+NbzF/ClJ87i/2XvvMPkqso//p3eZ2dbNr2QZNMDCRBa6EhoiihYaAr+VEAQBKRYEFGkKSAoShFsNIOA0gk1lBBKgkBoCUlI3d1smV7ulPv7Y/KenDttp+7O7r6f58mT2Zlbzr333HvP97ytu7sbf/3rX/HUU0/h5JNPxsEHH5z10CbBIrtVUQITGmhmvljrsWab0Wgs2IZc1jGyfsRiMVGAuNjjkDOj1TomgKwmwWAwK/GIy+XSlH6Q8Xq9YiBvMBiGZNr/cDgMv9+vSU9tNBrR0NCQUywPVoZTKiIcDAahKAqCwaBw32xsbBRtyrQ8lEsud0iCaizJfVSn06GxsZHrsQ0T8tVikxPzVONeT6VSGne7fGKN3O2KsSZnWtuG2jOpHhgqtdZqyfA7ImZIkmkJIcpxrQmHw3j44YfxyCOPiG3t2T4apxwxr6oxbEwV6ZDcJHPEtxWD1WzECYtn4qi9p+LJlevwyKufoLOzEzfddBMeeughnHLKKWhvbxfLU2IPSm5hNBqFYAPSoizzxVqPNdtylTiQkZOIyG6NFotFiNdYLFb0C24g3CTz1V0iMe1wONDb2yvaL4tyOc6F0v4PpZnWfG6RLpdL1BfMxWBkOKV4NpoACAQCsFqtcDqdWTGXlbRHzg6ZK2ZFr9ejubkZRqMRvb294p7Q6/VobGzkrJHDgEK12Ci1f6VCKJlMatwgc1GqWCMooyGwq+A3UxxDsdZaLWHhxgwa+QYg5Qq3ZDKJZ599Fvfdd5+wSMyc2IzTj5yPOZNbK2prOUSVBM68/jEAwN2XfBFWM99ueamCcCNsFhO+etAsHL3PNDyxYi0eefUTbN68Gddeey0mTJiAL37xi5g1a5bIHCm/BGiAbzQas9yq6qFmW642yO6DmdCAGkBWrJ88mI1Go0UNJCiNNZBdK6xaxGIx+P3+rGLpmYMzum6y2KbkHURjY+OQeYnnc4u02WzC5TAfg5HhNB6Po7e3F8lkUtQStNvtYlDrcrk0luFyhBtlhAyHw1npu/V6Pex2OxwOB7Zs2QKDwYCenh7hHktibqhcfyY/0WgUfr+/JrXYSKyRG2QujEajJmatHOTn0kAV/B7qkFiLRCI5xRpl6Szknjoc4ZEkUxfIFjd5trQYK4Cqqnjrrbdwzz33YMuWLQCAsc1OnH7kfOw7e9ygxrWEooWtIQzSLpK9O19qDQ7AXh2XJrvFhJMOmY1j9p2G/772Kf7z+qfYvHkzbrvtNkyYMAFf/vKX0draKl76er1efHa73Vn9ph5qtpXahkLWMZPJBIPBIJKzkItof9uj/Vfb2pZIJOD3+zVlGABtmnuZTDfJWCymybLZ0NAwZNzjSnWLlBmMDKcUz6aqqnBnpf5I8YSANrNrKQJKURSEQqGcrlBysWzqr6lUCj09PULcGQwGYYFjhi61qsVWilirhgVHURTNxApb2/JTTK01uTD2UItbrgb8VGMGjVw3HKWPBoqLb9u0aRPuuusurF69GgDgspvxzcPmYMneUzkF/1Chs3rWtlw4rGZ88/C5OHa/6Xj01U/w+Iq12Lx5M2699VaMGzcORx55JCZNmgSPxwMgbRnIHPDXS822UtogJxHJV2vNYrGIbGyKovQrEmrhJpmvLpnJZILb7c7ZJsokCKSFWyKRQF9fn0bUDoXBkaIo8Pv9JbtFysiuYzTQrBVyPBsRj8dFEhByTdTpdJpneTHxbf25Q1qtVjgcjqz+oCgK4vG42JfRaERzc/OQco9ltNSiFlt/KeKBXbFRVqu1qpZa2drmcrlGpNgoxHCstVZLWLgxdUWxbpLBYBAPPfQQli9fjlQqBaNBjy/t344TD54Jh3XkmMyHBUWWAagUt92C04+cjy/t346HX/kYT638DFu3bsU999yDyZMn46STTsK0adOyMjRSMWGxnUGo2VZOG2Q3UNk6IUPCDUjPbhcSbolEQtyfJpOp4oFNviQDJEoLZQDMdJvp6enR1JWr97T/qVQKfr8/y821GLdIGbJMAenBTS2PW45nI8iSRjWpGhsbhZWr2Gd5se6Quc5JNBrVCHaz2TwkE9Ewaapdiy2RSIgJr4EWa0QsFhP3DO2L2SWkI5FIzjjtoV5rrZawcGPqAnrx9ld4m+LY/vGPf4hZrH1mjcMZR+2OMc31lVpXpwPm7oytG+ETRPlJpYDOvvRnsxForn1qfY/TijOP3gPHHzAD/17+EZ55az02btyIG264AXPnzsXpp5+OWbNmieVli8Zg1Wwrpw3FWMcy49wKDfxlF8ZKz0EkEoHf789KMpBZlywfJNLIAkQDunpP+6+qqojDk8VnIetioW3JYr6WGU7leDZgl7tSJBIR5zqz/f0Jt1LdITOR3TWBdF9uamqq22vP5KdQLTaXy1WwH2RCYi2fIADS/Yus07V2p2Vr2y7kTJ3DvdZaLWHhxgwauR5ghV72H3/8Mf70pz9h/fr1AIAJo9z4v2MWYI9pbbVtaJlYTEZc/X+HDnYz6ptuHxDf6RoxunlAFW6z24bvHbcQJyyeiaUvf4jn3tmADz74AJdccgkWLFiAk08+GVOnTq2rmm3FtiEziUg+i4der4fZbIaiKEgkEkgmk3mtPbJ1qFzhlss1ECg9yQCJnmAwKF7ulPa/XgdGiqLA5/NpBpOV1JwKhUIDkuE0HA5rYiupsLHf7xffkVVMRo5vo/5HmQHlthPkzltM7GYoFNKI1nq/9kxuqBZfZrmPUu+LehNrhGzpo/2PNOhdNNJqrdUSFm5MXUBB7vTQNRqNYkDm8/nwt7/9DcuWLQMAOKwmnHz4XBy9aCoMHMc2tKliNslyafXYcc7xe+GrB83CQy9/hOdXbcDq1auxevVqzJkzB8ceeyymTp1aFzXbim1DZhKRQoMfi8UiBhexWCyni2I8HhexB2azueSXaz4XKIvFArfbXbKLUjKZFNYastDVa9r/fG6RlWTEo9T7RC0ynFKmSDnLpdlshsfjgdfr1cSvZU4mpFIpTfFzKrsRCoWy3Fz7c4fMJBAIaI7d4XAgHo+zaBtCVKMWG8VFRaPRgmKNXO0GI1HNSLW2ca212sJnixk0Mh9icuFts9mMZDKJZ555Bn//+9/F4OHwhZNx+pHz4XGOvJmrYQnFt+kAjG4a1Ka0NTrwgy/vhRMPmomlL3+EF1ZvxJo1a7BmzRrMmjULp5xyCubPnz+gL99y68aVkkTEYrGIAUY+4VZuUpJUKoVgMIhQKJRVQNvtdpc9A01uVcAua0u9pX0v5BbZ0NBQUVbSWmc4TSaT6Ovr01hGHQ4H3G63psZcPksX/U6CX362E8W4Q2bi8/k0QtLlcmWVHWDql/5qsTmdzoLifSiINUK2/uVKeDXc4FprAwcLN6ZukB/Emzdvxt13341169YBAKaM8eCsLy7EzIktg9U8ptoEI0BgpxWiuQEw18eDvK3JiXNP2BsnHjwLS1/+CC+u3oiPPvoIP/vZzzBr1ix8/etfx8KFC2su4Mq1qpSaRMRkMkGv1wuXllxlAfIV8c5HPtFSiWsgkSvtf70VWK62W6QMDYyA2mQ4VRQFfX19mni2hoYG2O12jWAmK2emZYRi7/r6+pBIJDSlHEpxh8zcptfr1UwgNDQ0DInMoUyafLXYbDZbQU+CeDwu3CBzZRwEdgmjwRZrBMXeEoORhXig4FprA8/g93CGgXa25r///S9eevFFpFQVDqsJpxwxF0ctmgrDEAtUjSeS+Nsz7wEAvrVkPkzG+nPjGlQ0bpKDa23LxegmJ847YW98/ZDZeOTVT7DsnfX46KOPcOWVV2LatGn46le/iv32269m7nk+n0+8CO12e9EvvVKtY5RqmV688Xhcs69CRbxzkWuAptPp4HA4Ki48G4/HNaKC3A3rBSqQnukSarfbq5KNlNwuiYaGhqoG8OeKZ2tsbITZbBaF0QmPx6OZFCD31XA4jN7eXo2rpF6vh8PhEGUDSkFVVfT29mpi5khIMvVPvrjWQrXYShFr9RgXJbfbYrHU3cRSpfRXa02v14trU0wpEKY0WLgxg4Z8M6uqihUrVuD+++8Xbi8HzZ+IM4/eHY2uoZk+N5lS8diKtQCAU78wD/VhT6oj6iC+rRhGNTrw/S8uxEmHzMJ/Xv0ET735GdatW4frrrsOY8aMwZe//GUcfvjhVX05yzXb9Hp9SWney3FrJOEGpIWaLNyK3V6+2kulprjPRzKZRG9vL1KpFFKpFCwWS9FxUbUmX8xONdwiZTJrtlXL/YqsZHIcntlsRmNjIwwGg6iTRzidTtEXMrNDqqoqBq1WqxVNTU0luUPKpFIp9Pb2ikE/lRwY7m5nw4F4PI5AIKDJRgvkr8VGYiAajRYUa+QGWQ/3fS6Gq7WtmFprJNZGamHsgYKFGzPodHd344EHHsD7778PABjT7MRZX1yIPaaNHuSWMTUjkQR2eNOfbRbAU/8vtyaXDWccvQe+etAsPPHGWjzxxjps374df/rTn3DvvffiuOOOwzHHHFNx5snMNO+lWFXkJCIWi6XowY0sOmOxmBhsFFPEm1w6M5NvVFIsNxOyupBoMRgMwno32AOEVCqF7u7umrhFymTWbKtWhtNC8Ww6nU6IJ7lOnsvlEslGMmONEokELBYLrFYrPB5P2ZYxEuq0fb1ej6amJna3qnPyJSLKFdc6XMSajFyP0Gq1Dun+WkytNbrXudbawMHCjRk0kskkli1bhv/+979QFAVGgx5fPWgmTjxoFsym+n9AMxXQ1Qckd1omalh0uxa4HRZ88/C5OOHAmXh+1QY8+uqn6PL6cd9992Hp0qU45JBDcNxxx2G33XYra/uV1I0rN2W/wWCAyWRCPB6HoihIpVLQ6/WIxWKaAXumlTwYDGal8jYajaL2UjVQVRV9fX1i4ECiTafTDepAgdwiFUXRDGqq5RYpk6sAezUGsYXi2Wi/Xq9XDKrpnHd2dmbFsxgMBtjtdiSTSeH+Vu6gNZFIaNwt6zUBDbOLYmuxKYoi3CBzudkB0IiBoSDWiOFgbaNaa5FIJG/Rcq61NriwcGMGhXXr1uHWW28VNdnmTG7FOcfvifGtxbuEMUOYIeImWQir2Yhj952Oo/aeitfWbMGjr36Cz7b1YdmyZVi2bBnmzJmDL37xi9h3332LHnzE4/GyrSrFWMcKYbFYhACJxWKiwDIhW04o8Yg88NLr9SKVdzWtYH6/X+M22tTUhB07dgDAoAzqyC0yEAhoBqjVtDBmklmzrRrxXaFQSFOLLZc4Ilc3RVFExtFcllWHwyGEfU9Pj+a3Usks9m0wGNDc3FwXSSeYbKj0Q1dXV1YtNnoekNvkcBRrMrKr9FDKnJhKpYSYVhSFa63VOfwkZAaUSCSCe++9F4899hhSqRScNjO+vWQ+Dl84BXr98PKJNhsNuOOiY8VnRoLKABj0wKjGwW1LhRgMehw0fyIOnDcBH2/uwRMr1uL1NVtEKYGWlhYcccQR+MIXvoBRo0bl3Q5ZVeilWWrdOEVRNC46pc6EWiwWUeg7FovBYrFoBJOcoEK2MFEqb5fLVfXZ11AopBGyjY2NGlE40LO9lNEyM/FKJS6B/UGuZ4TH46lIGOeKZ7NYLGhsbNScz3A4jO7ubhHT4vF4xICNsotmJs2hJFNAWnCVOsBTFEXjlmkymeq2Pt9IR47rTCQS4rlFtdhMJhNisRi6urpyijUqzE5ukEPdckPFxIl6t7bRRB/941prQwe+EsyA8fbbb+NPf/oTurq6AKSTj3znmD2GbU02vV6HtkZOV52FLwhEdiawaPEAw0TU6nQ6zJrYglkTW9DjD+PpNz/DM2+tFzGcDz74IBYsWIAlS5Zg0aJFWS/CcDgsBr1Go7HkVOflukkSlP2LiqfKL3Oz2Yy+vr6sRANWq1WT7r2aRKPRnGn/ZfedgRrs5csW6XA4oChKTTMcymKeBsTlkhk3RtuUiwMnEgn4fD5s27ZNCCjaL7lD5ssOKddrKzVZTzQaRV9fn6bP5So3wAwu+WqxARB9RI7zkhluYk1GtrbZ7fa6FDpca214UH89ixl2+Hw+3HnnnXj55ZcBAKM8dpz1pT2xZ/uYQW4ZMyhsl9wkh1h8W7E0u+045Yh5+Nohs7Hiw61Y9vZ6vLe+C6tWrcKqVavg8Xhw8MEH45BDDsHUqVORSqUqsqrIbpJ6vb6sDJcUaB6NRkXCEZpVTyQSmoGIyWSC2+2uWZprSvtPOJ1OIY7kAWGtLTGF3CIbGhpgMplqWvw5EomILJ2V1myLxWLo6+sTgzWyFJLIj8ViIp2/1+vVxDZSzbTMOMdMZFFdiptkZhkCq9WaZV1lBp9oNIpAICCEv6qqQqzT58wkFiMhgQWVwgDSx1tv1jautTa8YOHG1AxVVfHiiy/irrvuQiAQgF6nw3H7TcfJh8+BzcIzOSOWYRDfViwmowEHzZ+Ig+ZPxPaeAJ59ewOeX7UBXq8X//nPf/Cf//wH48ePx7777osFCxagpaWlpJpthGwdKzf1OgCNcOvt7RXFuJ1OJ4Bd4qGWFiZ538CucgJEZkHvWpHLLZJKMwxEDbFUKpVlcSz3umbGsxmNRjQ2NsJoNAp3VHJ38/v9SCaT4rqPHz++6P5YjnALhUKa47TZbBW7gzLVRa7FRgKNPtPzRr5eI0GsycgJmsqpVVgLuNba8IWFG1MTOjo6cNttt2H16tUAgMmjG3Dul/fG9PH1V2i5ViRTKXzemR6QTGprGHIFxGtCTAF6dg7SXHbAOTRr9JXDmGYXTv3CXOw3exw+3tSNjzb14O1PtmPLli146KGH8NBDD2HKlClYvHgx9t9/f4wfP77obVfqJklYLBbEYjH09vYiGo0KSxcN4imbY62g1PM00DCbzfB4PFnLELUYICWTSfh8viy3UIfDUZM4vnz4/X5NooNyapdRVkjZxdNisYh0/pkz8MFgEMlkEg6HAw6HA6NGjSr6HMvxbXq9vihXsUAgoLE0OxyOqpU5YCpHTioSj8cRi8WgKAoMBgMcDofGlY7EGtXxGu5ijUgmk+L5K09yDQZca21kwMKNqSrJZBKPPfYY/vnPfyIWi8Fk1OMbh87BlxfPgNEwMh7kRDyRwo/+uAwA8OAVX4HBPLKOPyedfQB5nA1TN8lCxBMp/Pj25wGk+0QqpWLFh1vw8v824b31ndiwYQM2bNiAf/zjH5gwYQL23Xdf7Lvvvpg6dWreATTFpAFpIVOuqwvNqodCITFLq6oqGhsbRSHmWpKZ9t9oNKKpqSlrcJGZybKa+89V3kB2ixwoYrGYGAyWWoCdyBXPRsfQ3d2dc3mz2Szq4zU3N5d0zROJhBCB/bnQkmUvM5lDvbmYjVQoIY7P5xNiLZVKiVIf9IwhIRCNRtHW1jZixJqM7EbtcDgG3NpGtdYoEUwmI1VQD2dYuDFVY/369bj11luxbt06AMDcKa34wfF7YWwLv4yZnYwgN8lisFtNOHzhFBy+cAr6AhGs/Ggb3vhoK95f34XNmzdj8+bNWLp0KVwuF/bYYw8sXLgQCxYsQHPzrnMnW0zKceHLTLyh1+uRTCaFO11LS0vlB1oEfr9fCFBK+59rkFELV8loNAq/36+ZpTYYDHC73VWrR1csmTXbXC5XyYNBOZ4tlUpBUZScwpOyQxqNRgQCAbEfj8dT8gRAsW6SuayAFEPHDC7JZBI9PT3wer2aGo7kIk0Df3KBpJhHv98/IgVBIpHQTLAMlLUtV621zKyQ5KrKtdaGHyzcmIqJxWJ44IEH8PDDDyOVSsFhNeGMo3bHEXtOYVM8swtVBTp705+NBqCFXaJkGl02HLVoKo5aNBXBiIK3P9mONz7ainfXdSAQCOCVV17BK6+8AgCYOHEi5s6dizlz5sBqtQq3ylJERiqVQjAYRCgUyiqY63Q6YTQaB8zKRO0AdqX9z+dqJxeLrnRAkkgkNHXiaLsOh0NYngYaSq8O7KqRVur6FKdGLlNOp1Mj/sjVzW63I5VKobu7W5O5shyxWoxwU1UVvb29QqBnJkhhBh7KEtnb26tJXgOkxYjdbheJK2SxxiDLzbeWzwuutcYQLNyYinj33Xdx2223Yfv27QCAA+aOx3ePXYBGF7+ImQx6/UBsp9tWWxPAs4B5cdrMOGSPSThkj0lIJFP4dHMPVq/rwOp1nVi3tRebNm3Cpk2b8OSTTwIAmpub0d7ejvnz52PatGmYPHly3pgoVVVFAe3MQRqlhY/FYiIxSa0haxfh8XgKutpRm/V6fdnty+cWabFY4Ha7By0NdiKRELX0yinA7vV64fP5xODOYrFokppYLBY4HA4R35JKpTSDdavVWra7Yn/xbRS/SMuRQC8ndo+pDHKvpmyecip7YFdtxsbGRk1/YXZByT+A2lnbKFswuUIWqrWmKApaW1ur3gam/mDhxpSFz+fD3XffjRdeeAEA0Oy24ftfXIh9Zo0b5JbVDwa9Dt84dLb4POIZAWUA+qOcPmE06DF7citmT27FKUfMgz8cw/vru/DR59348PNubNjuRU9PD1asWIEVK1YASA8kxo8fj6lTp2K33XbDxIkTMWHCBDidTgQCgawC0mRhosyCRqNR1HTLLAVQTRRF0aT9d7lc/VpfZOFWDlQfLrOswGC4RWYip8TPTP5QCEVRsH37dk1tLbKokTtkru15vV5NTGG52RwTiYTYr8lkyhmX2NPTI/oducJy6vGBQy64THW8Muut6fV6NDQ0oLm5WfQdJjeytU2ug1gpch3NUmqt9fb2VmX/TP3Dwo0pCVVV8dJLL+Guu+6C3++HTgccs880nHrEPNitnOJfxmQ04JuHzx3sZtQPcnxb28jJLipTjT7htltwwNwJOGDuBABAOBbHp5t78dHnO7BuWx8+29qHvmBUWOVefPFFsa7VasXo0aMxZswYtLa2YuzYscI6p9PpxAyy2WwWwigWi9VEuOVK+9+ftSeVSonlS3UFosLS5KYH7BKt1Rx4lUs4HBZtoyQQ/ZFIJNDb24uuri4xACfLKYk1u92eU+QGAgFN7b9Kil3LbpKZ1tJEIoGenh7RPoPBgKamJi7uOwDIYo1KhiiKIso/AOlrTwl4mpub+boUgaIo4t6hovTV2CbXWmOKgYUbUxKPP/447rjjDgDpFPc/+PJemDFhZFpPmBKIxABv2gUMTS7AVpvCzSMRu8WEPaa1YY9pbeK73kAEn23tw2fb+rCxw4vNOwLY3pMeqG/cuBEbN27M2o7FYhEZJBsaGuDxeNDU1ITW1laMHz8eHo8HHo+nKq5tqVQKPT09YoCSK+1/LsrJKKmqKgKBQFYsH7kR1sqaWAqpVErjLtpfzbZoNIpQKASv1ytcK4G04GtpaUFDQ0NB97ZIJJJV8L2S85Avvi0ej2uuM2UKrYdzPlxJpVKIxWJZ7nXxeByhUAjxeFwkGLFYLHA6nQOeNXWoUy1rG9daY8qBn55MSWzevBkAsHjeBPzoxH1GXIp/pky2S+nHR6i1bSBpctnQNNOGvWeOFd/FE0l09IaweYcfW3b40dkbQkdfEJ29IXT709aejo4OdHR0FNy2xWKBx+OBy+WC2+0WadzdbjecTidcLpeo+Sb/o8E6pf2nGf98af9zUWpGyUgkIhJ1EPXiFinj8/k0NdtyxfilUimEw2GEw2FRX0tO8tHQ0IDRo0f3m4o/Ho/D6/WKv91ud8ViXI5bIwEgZ7YE0i6UTU1NnDShBuQTa8CuzIfxeBxmsxl2ux0mk0nU8+uvvzBaYrGYxjJe6nOEa60xlcLCjSmLCa1uFm1M8chukmMGJr08o8VkNGDCKDcmjMquCRZPJNHlDaPbF0aPP4JefwTd/jB6/RH0+qPwhaLwBmNQEknEYjF0dnais7OzpP1brVaRsdBms8Fut8PpdKK1tVUIQLfbDbfbDY/HI1LEywOXTAGWj3p3i5ShATewK8ZIJpFIIBQKIRwOQ1VVUb4hkUjAYDDAarWitbW1qEQmlCBEdk+tNKlCMpkUA1CKb4tGo+jr6xP7MZvNFbliMtlQlsFoNJozcQWljFdVVSSd0el0wg23niYuhhLlWNsowysVMs9ErrXGWTuZ/mDhxjA1IhZP4NLb08lbrvv+YbCYRujtlkwBXTsTUFjNQOPIretXr33CZDRgXIsL4wrUXFRVFVElAW8wBl8oikBYSf+LxOAPKwiEYghG09+FonEEIwqCkfRnAGKQWQoGg0HjtulyuWC329HQ0IDx48dj/PjxaGlpgdPpFFkSc5U4qCe3SJnMmm1utxt6vV4kKAiFQhrxqSgKAoEAjEajcIcsNjMjWTrlJCLFuKf2R2Z8Wzgc1lj0rFYrGhsbeTBaBfoTa0BaBFAxdDnBCNVis9lsfC3KJBqNiv5OqffzkavWWiayWONJDaZY6ustxjDDCFUFNnR4xecRS7cXSOx0cWtrAkbwoGEo9wmdTgebxQSbxYQxzcVbaZKpFMJCyKX/D0QUhCIk/BQEwrGd/yvwhdLCMBJLiAQmvb29WL9+fd59WK1WNDc3w+PxoKWlBa2trWhtbcXo0aMxderUktLqDyRyhk+z2Qyr1SqEZ2bMC8XBkAA1mUwF691lIhc4pwQh1RjAy4NSRVE0QpNENguF8immfpfBYIDFYkEymYSiKDAYDMIiTanqM63XTGlQrCyRK3lQMdeK7nOutcaUCws3hmFqC5cBGNEY9Hq47Ba47KXF0ijxJPzhGHyhGPoCEfQGougLRNAXiKI3EEG3L4JuXxi+UDp19tatW7F169ac22ptbcWECROElW7ChAmYOHEi3O5st9GBgpJFAGlXKp1Oh87OzqzBnl6vRyKRgMViEbPyNputpNT9oVAoq8B5tQaNJNxCoRBSqZRoo8PhqFvBXO8UK9aoIHYikciqyzjYheSHG9FoVLg5kvgCSqu1ZrPZ6s7qzww9uAcxDFNbKL5Nr+PEJEzRmE0GtDTY0dJgB9CYd7lYPIEeXwQ7fGF09YXQ0RvE9t70/x29QYSicezYsQM7duzAqlWrNOs2Nzdj8uTJmDx5MqZMmYIpU6Zg3LhxAzIT7vV6xYDPbDZn7dNisQjXQ3ngTUlgikVRlKyMldVKJ55KpaAoCoLBIBKJBBwOBwCIhDVM8SSTSeEGWUiskWud2WxGJBKBz+fLqstI8aNs0akOmdY2p9OpqYdX6FrJtdYYphqwcGOYGqEDMMpjF59HJP4QEEwnXkBLA1AnMV2DBfeJ6mMxGTG2xYWxeeLz/KEYtuzMpLmlO4AtO/zY3BVAlzeEnp4e9PT04J133hHLW61W7Lbbbpg2bZr4N27cuKpZLVKpFLq7u0XtNTkzHdVqcjgcSCQSmgQfer0ejY2NJWUBzKyVR3XdqkUsFhPZLekYKKkM0z+yWJNdTGUyxRqQtv7s2LEjK9EF1UJkq051CYfDSCQSiMfjSKVS8Hq9eQtjy+n7GaYW8N3NMDXCYjbizouPG+xmDC5yNsnR7CbJfWLgcTssmO1oxezJrZrvw9E4Pu/yYWOHFxs7fNiw3YvPO32IRqP48MMP8eGHH4plHQ4HZsyYgZkzZ2LGjBmYMWNGyeKEXCNDoRB6e3vFwM/pdMJkMsHhcIjEEYFAQFOfrdR4NiBtJZD3Y7FYquoamkql0NXVJQSH2WxGY2MjZyvsBxJrhZJWGI1GIQBkaw1ZTzPXo2vLlp3qQ2VSKMbU4/FklSXhWmvMQMLCjWGY2sHCjalT7FYTZk1swayJu8pTJFMpbO0OYN3O4uXrtvZi/XYvQqEQVq1aJVwtdTodJk2ahNmzZ2PUqFFwOp1obs7u3xT/EgqFxGA7GAyKgV9m7TVK1y9bX0qNZyO8Xq+wyBiNxqpmdqQC6nLcXFtbG4u2PFQi1oB0SQi/35+VldVkMsHtdnMttioj11rz+/2in5vNZlHygmutMYMFCzeGYWpDPAF070x17rACbnafYuobg16PiaMaMHFUAw5bMBkAkEimsLHDi08296T/bepBR18IGzduxMaNGwEAf/3rXzF69GjMmTMH8+bNw9y5c+FwOLKyQyqKAkVRhDvkmDFjhAtmPB7XFCbX6XSimHmpBAIBTW24atZQSyaT6OnpQTweRyKRENtn0aaFanfJKeQzITdZq9Wa01qWTCYRCAQQDoez1uNabNUlV601SukPpO/H5uZmuFwurrXGDCos3BiGqQ1dfUBqZ9A2W9uYIYrRoMe0cU2YNq4Jx+47HQDgDUbx0efdWLNxBz78vBsbtnvR0dGBjo4OPP/88wCAlpYW4VY5c+ZMNDWlE/M0NjZCr9fD4/EIMRWJROD1eiuKZyOi0agmkYLH46lazFMikUBPTw+SySTi8bgoGM4xbWkSiYQmwUgu+hNrAPLWI+RabNWlP0toNBqF0WgUdSBzWdUZZqBh4cYwNSKRTOHJlesAAMfsMw1GwwhLycxlALIY8X1imOBxWrHfnPHYb854AOl4uY83deODjTvw/vourNvWh+7ubnR3d+O1114DAIwbN06IuHnz5sFut4tsdZnxbE1NTWVlBCSrHeF2u4sqzl0MiqJoYuZUVRXicyS76pFYky01mZhMJuFaV0hEq6oqBFtmHBXXYqsOcqmFfAlhzGYzLBYLVFUVRcy5tAVTL7BwY5gakUim8Jcn3wUAHLnXbiNrkK6qQOdO4WbUAy2eQW1OvTCi+8Qwxm41YWH7GCxsHwMgLeQ+/Lwb72/owvvru7B+e5+oM/fCCy9Ar9dj+vTpItnJ5MmTYTAYKipYnUqlNFkobTZbWW6WuYjFYprslCaTCSaTSVPXaiRRTbEGpAVbOBxGMBjUuNZyLbbqUEytNZPJJCyhRqMRfr9f3Idcf42pJ7gnMgxTfbxBILLT9aS1ETByPSFm5GC3mrDXjDHYa0ZayPnDMby/vgvvre/C/z7rxPaeID755BN88sknANIlCObNm4e99toLCxYswJgxY0ran6qqmvg4k8kEj8dTlWPJdOO0WCzweDzo6uoCkHbfGwn1wihhhVyIOZPMwX8xRCIRBAKBrFpslNp/JJzbWqCqKmKxWMFaa/kSwiSTSU3iHa5JyNQTLNwYhqk+nE2SYQRuuwUHzJ2AA+ZOAAB09oXwv8868e66Drz3WRcCkSjeeustvPXWWwCAMWPGYMGCBVi4cKFwqyyE3+8Xbl+ULKQaLnXhcBher1f8bbVa0djYiHg8LgbCw9naFo/HRcxaNcUakLZi+v1+rsVWRVRVhaIoQqyVW2stGAyK/m2321k8M3UFPxkYhqk+HN/GMHlpa3TgyL12w5F77YZkKoX127x4d10HVq/rxMeburF9+3Zs374dTz75JAwGA2bOnImFCxdi4cKF2G233TRuc+FwWGMdKDc+LpNgMAi/3y/+lt045UQOwy2+jcRaJBLRWMFkzGazGPyXeq4VRRFFy2UsFgtcLtewFsK1QhZrsqspUUqttWQyKbJ46nS6qrkbM0y1YOHGMDXCZNTj12ceIj6PGKIK0LtzwNfgAOzVSY4wHBixfYLJi0Gvx/TxTZg+vgknHTIb4Vgc76/vwup1HVi9thMdvUGsWbMGa9aswT/+8Q+43W7sscceWLhwIebMmaNxAWtoaKjKwN/v92sSpjidTk3xblm4DQehUWuxBnAttmrT3zUrt9ZaIBAQ95TT6WRrG1N3sHBjmBph0Osxb7dRg92MgaeT3STzMWL7BFM0dosJ+8wah31mjQMAbO8J7rTGdeC99V3w+/1Yvnw5li9fDgAYO3YsZs+ejYULF2LvvfeuaN+qqsLn82nqhrndbo3VgdzRgLQlY6i69MXjcWGlKSTWyA2y3AE812KrHv0lhdHpdLBYLOKaleounEgkxHXS6/Vc5oKpS4bmE5dhmPqF3SQZpmqMaXZiTPM0HL3PNCSSKXyyuQer13bg3XWdWLetF9u2bcO2bdvw3HPPwWQyYdasWdhjjz2wxx57YLfdditacFCCE9ki5PF4suLrEomEiB0aatY2RVFEzFo+sWaxWGC1WisSa0DhWmxOp1OkmWcK01+tNQAasVZJ9k25/iFn8mTqFRZuDMNUj1QK6NxZR8psBJrchZdnGKZojAY95kxuxZzJrTj1C/PgD8fw3mdpt8p313Wi2xfGe++9h/feew9///vf4XK5MH/+fOy+++6YP38+xo4dm1MsUCkBirvS6XTweDw5rUFDLb6NxFokEskZ/wRUT6wBaQEcCoUQDAa5FluZFFtrrVJrqAxZYAG2tjH1DQs3hqkRqqoiFE27czisppHxsu7xA/GdM9ltTQDPWGoYkX2CqRluuwWL503A4nkToKoqtnYH8L/POvG/zzrx/vodCAQCeO2110QR8JaWFsyfPx/z58/HvHnzMGrUKKRSKfT09AjXM0pwkk+UDYX4tv6SVeh0Ok3MWjUsK1yLrTLKqbVWTWRrm8vl4mczU7ewcGOYGhGLJ3HK1Y8CAB684iuwmkfA7cZlAAoyIvsEMyDodDqMb3VjfKsbx+47HclkCmu39uLddZ14f30XPt7cg+7ubrzwwgt44YUXAABtbW2YNm0apk+fjunTp2PUqFFobm7W1LTKpF7j24oVa9VwqcskVy02IJ2Jk2ux5afYWmt0zQr1y0ogqyyQdmXtr/wGwwwm9fPUZRhm6CPHt41uGrx2MMwIx2DQY+bEFsyc2IJvHDYHMSWBjzb14L31nXjvsy58tq0PnZ2d6OzsFBa51tZWzJkzB7Nnz8bs2bMxYcIEjcBJJBJCFJlMg2sxpiQp0WgUsVgM3d3dWctQsgpyg6y2xStfLTar1Qq3211XwrZeqFattWoil71gaxtT7/BThWGY6hCKAP50PSk0uwFLfbpRMcxIxGI2Yo9pbdhjWhsAIByN46NN3fhgww6s2bgDa7f2YseOHXjppZfw0ksvAUgnaJg1axZmzZqFGTNmYPz48WJ7g+EmKYs12bImW2pqLdYArsVWDtWstVZNYrGYsCKTdY9h6hkWbgzDVAfOJskwQwa71YQ928dgz/YxAIBILI5PNvfgw8+78dHn3fhkcw+CwSDeeustvPXWWwDSomjcuHHYbbfdMG/ePMyaNQvjx4+vqStgMRYaAJqaXbWKJUskEggEAiKJBcG12HLTX8mFcmutVRO2tjFDjUERbocddhi2bt2a8ze9Xg+LxYLGxkZMnz4dRx55JI455picPsd//OMfccsttwAAjj/+eFx//fX97vu5557DD37wAwDAGWecgcsuu6yCI2GY/Bj0Ohy2YLL4POyR49vaWLjlYsT1CWbIYLOYsMe00dhj2mgAQCKZwobtXnz4+Q58srkHn2zuRbcvjC1btmDLli2ijpzVasXUqVMxbdo0TJs2DVOnTsXYsWMrEnPFiDW5ZpeiKGhqqp1rNtVii0QiGuse12LLpta11qpJNBoVbaTEJwxT79SdxS2VSiESiSASiWDbtm14+eWXcdttt+Haa6/FokWLNMueddZZeOWVV7B69Wr85z//waGHHoqjjz4677Y3b96Myy+/HAAwf/58XHTRRTU9FmZkYzIacP5XF/W/4DBAn1KBHd70HzYz0Oga1PbUKyOpTzBDG6NBj+njmzB9/C5B1OMP45PNvfhkUzfWbu3DZ9vSdd/WrFmDNWvWiOXMZjOmTJmCKVOmYLfddsOUKVMwceLEgkkfKFEFuUHmE2vkAilb1mo1+OdabMVBJRAGotZatVBVNcvaxjBDgUEVbnvuuSfuvPNOzXfJZBLhcBibNm3Cc889h/vuuw9bt27Fd7/7Xfz1r3/FggULxLIGgwE33HADjj/+eIRCIfziF7/AwoUL0dbWlrUvRVFwwQUXwO/3w+Vy4cYbb6xZhiKGGSmoqgqHLoHRiShSsQh0BjN0nE2SYYYlzW479p9jx/5z0rFuyVQKW7sDWLe1D+u29mLd1j5s7PAipij45JNP8Mknn2jWb2lpwaRJkzBp0iRMnDgR48aNQ2trKwwGQ1FibaAsNIVqsVFq/5Eu2DJrrfl8vqxlql1rrZpEIhHhvkmlIRhmKDCows1gMOQscuh2uzF69GgsWrQIRx11FL7zne8gHA7j7LPPxrJlyzQzIxMmTMAVV1yBSy+9FD6fD5deeinuueeerIfqddddhw8++AAAcPXVV2PChAm1PTiGGcakomFE31uJ8Fsv4ZqGdDa3HZ9thsHkgN1zIKzRidBbOaUywwxnDHo9Jo5qwMRRDcIFOJlKYXtPEBs6vNiw3Yv1273Y1OlDjz+C7u5udHd345133tFsx+12o62tDaNHj0ZraytGjRqFcePGYdKkSfB4PAMmklRVFan9uRZbNqXUWrPZbHUn1ghVVREMBsXfbG1jhhJ15yqZycKFC/Hzn/8cl19+Ofr6+nD33Xfj/PPP1yzz5S9/GS+99BKeeuoprFixAvfccw/OPPNM8fvTTz+Nf/7znwCAk08+GUuWLBnQY2CY4UTssw/h+/edUOPZLjHJeAiBN59GcPULaPjqd2GZOnsQWsgwzGBh0OtFPbkD500U3wcjCjZ1+vB5lw+bOv3Y3OXH1u4AegMR+P1++P1+rF27Nmt7DQ0NaGtrE4JO/r+5ublqg26uxZabYmqt6XQ6Ees3FEoghMNhcZ0tFgsnlWGGFPV/hwE44YQTcMcdd2DDhg247777cN5552XNel111VV49913sX37dtx000048MADMX36dGzduhU/+9nPAAAzZ87kZCTMgKHEk/jtv1YAAC7+2n4wm4b+iz/22YfwPngbkP3u1qDG4/A+eBs8Xz+HxZvEcOwTDFMMTpsZsye3YvbkVs334WgcW3sC2LojgK3dAXT0Bnf+C8EfTrvg+Xw+fPrppzm3SwlC2tra0NTUBI/HI/41NDSIfy6XC06nM0uAFarF5nK5RmRIRWZymGQymfUPSLsYGo1G+Hw+2Gw29Pb2IpVKIZVKQVVVzWfaLv2v0+kwderUARdNbG1jhjpDQrjpdDocffTRuO222+D1evHhhx9i7ty5mmXcbjeuu+46fPvb34aiKLj00kvx4IMP4tJLL0UgEIDdbsfNN9/MMyvMgJFSVaz8aJv4PNRJRcPw/fvOnaKtv+NRAVUH37/vRMsPr2a3yZ0Mtz7BMJVit5owfVwTpo/LzgoZjsbR0RtEpzeEHd7wzn8hdHnD6PaF4QvFkEgk0NfXh76+vqL2Ry6PDocDFotFE4dls9lgt9vhdrvhcDg0iVBMJlPWP6PRCL1eD4PBID7r9XrodDoxuUx/q6qaJWBSqRSSyaQQOPSPxBEVPKfPmf/i8Tji8bjmc75/iUQCiqLk/D7f8nIbasH8+fNx9dVX12Tb+QiFQuJ4rFYr19xjhhxDQrgBwJw5c8Tn1atXZwk3ANhnn31w5pln4q677sKaNWtw6qmn4t133wUAXHnllZgyZcpANZdhBozgK08h9NrTtd9RKgWouWso5UaFGlew48ZLgTLjQhwHHAXngfkzxTIMM3yxW03YbWwjdhvbmPP3eCIJbzCK3kAUvf4IevwR+EJR+IIxeENReIMx+IJRBCIKQtG0RS0UCiEUCg3kYQxL9DodDAZd+n+9Hnq9Dga9Li1adTro9TrodemJ9/S/9Ho6AEoihW5fGNu3bx/QNlOWUIKtbcxQZMgIt3HjxonP3d3deZc7//zz8frrr+PDDz8Uou0rX/kKjj/++Fo3kWEGBzUFJLOLm9YNagpIliL4MtZlGIbJgcloQKvHgVZPdpKzTJLJFELROAIRBcGIgkBYQTgWRzgaRzgWRyQWRygaR0xJIhpPIKYkEN35WYknEU+mkEgkoSRSiCfSf6dSKhLlPtsKYNDrYDDoYdDrYNwpiowGPUwGPYxGPYwGPQx6PUxGPUwGA4zG9G+mnb+ZjAbxm0n8ZtD+rllOD+PO74zydnbuz6CnferEvvUV1KFcu7UXF//puSqeseIIhUIiS6jNZhuRbrDM0GfICDe5/ovX6827nNlsxuWXX47TTjsNQHq2hwpuM9Xj3698jCfeWDfYzahr5CDu/7vh8ZplRjtSXYcja7LlweeRVz7Gs6/WsSgtkYHqEwzD1A6DPi1eAECFClVNFyxPplJQVex0i9z1m8wuyxNZpbT/63Tp32RUFYgnUognUkBsII6wtiRzlH2oNalUSlhaKZkKwwxFhoxwk4s6FhrspFIp3HbbbeJvVVXx61//Gn/+859r2r6RwuTJkwGkkywo8dr4vQ9HApHcRUmrQcyYBIbpxGEsnoQ/MQxGKjmoZZ9gGKY+ISGnQkUqObLjXCdNmjRg+5Jr8g2V7JcMk4sh03MDgYD47Ha78y53++23Y8WKdNa2iRMnYtOmTXjxxRdx33334eSTT655O4c7xxxzDBYuXIhYrP/BdDweF9ZRytBVaBmHw6GxrPbH9u3bMWbMGKRSKfT09ABIW1wTiQRSqRR0Oh1aWlo063i9Xvh8PkSjUTQ2NsLtdmsKb4bDYWzduhWpVAoOhwNjx44VQebhcFjM2Lnd7qxEN93d3VBVFUajEQ0NDdi4cSPi8TgMBgM8Hg8aGxtzppROJpPo7e0FkE5NTP1bPjd6vR7NzbkLWxuevQ1Y9qeiz9tQ4oADDsCSky4XCQji8Tii0ShcLhfi8bhwdWlpadFM6FAsg9xPfT4f9Ho9nE4nRo0aVXC/Xq9XZJkrtV+WQygUQjgcBgA4nU7YbDbxm6Io8Pv9wlpH/Yn6Jd0HMqqqoqenB6qqavpOOBxGOBwW5ycej8NoNMLtdmedw2Lp7e1FMpmETqdDc3Oz2EYymcybQlyv14taT/3t0+fziYm75ubmqtfRSiaT8Pv9mjTwua65fD/qdDo0NTVBr9fnPP/FIl93nU6X9xmRj8y+YTKZ0NDQUJEll9K/x2IxxONxhMNhRCIRABAJPfR6vUijXg/uZpnXIBqNatzigF2FvB0Ox7C3dMdiMUSjUcTj8Zzp+41GIywWC6xWa1Xup3LuAZ1OpwmBqSXJZJKtbcywYcgIt82bN4vP+Ypnv/3227j11lsBAPvvvz+uvfZaHHvssQgEArjuuuuwzz77YOrUqQPS3uHM6NGji1ouGAyKwY/H48k5+I3FYuJ7p9NZUJTnYtKkSVAURYgoq9WKaDQKIC2CMsWOw+GAXq9HKpVCa2srRo8erXlxbdu2TQwSx4wZoxng9/b2im2PGjUqa8bOarUimUxCr9fDaDRi9OjRUBQFer0e06dP1whEmWg0Kn5zuVzipdLd3S3OTWNjo2YwLxNoaEAw5y9Dn1FtbWidNAmBQAA2mw2BQABGoxE2m02klAbS10qn0wlBEgqFNMLaaDQiGo3CaDTCYDCgra0t7z7D4bAYjBqNRrS2ttZ0oKeqKrq6ukSh4ba2tqzBlKIoItU2kBZvTU1Nop2ZM9ehUEhkS3M6nTCbzfD5fGJgHwwGYbPZ4HA4YLPZ0NLSUvYMtMfjEYOipqZ0ZsBwOKzp14TFYhGD/2LPaXd3t+aerOa1iEaj8Hq9op16vR4ejyer3aqqYseOHeJ+bGhogMOxK66qHMtBJBJBX1+fuO5NTU0lZT2ORCLwer2iHRaLBU1NTWWdn1QqJQorK4oCs9ksnkMknA0GA8aOHSv6U70xadIkUYvNZDJpBufDvRZbZq01k8kEp9OpWYaem7WyNg2k9axUgsGgELAOh2PY9gNmZDBkhNv//vc/8XnevHlZv3u9Xlx88cVIJpNwu924+uqr0dbWhp///Oe45JJLEI1GcfHFF+PBBx+sy5fOcER2b813zqsxCJNTFcszrLlmguPxuFiGZo4JVVWFZddgMGSJSDoeEmaZ6PV6YRlqaGiA0WhEIpFAQ0NDwfbLM/20XRpA0XHkE23pHRsAo5kOAiqyzyu9tOTvM7+TZ2ZzXpdUMv2vVPSG9L8cZO5TTpmt0+lgsaaPOxKJiNpCNGNuMBiEUAbSYiUQCGj6gF6vh8vlgt1uR2dnp7DE5oOsL0Sl1otiUBRF9OHMPkmYzWa0tLSgt7dXpOfu6elBY2PubHtkIUkkEohEIppMapFIRFh39Ho9mpqaKhrIWSwWBINBRKNRbN26Nauv6nQ6IRLLsc7QuaG06tUiEAhoPDmMRmPecxEMBsV9ajabNaKtHGTrHZDbgl+IUCgEn88n/rbZbEIAFksymUQ0GkU0Gs3rQUHPOrvdDqvVKoR5vZFKpTQCnxjOtdgya62lcsSNGQwGIdaG4zkohkQiobFqZwpahhlqDAnhFo/H8eyzzwIAxo8fjxkzZmQt85Of/ESklv3lL3+JsWPHAgCOP/54vPTSS3jyySfx4Ycf4qabbsKll146cI0fwZCrWT6hk0kul45iKEW4kcUMQNYAUx7AkBuQvA/adr4XoE6ng9/vh6IoUFUVZrMZFosFRqMRiqLktbjJhV+NRqNGQAL9pyx2nXgpXCdeikQiga6urnT7pUFWNBoVrphOlws2m00sZzAY0NjYiO7ubvh8PqiqCo/Hg5bW1qzjTIW86DprFlQljKyI+5wnRA+d2YZRf/4Ieocn6+dt27Zhx44dANKujgajEZs2bUJ3dzcsFgvGjh2LtunTNXWLDAYD9Ho9rFaruFaJRAI7duzQCGCdTifqNZXiCuT3+8V1ttvtA1L3kUQWkN0nZYxGoxBviqIglUoJN0WZZDKJWCyGcDgMRVE0956qqqL+FJC2llUykUXCkKyBNFAE0n2LXA4rcceSrYzVIJVKwev1ap4FVqtV434qk0gkhPDV6XT9TsT0B7lG0/PObreXJAQzBafD4YDb7S5KtBUj1uga0vOKzn89TnjG43HxzM2cKHS73XXZ5kohsRaJRHKKNdkNeTgef6nI1rZS3wcMU48MCeG2dOlSMcA76aSTsl5Qf//73/H8888DSKf+P+aYYzS/X3nllVi1ahU6Ojpwzz334KCDDsJ+++03MI0foVDRUCC/0Knmvgj5RZb50iJ3EgBi8C9DwgXItrT0Zz0ksSUvN2rUKFEUVhZnmWRa3MLhsPiO4hCKIZflLnPfJpNJU8PI4XCI2dp4PA6HwwGj0ZjzmukdHjRe9A/0XnsiVBXQFSrCrdMDOh0aL/5nTtGWSCQ08UJOpxPd3d2IxWIwGo3ChY8KwQJpt1o69xaLBYFAAKFQSIhNwmazweVyZU0W5LI6ypCrEZDuH6W67ZaDqqpCQOTqk5lQvFpfX5+IHYvH4wgGg2ImmYoRJ5NJIQhoMB4KhcTxu3aK+HKgGCK6nwwGgygcTDFzpbhD5iOVSonrVo0BVyKREFZLwu12F5yFl58L5VoNCVVVxbUB0s+SUoSgz+fT3L+ya3U+SKzJVvxMjEajKEAtH59sqa0nEZBIJBAIBDSTHgCEi2Sxz8yhQjweF2ItVzFssmpTsfDhHsNXLLK1Ta/XV2wpZ5h6oO6F2zvvvIPrr78eQDq26lvf+pbm9zVr1ojfJ06ciJ/97GdZ22hoaMC1116LM844A6qq4tJLL8V///tfzWCPqS7FuEkCtXOV1Ov1WTP0sVhMDMBytSnTRU4mU/zI0GCMjlmn08Hj8cBms8Hv9yOZTAorXK7jpUEkCY1yC4TK5yGfcDMYDBq3Ebvdjh07doi2W63WgoMeyx5HQHfWXUjd+QMgnhYcGgG38/h0ZhsaL/4nLLsfnnM7lGiEjpEGlslkUrhlmUwmYdEB0n2qoaEBqVQK4XBYiGK6lpXMsquqqnE9c7vdAzIzK7s4Wa3Wou4HcnP0+XziWsqW3s7OTo3rJQ1kKVkJsCvmpxTovMsTCwRZlin5Q7UGzplur5VA8WzyMyJXPJtMOBwW4tRoNFac1EBOtELW7mKuuaqq8Hq9GqGSGWcnIyeGKVWsycjrDoT1uT+SySSCwSDC4XCWmzU9c4eLaKFnXyQSybrfgJ2u5BaLEGzD5biriWyZZmsbM1wYVOEmZ/oh4vE4AoEAPvvsMyxbtgz/+c9/EI/HYbVaccstt2hmiIPBIH70ox+JzGi/+93v8r7I9ttvP3z729/GPffcg87OTlxxxRW45ZZbanp8Ixn5hV/sDHWlrpKpVEo8mHMN3uVBT+YghNyHAIhZS5l8QpQGVNFoVMTguN1uIZxMJhOSySRUVUUikcg6F/QbkB5IhUIhcTxWq7UkEdKfxU2v12sy/NntdhEvRdYsyrxWcD/T9kXvD/8N6/vPwPH2v6Hr3SJ+M4yaDMfRZ8F2yMnQ23NbEmKxmBBJBoMBFotFPAtoMEkWEMpsR8cQi8U0bqt0rIWStxCFLG6BQEATw1TrLJJEsW6SmdBA1WAwCMtjd3c3DAYD4vG4cBUdM2YM9Ho9uru7NbGdpVh5EomEyH6YeY9S/FNzc7PI7BqLxao2sy1PRlTiKpnpXmgymdDY2FjQhTuVSlU13jEzg2RTU1NRx6SqKnp7e4WAlEWKTLFijQb6xTyX+4vrHSjkpENyH6QY1ng8PmD3bC2haxiJRHJ6aeh0OpjNZnENWYjkh6yUwC63bYYZDgyqcHvnnXewcOHCfpcbP348fve732H33XfXfP+LX/wCn3/+OQDgvPPOw/z58wtu58ILL8Rrr72GTz/9FM888wyWLl2Kk046qfwDYPIiv3QGyuImv9Cz4rN2Zk0D0i/7zN8z3SRlyB0NgIixou/lWXByrzObzZq4ENqvnL6ekM+TXq/XWNtKddXLJdxkl1VywyQcDodImR2Px0XWtULXi7anWl2I7/cNRA4+HbuNbkIqEoTe5oTO2X9WO7/fLwahdK5SqRRisRhcLhfMZrMQjyTmAoEAotEokskkXC6XSHNvt9vR0tJStssfkL4GcqrogbLE0zEDuwRsqRiNRrEdVVVFraK2tja0tbXBYDBoXANJ5PZ3jcitWHaHlMmVHZKSxVBbqnFvV2pxyxXPVmwiDzne0WazVWRxisViGhHo8XiKEk4Uxyhb8xsbG8X9kUgkxKRTPrFmMpmEJb0UN085kdNguUmqqopQKKSpwQXscq+mLMHd3d2D0r5qkEqlhODOF3dIYs1ms7FYKxL5fnM6nWyRZIYNdecqSYG1bW1tmDFjBg477DAcddRRWS+Ohx56CI8//jgAYO+998b3vve9frdtNpvx29/+FieeeCIURcFvfvMbLFq0qK7T2A5FZKFjNBpr+qKhQT99phnszAEKWZrIvSQzuyI95MliJpNIJLJcLMm1jkQbzaDT37kSmSiKkjUrLIutWCymSYxR6gw3bUuv14tzLgtDWcRZrVYYDAbNzK4smPIRi8XEObZYLGkB5WqG3pW7xlwmsusPHZ9er0cgEBAZFSkjJ2V41Ol06OvrE9kkyS3PaDRCp9MV1b8KWXNl0e50OgfMskCZMoHSrG3ArqQMsVhMTBgEAgERJxeNRkUJADmuk2qP5aOQOyS51lIcZCYWi0VY5eQSHZUgD9ZLtbjF43H09fWVFM9GUHIXoPJ4R8o0K/exYq43JTGRLeZ0/YLBYF6rDLBLrFWS+r1Yd/daoKqqSO0vW12pD7pcriEtYGiyJRKJaFz4ZSibsM1m4/T1JaIoimZSbDhYYxmGGBTh9sILL1S8jRNPPBEnnnhiyevNmDED77//fsX7Z/IjF/0s5YVfjqtksRklC7lJykH7uZKB5Ipvk+OLaBZc3k+umepcgywaVCaTSSFmyikQqqqqxrKWa59UiwlIDx5JKMZiMSFm+xNuskuoyWQqqhC73EaKxQJ2WS/JGkpxGm63W8Qj9fT0iIG2wWCA0+lEW1sbFEXRFCcvBVm0U+ZFIH3eBjJVdDlukqlUCoFAIMtt0e12w+PxYPPmzcJlePPmzRqBWyjtP1kdZTFJyMliCp1rEm4ARJ+qFPn+LuU6U40zObFJY2NjUW3KFe9Y7sA5lUqhr69PE8dYzL2dSCTQ09Oj8SawWCzw+XwFxRq50FVj8mGw4tui0WhWQXRg6Ndiy6y1lq8wdi1rrY0UZGuby+ViaxszrOAnA1N1CiXyyKTSB2ou4WYwGDQvd3Lfot8y2yTXaMo1k5s585xLtFmtVs1ycjyVyWQScVqZLmS033A4LERDOQVC88W30fcU90THYDabhRVAURSRjKOQ0JazcppMppKFG8XvUeZIymJJg/x4PC5cS30+n3AbjUajwr2NCkXLrm/lWtwGo2YbkUgkNHX6inFhC4fDWRYIygDa1NQEn88Hj8cjRAdZ4FwuF9ra2vJmWe3PHbLYRCPy4L6UflGIUi1ulN1Vdjmm81PsPZVZs62S2Xqv16vxPijGRTMej6Onp0ekuE8kEnA4HFkZFIHqizUZ2T1zIERELBbLyswLDO1abFxrbeCR3YZJCDPMcIKFG1N1ynWxqcTilkgkxIAol5skkTlzTIkXgF0WnUxkIUpuZISclU4WEJnWP9pGpgsZ1ScjYaXX68uy+vSXmCQajWqEIVm5aL/kWlVoUCm79NDLUFVVTVKYfFByATpeo9EoBDZtk7avKAr0ej3i8bhw+2zdWVcus7YUUL7FbTBqthGlWNvi8bgmGyGwK8bH6XQKK1kkEhF9mPo0peiXIXdIORGOvN1C7pCFIOGvKAri8TiSyWTF1pFSLG5k3ZJFY6mFqam0AlB5vCPFZQLFuakC6cmNjo4OkWHVaDSioaFBsx65NNdCrBGUtIj2V8sJDdntV2Yo12LjWmuDR2YNVLa2McMNFm5M1SllprZaFjcqzgwUdpO0Wq1C5KiqqnGVy5VNkrJBAtlWhMxMhvKxZNaTI7EXj8fFPig7Yjgc1rgwlhO7IQs32lYqlRIDsGQyCR0A3e9PhnfDKti+czPUeUcV5SYZffNxhF/6J2Jr3waCfYDNDUvrRKi7HwXMXYJkMtlvmwOBAFKplCgInUgk4HQ6hTij1P7kjkcJZAwGA6yr/gPzTTdDN30fGK5elnV+y7G4DUbNNplihBu5RWZm3rXZbFnue+T2SqUnPB4P/H6/GCBS3TedTleRO2R/WCwWcT/FYrGKY0voOvcXy5gZz0Yut6VmkqtWvCPFZxGFMljG43HhGizHwplMJmEFJrE2UPFOAxHfNtxqsfVXa43iT0mssaCoPnLcJ1mjGWa4wcKNqSqZhbdr/XKShRsJNnmgkemSZjAYNMKNHvRUwDRzkELujZR9kURXrnTchSxuhDwgIktbLBYTA7JyUxbnsrjRCywSiaS/e+52YMOqdDt2Hhe5SVLSD5lUNAjvTd9GbPWz4jsdAAR7gGAPdBtWw/nSXxH74V9gmru4YNvk2Cd5cB0MBkWRbbLy2O122O32dJzL1k/hefkOAPnPbzlCYzBqthHk/gakhU6ugXg4HNZYBAEI60suyyCJMb/fL5K4tLa2Qq/XiyQ6O3bsgM1m01h0qe5atayNVBgdqI5wo/u70PWpJJ5NJhQKVSXeMR6Pi/hLAKIYeeYy5D5HmSGDwaA4BovFgubmZlHTcKDjumop3PLVYjMYDKIo/FARNcXUWiPrKNdaqy3kJk1UWnORYeoVFm5MVSm2DEAuKnWVpMGdLJQyLRvygCQWi0FRFCSTSZGUJPPFqiiKqL9ELwKPx5NzQEpWATnTJQCRHELOtkltJmsKFfct98Weq/g2ic5oNIqGj5+D7snf71pecs/M5SapplLou+EUKO+/mP7CbIM67wigbTcYIz6kVj8N1dsBvXc7Ar87Gdarn4dx7PScbfP7/cJySRk6KRkGWYn0ej0cDgfa2trg8XjQ29sLo78Ljf++HHplV3Y/ohKLWygUEgPhgazZRhSytuVzi3S5XEKQZUITED6fD6lUSkxQmM1mMaCURTwAjBo1qiYZNE0mk7gHKi0LQG64QO74tmrEsxHJZFIz6Cs33pHS98suxSQAM8UaEYlExDGQla2trW1Qk1PIXhPVEm6pVEqk9s9Vi81utw8JYcO11uoTWTwXkyGZYYYqLNyYqlLqTG01XCVpgEeumfKLMnOQnBmvRn/ncpME0skFyFpEFo9CA306nsyaQ2azWRSPptifUCgk9m+xWCpy66AXlsFgEG2Ix+OIRiKwLb8HplfuBqTBUmbK9syXXPjZO4VoM7ROhPmCfyLsGAUAcHo80J98JXr/8D3o/vcsEPLCd/sP0fzLp7LaFYvFRJwPDXbIEkMDfFVV0draCrfbDZvNlh70f/wGGv56HgyBHVnnFtC60JXSh8i6QXV9BqpmG0EiC9g1Gw+U5haZCa1LdQLD4TBsNpsQ85TsJhKJiJn/Ytxby4EstxTbQwlnyqGQOM8Vz2a328sWXNWId6RC2XJsGFmO87nPkau20+mExWKBy+UqqTh6LaAMt0B1vCaKrcVWz1CtNTkDcSZca23wYGsbM5Jg4cZUlVIySlYDstjkSkxCLnhA+qUqD37JIkAxViaTKWuAGQwGhUsdpVPvz5VRr9eL2DUZOQMj7VN21yumKHI+5P3Js/RKzzYY774Q5nUrcq4Xi8XEQFcWbmoijuDDvxV/e374F3gbxgI7z6XVakXKbAZO+y2S20+AoeszKB+9hth7L8Ay/7Bd29npukfZCzs7O8VMu8vlQjweh9PpFPFuAGA06OH7z++h/uvXMCRzpz2nbQPFu0nS8sFgUBNPONBWDblWH1k5S3WLzLXNYDAoBHHmQNhqtaK5uRmqqooYqlgshp6enrKsU/1Bwo3aVm3hFo/HNQKp3Hg2IhqNViXekUpd0HNHp9Ohp6cn57JmsxnxeBxWq1VMBLlcrroYcFbLTbK/WmxOp7OuU/tTAicSa1xrrX4Jh8Oij1kslgFNNMUwAw0LN6ZqkBUHyE7JX+z6pS5PLni5EpMUckmjumGKoogYEnldEm0k/DweT1ExL5nufPS3PABSFEUIR/qtEne9TDdJVYkg+N9bkXz0Rph3uhnC6oR5tz2gfPiqWEev18NoNGYVJI+9+xxS3s502+YcCP3UPZHo6hJtpTpuMJoRO/hM2Jf+FAAQWf6ARrhFIhGEw+GseBaLxQK32y3c6Chphu7jV+F98iYkNq0BtSYydV/YPnsj65hpUF/KzDZl0aTaZgNZs42Q+6TBYEB3d7dmsqM/t0iZVCqFzs5OcZ4pcyG5K1KsoCxOm5ub0dvbK6xh3d3dYp1qkVkWoFwxIvdrur/D4bAmgYjBYEBjY2PZAiNXzbZyrCV9fX3YsWOHuLcbGho0zzPZfc5isQiRR/tqaGgoW3hWm2oIt3y12Gw2G1wuV93WKCPXcvrHtdbqH1VVNe7SA51oimEGGn7qMFWD4peA4l/4lbjh0L76i2+jxCMy0WgURqNRJByRs3yFQiH4/X4xoHY4HEW71OUTbpkJSuRU7JW6RmUmJom8/jCC//q1ED/JsTPR9MO7kHz7cSHc4omEsLJluknG/vec+GxZeJSmnAItS/F8ifb9Ab0BSCURffspqKkUdHo9FEXB5s2bxeCHzqXZbMbUqVOFpVRV1V3n5s//BzoS1eKA79CzEN9tEWy3fV3TPnKvBIoXbnI8ITCwNdsImsEnt6vM+K9i3CKBXcWyfT4furq6xPLkxutwOPLGC5nNZrS0tKC3t1dkHCXLW7VimQwGg8gcSkKmHDGU6Vbn8/k019BsNqOxsbEiS4dsDbJYLEVPoFCfJvc5WUxSzbFcsU7kTkmTNuSuW0/Z7yoRboqiCFEqU8+12LjW2tBGfpdarVa+Psywh4UbUzXkl3UpD09K3FGOxQ2AcDmS90tJR4D0gCzT1YoKQZtMJmF1AnbN6AMQmQ7tdnvRx5MvgQZZIJPJJHw+n5ipNZvNFc+0Zwo3slOkbG6EDjgdliO+A/tuUxF4+3GxXDKRgC1PGYD4+nfFZ3P7IvhzCDc6JlidwKjdgI61UMM+KJs/QqRhHLq7u4XgI7cxqjtmtVpFPILRaNTG/uh0UPc4GrElPwRsjdB1fZ51vOVklKRyBACE1WOgIctYOBzWWDkpIUV/xc+j0ajIfEjFwxVFEQkwpk6dWlRfMhqNwvIWj8eRSqXQ09NTVQFhtVrFLDhlTS0V6hepVAo+n08jRB0Oh8iGWi4kgIH0M6i/CRQa4JM1hlyUKQkQkI6Po9qOmYkpKHGJnPijqamprty66HiAXYlmiiEej2vq1hH1XIuNa60Nfag+KFEPrsYMU2tYuDFVo5KMkuVAYk92+6OBRn9ukkBaUNBvVGtNTuNtNpuFwKpUuNH2KEMlDdbKKXScSaZwUxtakTj6fPTNXgLV6kJbcwsAIJlhwaCCyZmDs8T2dbuWa50IJborRbrcVrJ0qE1joetYCwDo/XgVknMaREKXZDIJp9OJWCwmXPnkwR1lPgQA0/4nQX/4dxBtnoyozwebzYZgniyKRDEDy8yabYPxclcUBdu2bRPnhSYT+suml0wmhdgjIUOxg5Tkxmw2Y9SoUSVNABgMBmF5o+yPfX19SKVSVXHZs1gsFQs3EhGBQEBjxeovQVAxkItkfzXbKoKQlAAAXc1JREFUZLGWOcCX3SzNZjNcLhfGjBmT0wJIlk26V+leqDdBUKq1LV8tNqPRCLfbXXeZ/bjW2vCCMhMDKGmClWGGMizcmKohzyQPxAOU4ttUVRUihL6X3STlGW1KzACkX+Iej0fUdpNFm9PpRDgcFm5exYorWUhkWhBNJpMY/FFGx0wxVA5yXTSDwQB13mHwm8dB3Rn7R0IlIQlr487rkzmwUhMK1GC6GDYMRsStDUDUm3NZGqAmHc3iQaL6dyAUCokEBPJ5IJFK1yZzYOQ6+7Z08oydWe1MJpOmH9HZLMXilhnDRBkWBwpKM09xfsCuxCMulytvW8idNjPOhjL00cCSLMblCBmy+Ph8PtE2n8+HZDJZcZwIXVv5fisVcgel+7vSeDYZyuYIZNdsK8Z1TqfTCUGqKAqamprQ0tKSsz8mEgn09PQIoWAwGKoeV1gtihVuQ6kWW7G11sgSXy/tZgojW9vIo4NhRgIs3JiqkEqlyk4hXYmrpJwGnwZCmZn75MEUDYQTiYSwdFBqcYLcI+mlUMpAsZDFjVKyUxvJNasS4UYWR3k7wWBQfOd0OkVsTVwatJjyCLdU2C8+68z2nPFt8vHF43Eo0IsHSSrsF4KA0s97vV5RV0ee5XY4HJqYJSrroCiK6EOypYb6RynCLRgMCnFP9eoGAlVVEQ6HhYsmiRej0YjRo0fnjJnMdIfMxGq1ikycQFpkNTQ0CCFXDhRjZTAYhPsq9R+Px1P2IDazBAaVKigGsijKRbWtVmvVMmBm1myja0GW2UJijcp2yFkGE4kEmpqacvbFeDyOnp4eTdbXpqamuk1q0Z9wK1SLjVL714PwKabWGtXu5FprQxO5D2YmYWKY4Qz3dKYqDHQZAGCXcCO3P9pvITdJGkAnEglYLBbxt+y62NDQoNlGqfF6RObgL9PyYDAYhCWhXCjBBwCRbCUQCIjEF2RtUxQlnTiE2onsEgkAgPiuNurMVtFm+fwmEgn4/X5hDVGNuyyaRqREin8SxZRlz2g0arKOWiwWYeU0Go1iX7FYTFwPu90O6lm5hFuhQWIikdDEPwzUjKyiKPD5fJp7gmp1Wa3WrFiqXO6QBGWHJJFLQjqZTIq0/5X2ISAdGyKXqCDLcGNjY9mDWqvVqrmmxdxHyWQSfX19IqkJkO4DLS0tVRMEVKhcVVUYjUaEw+GCYo0G91S+gcQ1kc9qHovFhPspUH5x8IFCTiJkNBo17cyciCDqqRZbMbXWSKxxrbWhTTKZ1MSnsrWNGUmwcGOqQjVSSJdqcQMgSgGQeyZZLYD0oFd2k6QsdwCEoAiHw2hqagKwq4AvUH68Xj6LG7nrkFsmFZ+u1O1Ldv0xGAx5i4pnxqDQ71noJFdPqBqLB1lCyD2KjlVzzDuXj0Qiwt3UbrenXTil6+twODTZIcmFkgaPJCSsFktB4VZo8CVbbBwOhzgvtbIIUMKQzHNN8U/kzkuD/HzukEB6kO9wOITLGZVVICwWi8baXA3IjbRatd4yywL0N7hSFAV9fX2aWD6XyyWKpVeDSCQCv98vajg2NjZmieVcYo2Ix+Ma67zb7c5p0YlGo+I8Auk+kM8qVy/IVsRMt/N6rcVGmVoVRdHUiZThWmvDD9naNtDu7wwz2LBwY6pCuRklgfIG0jToTyQSYjCs0+nE4B9A1qCL4niA9IAkEAiI5ByUDa7S48kn3MgKRjPZsVgMiUSiqolJDAaDcA0EIFJXy2JWJpd7nc66KzGFqqStJeSO2dXVpTkmiq+yymNRYzrZCLk6JhIJUbNNLphst9uzLJCZbpJWqxXGlBTjVoJwy4xhcjgcmhjGakJxZ8FgMOv8NDQ0iIElkO6T4XAYoVAoa8BPx+xwODSCPhaLZdUak2ebqykGqFh3Zq235ubmkvsq9fVkMilEQb57nUpwyGUeyEpb6aCMRGgoFEJHR4e4P+SslMXEOZFLNbWRsnlmFtnOTHJktVrR2NhYFy6EhcicfItGowgEAln9dLBrsdHzLBKJiMQ68iQQwLXWhjOJREK8y9naxoxE+InGVAV6uZeSyKMSKBU37bM/N0k5YQm98Knmi81m07ivZboMlTIwziXcKJMZkLZCUMxTNYSbPAtOdbnIDZTi/ijmT1VV4SqZr0C6ziq9BOMRYRWS453k1P6BQACI7xKFqsUJRVHQ0NCAeDwu4ggpAyIA4aYkDwhpMK0oihAtNpsN+pjWXQvoX7ilUin4/bti9RoaGop2rywVElWygM7MFknucrnEMy1P7pCZ14QsPHL2Q5PJpCl1kc8trFyo1hsl1Egmk6JQd6kWYovFIiy0dM/JUPIYeVLFYrHA5XKhu7sbQGlF1uXtZhZSlic15GLYxSSloKybchxvrjjFYDCo6Xtkxa930QbsEm6UyTPTEmmxWOB2uwclqQr1H7qeXGtt5EKToMCuGG6GGUmwcGMqJpFIiBdpOa5/NKgpxVVSju2ilOhyEgj6jqAkCTRbRwNkl8uVlYRBLiRe6gAgl3CTB3Iejwc7duwQ+6mmxU2O32toaBAxdCQaE4kE6IzkOy6dXg99wyikfF1AIg7/9s9h8owS58dut2uyIep0OsDXIdaPWlwiwyHFXtGMOK1D6eZl4UbXU1EUES9jNpuRlAe8OxPY9CfcKIYJ2FWzTRYG1SCfW6TD4RBukUB6kEExW5n3RqY7ZK59kOUL2FXEWLa+2Ww2TaKNamE0GkW5ALnWW2NjY0kJXuRznync5Hg2wul0wuVyafpGsRa3XGKNIFdlcp8eO3ZsSYk0yL0S2JXKP3Ndv9+vcWeleNmhAMWwhUIhkdGVIFffwag3V2yttVgshra2tgFvHzOwyJOger2+KqVLGGaowcKNqZhqxLeViizcyOImD9YyB8MU+xUIBMSgmupfZQ7AKj0eil+jWWI5o6Db7UZnZyeA9EuoWsKNZsfJskWp9GVLjzzwyXdcyWQSGD0N8HUBAAzeDljaJuSdbdfpdFB7tgpLXrJpAmw7C/dStlA5o6DFYhGfaXBObabzQbWUMq8LuVvSceRyE8ys2ZZr4FyJ9YPcIuVZXyB9PhsaGoRrKg2CqU4aHXs+d8hc++nt7RXX1WQyobGxEQA0MZy1zJJpMBjQ3NyMvr4+4ZLW29uLhoaGogdMdMyZZQEURdGIUspuSVZy2dpTaEa9kFgjdDod4vG4uD4NDQ0luVfRtaRt5Yr583q9mskBt9s9ZFy4aIKgt7cXwK7YxMGqxVZOrbVauUEz9YU8ScXWNmakwsKNqZjByCiZaXEzGo0ay5bsJkk1h2RLBaXlzjWbX+nxkHDLdNkjS0w5FsZcpFIpMbCJx+Mi8QkduyxmU6mURrDoM45bk+Z79HToPnkdAGDp+BhtBx2Xf6Ae8UO3Y2N6GxYnbBNmAtgVywVAxCECu6xtsgAja5ycTTLfYFG27uYSdpmxYPRir/RcA/ndIt1uN+x2u7DCUf0/siACu0RQMYH05JZH/ZDqflEMp2yBq7ULHlmXyq31RpMqiqIIV95oNKqJZ8tV16yQVZUs63KMU6790uBeTrpBVs5ioQyhRENDg0Zw0zWWRVspwnYwoQy0sjAF0v2KRPRAuXhyrTWmPxRFEZNWBoNhSNxjDFMLWLgxFVNJYhKgPFfJeDwulpfjxoC0KJPbIaeuJ2uUw+HI6/ojZx8sV7gB6Vg6mhGmWDo5G6NsZSoH2coWj8dhtVqRSqWE6DGZTGJAFovFYMmxn1xZ41KzDoLh5b8BAJyfv13wBan/+BXo1J2xfLvtDbPJBIvFIo6RMmgC6etCbct0k6RSAeS+lu/ayMItc0CfGcOUryh1qYO+/twiE4kE+vr6siw+iqIId57Gxsaii1oHAgGNVU2eYChU6qJWkDVMr9cLV8BSar1RHJ6qqsLaLP+Wq+SAbGkxGAzCchyNRosSa3TfkTWJKCXejNaVs9fJfYqSleSzGtYruWqx0X1qt9sxduzYAfGcKLbWGok1tq6MbDKtbSzemZEKCzemIuREHqadLnIDgex2ZbFY8g5o4/E4tm/fronBo6yF+QrMyolJynk5UMHrUCgk9kGDdjlJh9FoRDweL3ugRyIlGo2KbcrFz/V6vSbuzZbDrdDv92sGTTqdDoaZB0B1t0Ln3wH1w5ehfPomzO2LsvavhIOwLP+r+Du11/HQ6XRoaGgQmfai0agQfrIApLZTvTcq65DPTZKQBbvc1+SabTSAlinH4lbILdLtdiORSKCnpydvdkiqVQcgr4jMhAbUtJ3GxkYxeVAohnMgcLvdMBgMJdd6s1gs8Pl88Pl8MBgM4l6geLZ8GRwpy6jX683KGkgYDAaRtp/EmgxN2ADo1z1VhtxC6blBrsJy++RrTy6UgxEHViyFarEZjUYRu1rLflVsrTWbzcaFsRmBHHJgMBiKfp4yzHCEhRtTEZVa24DyLG5yogCLxaJx9SEhFI/H0dHRIQZXJBxCoVBeq0659dtk9Ho9otGocAe02+1iX3K8l8FgqCgjICVRiUajYgaSrHdy1sZUKpXbOpWRxpzSfPt8PsSOPg948AoAQN/vTkXT5Q/BNHm+WDYVDaH35jNh6FoPAIiNnonkjMVwOp3ChZMG3xSLJr9sqW2Kogjh3Z+bJC1PyMck12xzOp0FrZjFiPF8bpEOhyNrUE/QgIKySZJ1qdjBcDQazXLLk/tooRjOgaKcWm/kwipbSxsbG3NOWJBlrbe3V1iEMl3j+hNrRCQS0bhWUTH6YvB6vZoJHDmdfzKZRE9Pj+gbOp0Ozc3NAy6ki4WeEX6/P2ctNovFklW/rZrQNSWxlus5bzabhbWUa3IxmcjWtnyTPQwzUmDhxlRENYROqcixQ3q9XiNSqLZYPB5HT0+PiD0xmUwYNWoUOjrSGRDzWdOqEa9HBZOB9KBFHjCSSxKl4ycLUjkvokQiIVJjkzggQWsymYQVUlEUUUCb9pLIuG5ut1tk5lQUBdj3JODdZ4BPXkOqrwPdlx8K697HwThpDlKBHkRWPAK1L30uUyYbgsf/DFa9Hk6nU8T1RaPRdJ/45WFQ+7ahA4DzxMvg+tpPNMKN4pDIja2Q1SIej2cV/s6s2ZYrKUSxkwLJZBI+n0+Tup8EsU6ny5nB0Ww2w+FwaCyFoVCoJJGVWdjZ6XRmzSoPhptkLkqp9Ubp8eXC8w0NDZr25xrYUz1GnU6XtgLvFGvkBtkfmfGlcrxjfwSDQXGuyeopu/5SmQRglyivV9FWTC02edKrWseRq9ZaJlxrjSkGuQ4m9RmGGcnw05KpiGpnlCxGxGS6M8ptsNlsSCQS6O3tRSKRQCwWEzWX5Jlcs9mcczBRDeFGySmAtNuPfF5olt5oNMJoNIoi4uXsS06NTG55JNx0Op34TIlc1EgEcrQaua3JL0IxyNLpYDvvbqTu/iFiq54BknFE33gEeOMRTRuSrhb0HP9LoGWSsDSRZSgajeYsPEzHTINfWcz2F8tCJRRIsOeq2dZf/8n1O9X5kuN+1J3lB0hgZ27DZrPBbrfn7PeliKzMWCqbzZYVD5dMJvPGcA4G/dV6U1UVXq9XnAc6Ry6XS1hjySKWywpD591ms6GlpaXkZ4scs0mCrxjIMkXIrqqZmTCNRiOam5uxZcuWkto2ECiKAr/fn2XRz5UdtlrPcLLA0nXN59rKtdaYUmBrG8NoYeHGVEQ1Cm+X+iCOx+NIJpOi1pdsHTGZTGIwqSgKDAYDGhoaYLfbi3LrpGVkt8NSkNPvkyuSjFzAlwSK7D5ZCuFwWMTMUeITggat5H5nMpnglM6zzWaDJ0cpBPlc2hqaYLlsKaJvP4nIS/dB+ewdpHw7AJMFautkxNoPRHjh8dBbnDCbzcL1k0SGsGpm7IP6TCwWE+vQgLG/1ONUEoCEm9/vz6rZlm+9fNBgXS6tQH0nU+DL7pD5XLoSiYToR1SWIR+pVEqT9t9sNucs7Fwv1jaZfLXeXC5XVsKJpqYmkXQlHA5r6p3JkGWN7gmTyVSymKCi8QBEzGUxJBIJTVp5l8sl+mMsFtOIa5PJhObm5rqLwaKSJ5nF3gvVYpOfeeU8h6gwdr5aa6VaSxmGkJ8jlOCLYUY6LNyYsqGZdqB6ZQCKsbjJgxJywQLSA0mv16ux5FDWO7vdLlzR8oky+XgKxc4UQh6QZlqPyNIEaFO5K4pScrB1MpkU7piUGll2eYpEIggGgwiFQuJYk0eeA/tpv9QUiJaRa22RKAYA617HwLrXMQDSQqOrqwuxaBR+nw9msxl6pF375DZFIhFxTM2/f1czYKN2UnxbNBoVFqZM4WYcNQlj/uVHKBRKJ5qIRoWFMrMYa7EDdDrviUQCfr9f9CfaHsXjyecolztkPooVWWSVkq3HuQo7l7LNgSaz1puiKNiwYQPsdruYTLBareJcU/8n6yywK9soxawlk0nRR0qNd8osCSEXiy8ECWh5EoBcnCORiCaGMl8mzMEkmUyK1P4ylHQkX5+RLd+lPPPKqbXGMKWgqqrG2lZsVl6GGe6wcGPKZjAKbwPImk0GdqXEJwGp0+lEanmybiWTSfF3LeLbaMBJtdocDodmBlquPWe1WoWLWK5U2P1BBcWB9EDSYrEI0SoPkIFdCVzGjRtXMLW/oigaF89c54jSwAeDQbEMJUahrHVkEaXYn8y+Qb/H43FhAejPTVLOrEgZBoPBoBiU9xfDJFvNaEAQDAY1aebpWEjokjukw+EoqT8UK7Jk0Uhp/3MdQzweF9fabDbXXTwQtX3btm0iA2RPTw9MJhNaWlrEZIDZbNZYNSl1fua5LVTDrT9CoZBmhr6YWk9UN0+2hpPVMxwOa6xwVqs1p/vvYCHXYsu0DJNgK9TWUp7hXGuNGUjkfmY2m+s6YyvDDCT1NQJghhTVyCgJlO4qKZcCoOLSfr9fDLaMRiPMZrOYfbbZbJp18rW1UuFGaeNpwE8xWIQ82DGZTCJWj9zMShmkZs5EkkWDEjyYTCaoqgqj0QiPxwOLxdKvVU8WxLlcFkmYUgyfw+EQGS1DoZA4Prl+Xa6Bs+xKmEwmi8omSWKFMlZGIhGYTCbhzlisxTIWi6G7u1sj2IC0xVAWkVSzq1ThEI/HxXmwWCx5rT2hUEjjztfY2JhXkNWrtY1QVRU9PT3CrSkUCsG0s55fOBwWwp7abrFY4HK58s6gy/dMKRa3RCKhuS+KqTEHpO8l2dJMVk9KrELY7faS6sDVEpq4CIVCWULX6XSKSav+6E+4lVJrbSAKwjMjA7a2MUx+WLgxZVOLjJLFZP+jQRaJJK/XK+KpyN2su7sbwK44M3nWPF9bK7EgxuNxjesiDVLl48kUbrK1TbY+9YfsSkap9rdt2yb2n0wmYbVahcWhv9pohBybl6stVPMtEomITHAknoD0daGkKFTzK1OM0TFTDFw8HhcuaYWEG8W0UUbReDwuBubFuEjG43H4fD6Ew2FYLBZhfaUEI1TXr1h3yHzIrmr5RFZ/af8zkTMc1pNwo9p5XV1doo0U+0XucyRkR48eDbPZjI6ODo1Lbi5k17tShHNmzbZiJl/keDsS0AaDAX6/X+P27HQ662LwWKgWm8PhgNPpLOmcyfFt9MxLJpOaLJ+54FprTC2h+G0gO8EXw4x0WLgxZSGLDsryVy6lDJLJxY7aQLWJHA6HyPJGFixgVyyZXD4gn2WDtkup+kshc3aQ/s5ncaOMkvK+ixVuZN2iOm6hUEgcXyKRgNPphMfj0aTO7y/pB7kvAmnRmnk9FUURcXMkIJxOp0jiQYNxEtMkgjKvLcURxuNxcQ7IQtrfOTcajUilUggGg2KA2t8AnWKetm3bBp/PJ+rKyS6Q5bhD5tuXLLJynfPMtP8ul6ugtZCSvQC5r8tAQyUoIpGIKE5OfVyv18Plcgkrq1wY3u/3iwLVVMIiX1KeclwlKe08sCtban8oiqIR0FQSw+v1agS42+3OWWJioIlEIggEAprnCPXlYmP5ZJLJpNiWwWAQlrV8otpsNguxxrXWmFpB1mSiHiZMGKaeYOHGlAXVHwMGNr6NRBn9A3YNXpqbm2EwGDSDMZvNphFy8kAx0xqWa5liUBRFU+jX6XSKtPKFhJssaootxE1uoV6vV9Q+owQtVqtVuDACuwRxf7XRgP7dJMkNMx6Pw+lMZ5F0Op3o7e0V+6dskh6PJ6vgNhGPx8WxplIpsa/+hCWQPmfRnclJKEFJoaLKgUAAXV1dCIfDwqVMp9PB7XYLgVGOO2Q+5BjBXJaIXGn/+ysKXQ9ukrJYo8kNOqeExWJBW1ubqA0GaNPnU603+TpTqY5MZItbsYlFMi2Y/U0GJZNJUUQcgOgLfX19mnvB4/GUnDio2hRTi60cYrGY+Gc0GnPGrZlMJhG3Vm+xlczwJBQKiWcAl41gmGz4ScyURTXqneWiP1fJWCyGZDIJr9crZuKo1pPBYNDUu6KEF5kDzFxZ0Co5HjkOhurMUDKUXMLNYDAIt06KgysmQUkymURnZye6u7tFCn2KJXK73VBVVSOkaKBVipskLZ/5G6VxJ/dHeXBsMBhERsFEIiFm5vMl2qDrk0qlii4DAKStL2QJUVU1bwwTue9lFsum2Kq2tjYRd1VNCrlJFpv2X6YYC16toEQU0WhU0zcp9oQG+xaLBQ0NDTlT4+eq9RYMBoXopvjITEq1uMklISg7ZSEoGYl8LZxOJ3p6ejSug42NjQN6zjNRFEUTf0fkqsVWLHKtte7ubtFn5eOkLJ88aGYGGvKoIPqb2GKYkQgLN6YsqplRspQBNLnrxWIxEfc0ZswYIVIikYgQfxS7lJmYJJdwK/d4qIAwkB7w0IA9U7jJn+XBkMlkEmKUarJlkkqlhFuanLLcbrejubkZTU1N2L59OxRFERkt6R/QvyiS3U/lmDVglzsqWazcbjccDofGgqfX6xGNRsU1ITfJXMRiMWGt1ev1olZXMbP5cupxEqyZ52jHjh2aItrALsFGRc/lVPTVggqOA7uyeMq/9fX1FZX2XyYWixW04FWbfGKNoHIPRqMRDodDuCQWciPMrPVGST/IjTZXUp5SLG6KogjxUWxJCJ/PJ+5ZqvNI7QPSzyNy6xwMMstUECaTCW63u+R2UVwoXVvqU/Izj5IXca01ZjCRk+3Y7Xa28jJMDviuYMqi0kLV+ShkcUulUtixY4dwy6NZcVkkZFo9aNACpAdpNNOfSTkWNxI1hNvtFoNxGoyqqqqp3wZAc75IuAHpcypbasjiEggEhCWRBrUNDQ1oaWmBw+HQFLSWU9nT/5W4SYbDYYTDYUSjUREPlhlzQJa2VColZuvznUNKW55IJPLWbsuFLPhk61MikUAoFEJvb69I70+YTCaMGjUKTU1N8Hq9Oa97tZAnDDJTsPt8vqyshcWIsIFwk4zH48INMperHLCrvlc0GhUz4Hq9Ho2NjUWJiMxabyaTCT6fT8RIZl5/OWaukLilOnhEMXFewWBQPCMosY1cCkCv16O5uXlQLE3l1mLLhyzWMieryNXcarXC6XRi9OjRnBGSGVRkaxuVZmEYJhsWbkzJZBbervSFX8z6qqpix44dIsaJRBi5JgK7BASQnkE2Go0iYQZ9l2/b5SRayawzIw9A5WNKpVJ5hZs8ux2Px8XgLBaLiSyORDQaFfE2NJimzIwkUB0OB5LJpDiGQrXRCNkiKR8DxdPRy5Qy62VuT459InGXC0otTp+LKQMAZBdVprjFnp4ehEIhBIPBrAynLS0tOd33gNLLTxRDPpGVKRSampqKmuigcgVAtgWvUooVa+QuR1kM6byZTKaC5QtyQYKVYjPJct7T04Nx48ZplpWFWyHI7ZLa218sGt1ThMPhgM/nE88yEpgDPctPA9ZQKJRVi83pdJZkIe7v2tKkh2yZL7Z0AMPUEtlTgq1tDJMfvjOYkqlFGYBCUOwWuVEkk0khIORBsjxTTYM4WZRYLJacgxnKzgiUZm0rVGdGHnQWEm7y/ijNfSAQyLIOGY1GOJ1Okf2N1qP4MhKolPSkWDdJOTU7iWEiGAyKgttWq1UkcMhEzixYyNpGiUnIakaui/2dcxqg0z7IPdVsNiMajWqscBRvlfnSL6bMRLlkxlXSPRGJRDRCwePxFH2/0HEB2Ra8cqAyDjTxkYvMrIGpVCrLUmmz2Yquj5YJWcj1ej38fr9wIXU4HCJmUk7oU0i4USkC2m5/CUkSiYQmm6fVahX1CAGIjLQDmS2xWrXYciWPkclVa83v94tts2skM9hklrhhaxvD5IeFG1My1YxvyyRzgE2iLRaLicG7yWQSqdFp/6lUSlg9qG5ZrrbmGrSWczxy5iur1Zq1XqZwkwdUsqig0gM0sKRaaHJ73G63psC1bH0hixutJ2+7mIQWskCQl00kEvD5fIhEImJWPlf8EFmG6FrYbDYkk8mcYoxmVMmyWEz7EokEvF6vSD9vMBhEnI7NZoPFYhGi0uPxFOV2WW3rgmxtI2GrKEqWG18prm7VcJMsR6zJ68ouhED10uI3NDSgsbFRJGuhkh5NTU1Fx7eVUrONEsPIglC2xJvN5qLdV6sB1WKjSRGilFpsldZaq+UznGFKJRAIaO5nLjfBMPlh4caURGzNKzAmkzCPnQtFUaoSC6LT6WB+/X5YrBbojzpLfC+LNvPr92OMxYKGxomwNtiwo8Ej2gMA6tS9NRYK5cNX03+Png1Am3TDuHFVeja75QDE1rwCfSoF85g5BY+H9mOZc2C/ma9ia16Baec2U5++AeWdjXBMngeMmaNxY6RjpEGys+NDeFpbERk/P12bbvuadLsb9xdWD/Pnq2F3OBAeOzc9S/nkbbACaFj4FcQ/eg0etQWRcfOERSrfAJCOJzlpwa4vX/4bggYDnMf+AIFAQLxM7XY7GhoaNKIw+MQfAQDORV/D9u3bRaIRV+dHQHgLMO9gsZ/4xvfSbU+mMG3iHGywjdKUAZDPLX02z16MWCyGLVu2iJlYvV6Pxu5PMXbcOPT19WGU3oho23xYLJZ+s0TqPnsLDRYLIuPm5V2mVOjYDBPniv5js9mQSCQ0af/tdnvB7GiZx6+qKjBmDgCtBa8YqDSFnMglE9uW99IZA2cvzjlAikQi8Hq9ov2lxLMVS2NjIxKJBAKBAIwb3oHV44HPsADGjavQoNMhMm6e6Lt0nk2T58My50CEw2Fh4aT4LyJzWYqDIwGaTCaFhRZIixuyApZDbM0raInHgUmTCi4DpK9vJbXYaJKk0lpr0Q+Ww5ZMAuPmIZVK8SCZGXDke4KSIgG7rM0Mw+SHhRtTEoGl10AHwP2TR9HX11e1l37i8ZuQAOA89gcAdmXiowEK/W6cPB8JAGMveQidnZ0ILr0GAOD+6X/Etux2O/z0/U8eRXd3t2bwq3v6D4jq9bDNOxh90nI9PT15hRvtxzLnQE0SDLvdnrVOUNqm7+k/ILTpA5gmz4f7J49q3CvJlY4sSI7ldyNqMKDx549Br9ej909nIAag4Wf/3WWJfPoPiOh0cP/k0XS2x4euBQC0HX02ev78F0T1enHMhaxP1EbXTx4FkH5hRh6+HgDgOOYckWWPiipnvkxp/VFHn41t27aJF67u0T8gpNfDulO4BZdeI4QbAJgmz8ekix6Ez+cToqRHOreif/30P9i2bZuI+1EUBTqdDqPfeRCpdw1wqypiOh2af/FEURa05BO/R3TneauWxY2Oja6tz+eDXq9Hd3e36B+ULr+/7QA7+5Z0XWKxWFHWNopFLCTWyDJptVrhveMWhAA077xGBLn/ypMSJpMJTU1NVR/cWywWEX9pfOh2RPV6uH7yKHwPXw/9zj5MViH5PJtnL9a4n2a6SMrLWuYcqHE7puQ9tHwlbp/y/gDAscdh/S6jb99X465JbShUi42ylZJYy+XyS3GlVqu1qLig4NJr0vUMf/JoViIUhhkI5GdeprVtoCzfDDNUYeHGFE1szSuIf/gqAEBduxL2iXtUZbvJF+6BGk4PxkJP3gbHMedoCuGaX78f8Z2/0/6dG96BKxCAsvPv1KdvwDJ6djpb2qdviO/VtSthGT1bWAvUT99A4uPXAQDhp/6kWc42bl7OQVxszStiudiaVxDfacXT6XQ5rW20rLLsTrEv5cNXoa5dCevkhVAUBX6/XwxMTSYTGro+RvKTFUgCmnUAIPnJivSxffoG4h+9JtprXLdKnJfAXy9B8uPXkZSOOZ9wk9tIy+qX/x1R+Rrs+VWRtr2xsVE7OH7ij+J6hZ/6E0Yt/Ar6+vrQsOMT0XaaUaX9EMqHr8K56V3YRs9OJ4+R2hJ84o/i+qY+fQM6+3goigJFUdDU1AR318dIfbICSsb2LHMOzHmc8vEmpPOGfpYvBs113nltnbvtpXEvNBqNWeeu0HaCT/wx67rkE25yxkBFUTRiC9hpxZasL7L1Su7LdO5SqZRmogSAsLTWInEFWcBNn69GYme/le8Xde1KGKfunXWelQ9fhWn0bCFqZStg5rLyvRqJRGA0GoUArcax5TuXMqH/vZj1jIrFYqJERS5rqlxrTXZllim31lpszSuaZ4h16t5Fr8sw1SDzvkntvEf1en3eMjIMw+yCpzaYoqFZMiBteesvi1uxxB69IWsfJNp0Oh0Sj9+UtU5g6TXAU7dq/na5XLDb7VntlAdIyn9vzHs8+Vw0cm0PSA/+Mi0R8rLyZ1rXZDKJAtqE0+mE45V7NOvl2qfu6T9ovos+suu8RZbdrfnN7XbntZLk2nbsP7/T/N7a2iqER6aLXOYxjho1Ku1u9swf8x5D5nlwuVywWCwI/Os3edvV2toq0rM3NDTA/PztBY8lH5nbrQb5rm2paf/z9Re6hjQop8G8z+cTRdjlOEtgVxIKj8eDtrY2NDc3w263a9qQa3/xeBzd3d2i7ZTso1JrVH9YLBbNPZx5/EajMed5drlcaatcRkKgfMuSqKVz6XQ6q3Jshe51cpcN//s6TXvcbjeam5vR3NysEW10fb1eLzo7O9Hb26spMQGk3WYdDgdaWlowatSosopwZ57jwapVx4xc8r1Pi4ntZBiGhRtTJPIsGZC2fNHseCXI1hsAUMN+hJ68TcQs2d9aqvld3j/NHNPf6tqVMKx/O6ud6tqV6YQE0mwz7UteLvnJiqz95Dpude1KWK3Wgta2zO3L69JgyWQyobm5Gc6ODzXnkiwL8nrKsjuzj1fefiqp+S316RtZx5LveML/uDzrGkSe/jPGjh2b5eaX63pFnv6zxtqW6xhynYfI+y8XvB6mz1ejqakJTqcTtq3v5+xvZFnJR67jLbR8MWRuk7ab/GQFLBZL0Wn/C/UXuoaKosDn86Grq0uUP8gUa3q9vqBYy7c/Ond+v19Tx6ypqWlAZr6NG1dprmnm8UefuT3neVbXroTH49FMTOS7JuralbDZbMJy6Xa7swRfOeQ7l8lkEl6vF11dXUh+skJzfHRNZbGUeX3lTJdA+nrY7XY0Nzdj1KhR6QmMMpOJ5LoX8t2jDFML8r1PbTYbW9sYpkhYuDFFkcuyUQ3rRa7tBndavxobG0XcVTEEll6DkDTDLX+fb1+5luuvfYGl16ChoSFrcFyM9Ydm3T0eD1paWmCxWEq2GhVDvmuTazuytU5ezuPxFHWMwaXXlHSdqH3Jx2/ud5mxY8eitbUVqSd+n3e5QucmX3sroT9LYrFp/4vtj7nEmtVqFWKNapiVYt3LbDOQTmzR2to6YFaY6KO/Lfh7ofOc6QZcaNm2tjbodDp4PJ6qJT7Idy6TyeSuuDHJmiivRzGJXV1deS2nNpsNTU1NaGtrg8fjERMC1W5zpfcCw5RCvvvG7XZzLUGGKRIWbky/5JrNBiq3XmRabwg17Ef4qT8h8fxfcv6ej3wzyPEPX9XEDxVaXz6eQsedaf3Jt2yudVOfviGK6ha7XinngfaTeW3y7iuVndCCLJ8yha5XqTP3mRbTfMtg3ZvAujcLbj+f1S3f8fZnpStEoeslzx5Xsp3M7dGg3Wq1orGxEW1tbWhqaipKrPW3P9oHufANVIZBOV42H/n6fLH3KS2rrl0pLJHVoL9zabFYYN4Zu5eJsrPfx+PxrMySdH1Hjx6NxsZGUXOtlm2u5F5gmFIo5X3KMEx+WLgx/VKqRaNa263mbHCx2yoUt1Joe6W0tdh9VEolbazG+tWg2H5QqjWh3GPpb71y+lkhAkuv0Yg1m81WVhxIof1RfOdAznhX0+pZjOWymrXK+juXDQ0NgBTvmWsZsnLKMYl0fWtxHWr1DGeYYuE+yDDVgYUbU5D+LAPlztjms94QathfspWpEMVui46nlOMu1mpW6j4qpZI2AulzRvXa+rtetaJQrFzmcsVaYXItXwzFnMNitlvKtSDLZCVB+/3trxpxf9VsTzGUcg9V8/iKOZfRZ24vaE0ky1x/MYnVolbPcIYpFu6DDFM9WLgxBalF/FW56wwUpVp5yj3+gTgHlbSxGusPJKVaMiu1QJa7XK32W8n6A3l9q7WvUu6hau6zWssMlFtqvV1/ZuTBfZBhqgcLNyYvxc6MlzpbNljWm2IpxcpTTOxcJfuolEraCKStbj03fLOurxdRqiWzlH5bipWo0HbLsTbVKiavWvuoRXuKoZR7qBrHV8141Ho732zxYGoF90GGqS4s3Ji8lBu3Vc1l652hcCyVtlF564kqtaT2lGrJrJXFJt/ytYqtq8Z6A2kBHgwGwnI5mNurdB9D4VnGDD24DzJMdWHhxuSk3Lit/qh3a1upDIVjGQptrBalWjKrHZNWaLuVWJtqFZNX6T5q2Z5qMxCWy4FqTzHU2/VnRh7cBxmm+rBwY3JSq7g1nlFj6o1qx6TlW2+gLT71FntaD/f+QFs8B2u75W67Hq4RM3zgPsgw1YeFG5NFubPL/c2WDTdrGzM8qHZMWq7tVjOTYjHU6h4ul8G2thEDYbmsdXuKod6uPzPy4D7IMLWBhRuTRS1rszFMPVLtmLTM9Qcyq2Gl+6vFfVpP9/5AZ/UcjO3X2/VnRh7cBxmmNuhUVVUHuxFDkffffx+JRAILFiwY7KYwDDOC+fzzzzFp0qTBbgbDDBp8DwwufP4HH74Gg89AXQO2uDFDlkAgMNhNGPH09fUNdhMYZlDZvn37YDeBYRiGGSGwcGOGLL29vYPdhBGP388xi8zIRlGUwW4CwzAMM0Jg4cYwDMMwDMMwDFPnsHBjGIZhGIZhGIapc1i4MQzDMAzDMAzD1Dks3BiGYRiGYRiGYeocFm4MwzAMwzAMwzB1Dgs3hmEYhmEYhmGYOoeFG8MwDMMwDMMwTJ3Dwo1hGIZhGIZhGKbOYeHGMAzDMAzDMAxT57BwYxiGYRiGYRiGqXNYuDEMwzAMwzAMw9Q5LNwYhmEYhmEYhmHqHBZuDMMwDMMwDMMwdQ4LN4ZhGIZhGIZhmDqHhRvDMAzDMAzDMEydw8KNYRiGYRiGYRimzmHhxjAMwzAMwzAMU+ewcGMYhmEYhmEYhqlzWLgxDMMwDMMwDMPUOSzcGIZhGIZhGIZh6hwWbgzDMAzDMAzDMHUOCzeGYRiGYRiGYZg6h4UbwzAMwzAMwzBMncPCjWEYhmEYhmEYps5h4cYwDMMwDMMwDFPnsHBjGIZhGIZhGIapc1i4MQzDMAzDMAzD1Dks3BiGYRiGYRiGYeocFm4MwzAMwzAMwzB1Dgs3hmEYhmEYhmGYOoeFG8MwDMMwDMMwTJ3Dwo1hGIZhGIZhGKbOYeHGMAzDMAzDMAxT57BwYxiGYRiGYRiGqXNYuDEMwzAMwzAMw9Q5LNwYhmEYhmEYhmHqHBZuDMMwDMMwDMMwdQ4LN4ZhGIZhGIZhmDqHhRvDMAzDMAzDMEydw8KNYRiGYRiGYRimzmHhxjAMwzAMwzAMU+ewcGMYhmEYhmEYhqlzWLgxDMMwDMMwDMPUOSzcGIZhGIZhGIZh6hwWbgzDMAzDMAzDMHUOCzeGYRiGYRiGYZg6R6eqqjrYjRiKrFq1Cqqqwmw2D3ZTRiyJRAJGo3GwmzGi4Wsw+PA1GFz4/A8+fA0GFz7/gw9fg8GHroHZbMaMGTNqth++ymWi0+kGuwkjHn5IDT58DQYfvgaDC5//wYevweDC53/w4Wsw+AzUNWCLG8MwDMMwDMMwTJ3DMW4MwzAMwzAMwzB1Dgs3hmEYhmEYhmGYOoeFG8MwDMMwDMMwTJ3Dwo1hGIZhGIZhGKbOYeHGMAzDMAzDMAxT57BwYxiGYRiGYRiGqXNYuDEMwzAMwzAMw9Q5LNwYhmEYhmEYhmHqHBZuDMMwDMMwDMMwdQ4LN4ZhGIZhGIZhmDqHhRvDMAzDMAzDMEydw8KNYRiGYRiGYRimzjEOdgOGCqqq4vHHH8e///1vfPTRRwiHw2htbcXee++NU045BfPnzx/sJg5r3nzzTZx22mlFLbt48WL85S9/qXGLRga//OUvcd999+Hcc8/Feeed1+/yL7/8Mu6//36899578Pv9aG5uxvz58/HNb34T+++//wC0ePhR7DXYunUrDjvssKK2OWXKFDz99NPVauKwwufz4YEHHsCLL76IDRs2IBQKweVyYcaMGTjqqKPwla98BWazOe/6/K6onEquAb8rKqe7uxv/+Mc/8PLLL+Pzzz8HAIwePRqLFy/G17/+dUybNq3g+nwPVEYl55/7f+3w+/049thj0dXVhRNOOAHXXntt3mVreQ+wcCuCaDSK888/Hy+99JLm+61bt2Lr1q147LHHcMEFF+B73/ve4DRwBPDJJ58MdhNGHCtWrMADDzxQ1LKpVApXXHEFli5dqvm+o6MDHR0dePbZZ3HqqafiZz/7GXQ6XS2aOywp5RrwPVI5K1aswIUXXoje3l7N9729vVixYgVWrFiBf/7zn/jTn/6ECRMmZK3P74rKqfQa8H1QGa+99houvPBCeL1ezffr16/H+vXrcf/99+Pcc8/FWWedlXN9vgcqo9Lzz/2/dvzmN79BV1dXv8vV+h5g4VYEP/3pT8UFmDp1Kr72ta+hpaUFa9aswQMPPIBwOIzf/e53aGtrw/HHHz+4jR2m0MPI4XDg+uuvL7hsc3PzQDRpWPPBBx/g3HPPRSqVKmr53//+90K0jRkzBieffDLGjRuH9evX47777kNvby/++c9/orm5Geecc04tmz5sKPUayC/sG2+8ERaLJe+yDoej4vYNNz7++GOcffbZiEQiANKz0Ycffjg8Hg+2bduGRx99FGvXrsXatWvxne98Bw899BDcbrdmG/yuqIxqXAN+V5QPnf9YLAYAOOSQQ7B48WI4nU589NFHePDBBxGNRnHTTTfB6XTi1FNPzdoG3wPlU43zz/2/Nrz00kt45JFHilq25veAyhTk1VdfVdvb29X29nb19NNPV6PRqOb3devWqYsWLVLb29vVRYsWqYFAYJBaOrw58cQT1fb2dvXrX//6YDdl2PPSSy+pe+21l+j37e3t6i233JJ3+XXr1qmzZs1S29vb1WOPPVb1er2a3zs7O9UvfOELant7uzpnzhx1y5YttT6EIU+p10BVVfX8889X29vb1QMOOGCAWjm8OOWUU8S5vu+++7J+j8fj6kUXXSSWueaaazS/87uiciq9BqrK74pKOPnkkwue/w0bNojn0sKFC7P6MN8DlVHp+VdV7v+1wOfzqYsXL9a8jy+99NKcyw7EPcDJSfrh7rvvBgAYjUb8+te/zprFnjp1Kn7+858DALxeb5arGFM5qVQK69atAwBMnz59kFszfFEUBbfccgvOOuss+P3+otf729/+hmQyCQC48sor0dDQoPl91KhRuO666wAA8Xgc99xzT/UaPcwo9xoAu2Za29vba9G0Yc1nn32Gt956CwBwxBFH4Jvf/GbWMkajEVdffTVGjRoFAHj44YdFvwf4XVEp1bgG/K4on3Xr1uHtt98GkLZ05jr/kydPxv/93/8BAILBIJYvX675ne+B8qnG+ef+XxvIRTLTup+LgbgHWLgVwOv14vXXXwcAHHjggTn96QHgmGOOESZnDvivPps2bUI4HAbAg9Ja8frrr+Poo4/GH//4R6RSKdjtdpxxxhn9rpdKpfDMM88ASF+bvfbaK+dyCxYswJw5cwAAzzzzDFRVrV7jhwnlXgMAiMViIoid75HSWbFihfhcyHXFYrHg0EMPBZBOoLFx40YA/K6oBpVeA4DfFZXg9Xqx7777oqWlBUcddVTe5WbOnCk+b926VbM+3wPlU+n5B7j/1wLZRfKSSy4puOxA3QMs3Arw9ttvi/iSfffdN+9yer0ee++9NwDgf//7H3w+34C0b6Qgx+7ww6g2/Pe//8WWLVsAAHPnzsXSpUvF4KgQn376qQiiLnSPyL93dXVxAHUOyr0GALB27VpheeB7pHT0ej2mT58Op9OJyZMnF1xWtiiTVZTfFZVT6TUA+F1RCXvttRf+9re/4bXXXsNJJ52Ud7mOjg7xubW1VXzme6AyKj3/APf/auP3+4V17Otf/zr222+/gssP1D3AyUkKsHbtWvG5v5uA0rOqqopPP/1UXBSmcj799FPxmcz/mzdvxueffw6DwYBJkyZh7Nixg9W8YUNTUxPOPfdcfOMb34DBYEBPT0+/68jXpth7BEgHYcszh0yacq4BkPse6ezsxGeffQZVVTF+/HhMmjSpJm0eDpx88sk4+eSTi1qWXJEAwOPxAOB3RTWo9BoA/K6oNV6vV7iC2Ww2HHTQQeI3vgdqT6HzD3D/rzbkIjlmzBhccsklWZk+Mxmoe4CFWwFkM/S4ceMKLjt69GjNevwgqh40i9TS0oLly5fjjjvuwGeffaZZZs6cObjggguyHmRMcZxyyim48sorYbVaS1qvlHtkzJgxOddj0pR7DYBd94her8fmzZtx1VVX4b333tMsM3nyZPzgBz/Al770paq0dyTS2dmJV155BQDQ2NgoxDC/KwaOfNcA4HdFLYjFYti6dSuef/55/P3vfxfp0C+77DI0NTWJ5fgeqA3Fnn+A+381kV0kf/nLX8LpdPYr3AbqHmDhVgC5jkxjY2PBZeVZv/4uLlMa9DDq7u7GpZdemnOZNWvW4Lvf/S7OO+88nHvuuQPZvGHBvHnzylqvlHtEdm/ieySbcq8BsOseSaVS+NGPfpRzmY0bN+LHP/4xVq5ciV/96lfQ69lTvlSuu+46xONxAMCxxx4rziG/KwaOfNcA4HdFtfnggw/w1a9+VfNda2srfvKTn+CYY47RfM/3QPUp5fwD3P+rhewi+eUvfxkHH3xwUesN1D3Ab+4CRKNR8blQTSQAMJvNOddjKiMcDmPz5s3i70mTJuH666/Hq6++ivfeew///e9/ceqpp4qizrfeeisefvjhwWruiEPu6/I9kAv5HuJ7pLrILjKtra244oor8PLLL+P999/H008/jXPOOUdcn4ceegh//OMfB6upQ5YHHngATzzxBADAbrdriqfyu2JgKHQN+F1RfeR4KsLr9eKpp57SJIUB+B6oBaWcf+7/1YNcJEkkF8tA3QMs3AqQSCTE5/4GpfLv8npMZWzatEncAAsWLMDDDz+M448/Hq2trbBYLJgxYwZ+/vOf45prrhHrXHfddQiFQoPV5BEF3yODj9frFQ/+SZMm4ZFHHsEpp5yC0aNHw2w2Y8qUKTj//PNxxx13wGhMO1ncfvvt2LZt22A2e0jx3HPP4aqrrhJ/X3nllWhraxN/831Qe/q7BvyuqD6NjY34+c9/jptuugmXXnopZs6ciXg8jmeffRYnnngi1qxZI5ble6D6lHL+uf9XB9lFMld5o0IM1D3Awq0AcqwJuWbkQ1EU8bm/C8YUz8yZM/Huu+9i+fLluPPOO+F0OnMud8IJJ+Cwww4DkB7ILlu2bCCbOWIp9x4xmUw1a9NIw+PxYNWqVXj99dfxwAMPZGUaI/bbbz9RGygej/Nsa5E8++yzuOCCC0TWztNPPz0rXT2/K2pLMdeA3xXVZ88998Spp56KY445BmeeeSYeeeQRfP3rXwcABAIBXHzxxeKa8D1QfUo5/9z/K0d2kTzmmGNwxBFHlLT+QN0DLNwKYLfbxedYLFZwWfki9GciZUpDp9Ohra0NLper4HLyi5wKWTK1Rb5H5HsgF/I9VE4CDqYwzc3NWcHqmcj3yDvvvFPrJg15li5dih/96EfiJXzCCSfkdJ3hd0XtKPYaAPyuqDV6vR6/+MUvMGPGDADA+vXr8eqrrwLge2AgKHT+Ae7/lUIukmTpLJWBugdYuBVArpLeX/Cg/Ht/gyemNuy2227iM2VeYmpLKfeIXKuE75HBge+R4vnDH/6An/3sZ8KN5atf/Sp+85vfiBgRGX5X1IZSrkEp8H1QPgaDASeeeKL4e9WqVQD4Hhgo8p3/UuD+n43sIvnzn/+8rH45UPcAC7cCyEVAt2/fXnBZOYiU62QMDhS/wwwcpdwj8u98jwwO7KLaP6lUCldccQVuvfVW8d3pp5+Oq6++Om8mTn5XVJdyrkEp8LuiMqZMmSI+UyY9vgcGjlznvxS4/2fz9NNPi88XXnghZsyYkfXv8MMPF8s88sgj4nt6Tg3UPcBXrwBTp04Vn9euXYtFixblXZYK7+l0OlH4kKmc5cuX4+OPP0ZPTw/OOuusgilWOzs7xed8cT5MdZGLasvFJ3NRSnFKpnjeeecdrF69Gj09Pfja176mealnwvdIYVKpFC655BI89thj4rvzzz8f55xzTsH1+F1RPcq9BvyuqIxHHnkEL774IjZv3oy//e1vGutBJrIbmMPhAMD3QKVUev65/w8+A3UPsHArwIIFC2AymRCPx7Fy5UqccsopOZdLJpN46623AKQDRAvdcExpPPvss1i6dCkAYK+99sIXvvCFvMuuWLFCfN5jjz1q3TQG6Zm/lpYWdHd3Y+XKlQWXpd89Hg8Ltyry5ptv4uabbwYAtLW1FRRur7/+uvi8++6717ppQ44rr7xSCAa9Xo8rrrhCJHQpBL8rqke514DfFZXx6aef4plnngGQPj9LlizJu+x7770nPpPbHd8DlVHp+ef+XxmnnXZav8lIenp6cMUVVwAA9tlnH5x++ukAdllAB+oeYFfJArjdbuy7774AgBdeeCFv+uwnnnhCmKuPPvroAWvfSIDOP5Cu4ZOP3t5ePPjggwAAm82GI488suZtY9IDK3rYffDBB3j33XdzLvfOO++I1MVLlizh4s9VRL5HHnroIZFlLJNYLIa7774bQHqW70tf+tKAtG+o8PDDD4tniF6vxzXXXFOUYAD4XVEtKrkG/K6ojAMPPFB8vu+++/Iu5/V6hUAwmUyiODHfA5VR6fnn/l8Zc+bMwRFHHFHw3wEHHCCWHzt2rPieLG0DdQ/w6Kkfvv3tbwNIp/a88MILEQwGNb+vW7cOV199NYC0yfqkk04a6CYOa4444giMHj0aAPDqq6+KgaeM3+/HueeeK4I9zzjjjJJqbzCVcdpppwmf+UsvvRQ7duzQ/N7Z2YnLLrsMQDqw+lvf+taAt3E4s2DBAsyZMwdA2v3immuugaqqmmUURcGPf/xjUbT1S1/6ksatY6TT3d0tnuNAuh9/+ctfLmkb/K6ojEqvAb8rKmPffffFzJkzAQBvvPEG/vKXv2QtEwwGcd5554lB59e+9jVNLT2+B8qn0vPP/b8+GIh7QKdmvuGZLH74wx8KE/b48ePxzW9+E2PGjMGHH36I+++/XxQw/OUvf4lvfOMbg9nUYcnLL7+Ms88+W1gSDjjgACxZsgQulwtr167F0qVLhVhYtGgR/vKXv3BtmCqwcuVK4Qpw7rnn4rzzzsu77HXXXSdeFM3NzTj55JMxZcoUbNiwAffddx96enoAAN/73vdw0UUX1b7xw4Rir8GaNWtwyimnIBKJAADmz5+P448/Hk1NTdi0aRP+/e9/Y9OmTQDScYn33nsvPB7PgBzDUOB3v/sd7rjjDgDpmdSf/vSnRa03e/ZsTWA5vyvKpxrXgN8VlZH5HFm8eDGOPPLInOdv9uzZuO+++2Cz2TTb4HugfCo9/9z/a8uWLVtEgpITTjgB1157bc7lan0PsHArgkgkgnPPPVdTL0NGp9Ph3HPPxbnnnjvALRs5PPXUU/jJT36CcDicd5nDDjsMv/3tb0WwLlMZpQi3ZDKJn/70pyKdbi5OOukkXHXVVewmWQKlXIM333wTF1xwgRDJuVi4cCF+//vfY9SoUVVv61Bm8eLFWZbiYrjmmmvwla98RfzN74ryqdY14HdFZRTzHDnwwANx44035ozN4XugMio9/9z/a0exwq3W9wAnJykCm82Gu+66C4899hgeffRRfPTRRwgEAvB4PNhzzz1x+umnY8899xzsZg5rjj76aOy555649957sXz5cnz++eeIx+NoaWnB7rvvjhNOOEH4ejMDj8FgwLXXXoujjz4aDzzwAN5//314vV44nU7svvvu+OY3v4lDDjlksJs5rFm0aBGefvppPPDAA3jhhRfw2WefIRKJoKmpCbNnz8Zxxx2HY489tuIaWMON3t7esgRDLvhdUR7VvAb8rqiMRYsW4amnnsK9996LF154ARs2bEAsFkNzczMWLFjQ7/nje6AyKj3/3P8Hn1rfA2xxYxiGYRiGYRiGqXPYZ4lhGIZhGIZhGKbOYeHGMAzDMAzDMAxT57BwYxiGYRiGYRiGqXNYuDEMwzAMwzAMw9Q5LNwYhmEYhmEYhmHqHBZuDMMwDMMwDMMwdQ4LN4ZhGIZhGIZhmDqHhRvDMAzDMAzDMEydw8KNYRiGYRiGYRimzmHhxjAMwzAMwzAMU+ewcGMYhmEEyWRysJvAMAzDMEwOWLgxDDNozJgxQ/ybOXMmOjs7i173nHPO0aw/0Dz88MM5971lyxbx/cMPPzzg7aqEJ554Aj/+8Y+zvr/11lsH7Twzw4Ph0odCoRCOOOIIHH744VAUZbCbUxannHIK9t57b2zfvn2wm8IwTImwcGMYpi5QVRVPPfVUUcsGAgEsX768xi0aWdx000248MILsWPHjsFuCsPULb/5zW+wefNm/PjHP4bZbB7s5pTF5ZdfjkAggMsuuwyqqg52cxiGKQEWbgzD1A1PP/10Ucs9++yziMfjNW5NeZhMJkycOBETJ06E0+kc7OYUTSFrZ0NDgzgmhhmpvP3223jooYewYMECHHXUUYPdnLKZO3cujj32WLzxxht49NFHB7s5DMOUAAs3hmEGnfb2dgDAu+++i46Ojn6Xf/LJJwEAra2tNW1XObS1tWHZsmVYtmwZjjzyyMFuTlU4/fTTxTExzEhEVVX86le/AgCcf/75g9yayjn33HOh1+txww03IBQKDXZzGIYpEhZuDMMMOgcffDAcDgdUVe3X6tbb24s33ngDRqMRX/jCFwaohQzDjGSefvppfPzxx2hvb8d+++032M2pmClTpuCggw5CT08P7r333sFuDsMwRcLCjWGYQcdiseCwww4D0L+75FNPPYVEIoF9990XTU1N/W573bp1+PnPf44jjjgC8+fPx1577YUTTzwRd955JyKRSMF1t27dil//+tc4+uijsfvuu2Px4sW4/PLLsXnz5rzrFJOcZPny5bjkkktw5JFHYuHChZg7dy72339/nHnmmVi6dGlON1BKhkLn6b333sMFF1yAxYsXY+7cuTj44INx+eWX47PPPuv3nMhQ0ohHHnkEAPDmm2+K9m/ZskWzTGZiiZUrV2q+/+ijj3DhhRdi8eLFmD9/Po488kjceOONCIfDANIZK//xj3/ghBNOwB577IE999wT3/rWt7BixYqCbVy9ejUuvvhiHHLIIZg3bx4WLVqEU089Fffff3/ZLrPU7uXLl2PHjh244oorcNBBB2H+/PlYsmQJrrjiCmzatKngNuLxOO6//36cdtpp2GeffcR1uOiii/Duu+/mXEc+Z4qi4MYbb8T+++8v9vvYY48V3Oc3vvENzJgxA9/61rdy/v7444+L7T/wwAM5l6HEPuedd17Wby+99BJ+8IMfiH61zz774LTTTit4rg877DDMmDEDS5cuxauvvoovfelLmDt3Lg444ABcdNFFBY8HSPeL8847T7T7r3/9a9Yyg3GuZe666y4AwFe/+tW8y7zyyiv44Q9/iAMPPBBz587F3nvvja985Su48cYb+40draSP9/b24vbbb8dXvvIVLFq0CPPmzcOSJUvw61//umACkhNPPBEA8Le//a1uXc8ZhtFiHOwGMAzDAMCxxx6Lxx57TLhLjh49Oudy5CZ53HHHCWGRj3vuuQc33HCDJsV9LBbD+++/j/fffx/3338/7rjjDkybNi1r3eXLl+P8888XogMAotEoHn74YTzzzDP44he/WPIxRiIR/OhHP8KLL76Y9VtPTw9ee+01vPbaa3j88cdx9913w2Aw5NzOAw88gKuuukpzXB0dHXj44Yfx+OOP44477hhwq8CTTz6JSy65RDMA/Pzzz3H77bfjzTffxN13342zzz4bb7zxhma9N954A2+++SZuvfVWHHHEEZrfUqkUrr/+etxzzz2a7xVFwVtvvYW33noL//rXv/DnP/8ZbW1tZbV769at+OlPf4quri7x3caNG7Fx40b85z//wc0334xDDz00a73t27fje9/7Hj799FPN9x0dHXj88cfx+OOP47vf/S4uuugi6HS6nPu++uqrNeJq48aNGD9+fMH2HnzwwVi9ejVWr16NWCwGi8Wi+V0WwW+99Ra+8Y1vaH5XFEUsIx9XJBLBRRddhOeff16zvNfrxZtvvok333wT9913H/785z9j3LhxOdv27rvv4pe//KXoA93d3XA4HAWPR1VV/OQnP8Gzzz4LALj00kvx7W9/W7PMYJ1r4oMPPsAHH3wAAFiyZEnOZW666Sb8+c9/1nwXj8exZs0arFmzBvfeey/uuusuLFiwQLNMpX38zTffxAUXXICenh7N99SHH374Ydx2223Yd999s9Y9+OCDYbPZ0N3djWXLluGYY47p/2QwDDOosMWNYZi64IADDoDb7S7oLtnZ2Yl33nkHZrO5XzfJpUuX4tprr0UymcSiRYtw9913Y8WKFXjppZfw61//Gq2trdi6dSu+853voLe3V7Puhg0bcM455yAcDmPs2LG46aab8Prrr+O5557DBRdcgHg8nteaUYgbbrhBiLZTTz0VjzzyCFasWIFly5bh5ptvFgLyjTfeyGsN2LFjB6666ipMmTIFt956K15//XU8//zzOO+882AwGKAoCq644oqis8V9//vfx6pVq4QQ3XPPPbFq1SqsWrUq7wA9F5dddhna2trw+9//Hq+++ioeffRRLFq0CEDamvCVr3wFb775Jv7v//4PTz31FFasWIEbb7wRDQ0NSKVSuOGGG7K2ecstt4gB7ZFHHon77rsPK1euxHPPPYfLLrsMTqcTH374Ib7//e8jFosV3VaZ66+/Hl1dXTj55JPx5JNP4vXXX8fvfvc7tLa2IhqN4rzzzsP69es164TDYZx55pn49NNPYbfbcdFFF+GZZ57BypUr8a9//QvHHnssAODOO+/EnXfemXffDzzwAJYsWYJnnnkGL7zwAq688sqsgX0mhxxyCID0BMQ777yT9bss3N58882s39955x2Ew2Ho9XqxLQC48MILhWg7+uij8eCDD2LlypV48skn8b3vfQ9GoxGffvopzjzzTASDwZxte+ihh9Dc3Iy7774br7/+Om677TacdtppBY/nqquuEgkyLr74Ypx55pma3wfzXBM0WTRt2jSMGTMm6/fVq1cL0fbFL34RDz30EFasWIHnnnsOv/jFL2C32xEMBnH55ZcjlUpp1q2kj2/evBnf/e530dPTg+bmZvzyl7/Eiy++iOXLl+O6665DS0sLQqEQzj///CxhBwBmsxn77LMPgOITQzEMM8ioDMMwg0R7e7va3t6u3nLLLaqqqupll12mtre3q1//+tdzLn/33Xer7e3t6g9+8ANVVVX1lltuEduQ8fv96oIFC9T29nb1nHPOUZPJZNa2Nm/erC5cuFBtb29Xf/WrX2l+O+uss9T29nZ10aJF6vbt27PWXbZsmdhv5r43b94svv/3v/+tadOcOXPU9vZ29ac//WnO4+vq6lLnz5+vtre3qxdeeKHmt3//+99iu4cffrgaCASy1r/++uvFMu+//37OfeTj0ksvVdvb29VTTz0167d85/mNN94Q3y9YsEDt6OjQ/N7R0aHOnj1bLHP77bdnbfvee+8Vv2/dulV8v2HDBnXmzJk5rw/xv//9T501a5ba3t6u3n333SUdr3z9br311qzfP//8c9E/zj77bM1vN998s9re3q7OmTNHfffdd3Nu/1e/+pVYpqurS3wvn7NDDz1UjcfjJbVbVVX1wAMPVNvb29Xf/va3mu83bdqktre3q3vuuac4dxs2bNAsc+2112bdYy+88IJo07XXXptzn08//bRY5rrrrtP8duihh4rfXnvttZzr5+pDN9xwg/juz3/+c871Bvtcq6qqHnfccWp7e7v6k5/8JOfvdE6/8IUvqKlUKuv3hx9+WLTjgw8+EN9X2sfpObVw4UJ148aNWeu+9957Yvt/+MMfcm7/j3/8o9hGruckwzD1BVvcGIapG8hV5913380Zm/HEE08AgJhlz8d///tfkSntsssug16f/agbP348Tj31VADp+LFEIgFAWyPujDPOyOmyecQRR2Dx4sXFHpbY7re//W0cffTROOOMM3Iu09raiilTpgBAlhVQ5tRTT81ZakB2fevPjbTafOlLX8py5Wpra8OECRMApOMYTz/99Kz1ZKuHXJLgwQcfRCqVgs1mw49+9KOc+5w/f77oM//617/KavfEiRNx9tln5/yeXPaWL1+OQCAAIO3a9+CDDwJI98Pdd98953YvuOACWK1WxONxET+YyRe+8AUYjaVHLBx88MEAgNdff13zPVnb9tlnH2G9zbS6vfLKKwCgsbbRuWttbcWFF16Yc59LliwRrqz/+te/NG66hMfjKdpF909/+pOwkJ1//vn4/ve/n7VMPZzrYDCIdevWAdiV/TYTcg0Nh8M5MzQeddRRuO222/DYY49h+vTp4vtK+rjf7xfX8owzzsCkSZOy1p03bx6OPvpo7LnnnrBarTm3T8cUDAZLjo9lGGbgYeHGMEzdsN9++8Hj8UBVVTzzzDOa3zZt2oT3338fdrs9Z8yRzMqVKwEAjY2NaGpqQigUyvlv/vz5AIBQKISPP/4YQDouiETcQQcdlHcfmfFY/TF27FhcfPHFuPnmmzF16tSs3xVFwdtvvy0SplAbckHtzkRO1hKNRktqX6XkG1Q3NzcDAHbbbbecg0dZgCqKIj6T4Nhtt90AIO81pP2uX78efX19Jbd7yZIleWMJqZ/F43HRnnXr1gm3s1mzZuVtl06nE0lbVq1alXP7s2bNKrm9wC7R9eGHH8Ln84nvSbgtWrRI9BFZuG3fvh1r167VHBuQ7vNAOsmIyWTKu18SEIFAQNwvMjNnzswbYybzj3/8AzfffDOAtCA755xzci5XD+d648aNwr2R+mIme+21F4C0G/NXv/pV3H333RoRZLPZcPjhh6O9vV1TtLuSPv7WW28JwSiL8ExuvPFG3HffffjOd76T83eaKALSLuIMw9Q3nJyEYZi6wWg0YsmSJXjwwQfx1FNPaZIUkLXtiCOOyDt7TJC1qa+vDwsXLixq3x0dHZg7d67G0leo4HQu8VUsn332Gd5++21s2LABmzZtwueff46NGzcWFGsy+bJpyoPCzFiaWtPY2Jjze7J2ulyunL/nG+jTNVyzZk1J1zBfO/KRz4oCAJMnT9ZsG4Amo+g111yDa665pt995MvsV0xW1Fzst99+MJvNUBQFK1euxJFHHglVVUXil3322QculwsPPfSQEGUAhCV53LhxQugEg0FhTeyvT8u/b9++HXPmzCnreK6++mrx+ZVXXkFXVxdGjRqVtVw9nGu5rmRDQ0POZY488kgcccQReO6557Bx40Zcd911uO666zB27FgsXrwYhx56KBYvXqy5P4HK+rhsnZb7aal4PB7NthmGqW/Y4sYwTF1x9NFHAwD+97//aQZhcjbJ/siXPKGYdWgQC6RnyvORT4gUYuvWrfjWt76FY445BldccQXuuecePP/881i3bh2cTieWLFkiXAsLUY7LV60pdK7KoZJrWAputzvvb/IEAW27mu3KzAhZLHa7XSR+ISvbJ598gt7eXng8HsyYMUMknejo6BBlDUi4ydY22bXPbrcX3K98jXO5BBZ7PKqq4rjjjoPRaITf78dVV12Vc7l6ONdyVtlc7slAenLi1ltvxa9//WuNmN22bRv+9a9/4eyzz8bBBx8skrD019ZC0DqypbWSe09+jsnHyjBMfVJ/b3+GYUY0++yzD1pbW7Fjxw48/fTTOOOMM7Bu3Tp8+umn8Hg82H///fvdBg24d99995Jjn+RZ9UgkknewJrv1FYPf78dpp52GrVu3Qq/X46CDDsJee+2F6dOnY+rUqUKwffOb3yxYJ26kYLVaEQwGccwxx+Cmm26q2X4KZaOUB7JkyZMHyXfeeWdBd9pacsghh+DVV18VcW6ytU2n02HcuHEYP348tmzZgpUrV2LMmDE5ywDIYq2/gXspIq8QZ555Ji699FKMGjUKd999N5YtW4Znn30WRx55pGa5ejjXskU4n0stkBZvJ510Ek466SR0dHTglVdewYoVK/Daa6/B6/Wit7cXl156Kdxut6jFWEkfl89NoedUf8ixisW4uTIMM7iwxY1hmLpCr9eLWkmUopqsbUuWLCkYg0OMHTsWQNrCVQg1R8p8WhdAVhp4mVLF1X333Sfa8/vf/x633347vvvd7+KQQw7RWNnKidMajlRyDUuh0HWUY36oNIKcDr7WbSsExTVt3LgR27dvF3GdZGkDIGp3vfnmm3jnnXfw/+3de0hTbxgH8O/KTacUS0uNLopgpVISlkYWZP5RVIhkilgZdFfTGS2ESqMbZKkVWlLCkhSslNQRdkGEwAiyqCHrIhSmtSxX2kUrhfz9Ief8Nj3q1NITfD9/RedsO54d5X3e532fp6OjwyZbB/RmkYSs41DFKYQiHYDt78lwpaWlAQCSk5PF9zl69Ci+fv1qc54c7rV1Hzph/+lQPD09ER0djZycHLG9hLBMsqioSDxvNM+49b0Z7Bl+8OAB8vLyUFlZKXnc+u/NUD33iGj8MXAjItmxXi7Z0tKCW7duAbBvmSTwf7EAi8UCo9E44HkXL17EokWLEBERIS4nCw4OFmezq6urB3ytsOzMXk+ePAHQm7npm1kQtLS0oLGxEcDY71GT22y78B2aTKZB995kZGQgJCQEUVFRI1p6Ntj3KPQ1c3FxweLFiwH0FuAQsht9m1Vb6+joQGhoKMLCwpCVlTXs6xrKrFmzxD1n9+/fF4tyWAdlQhBXV1cn/px991opFAoEBQUBAGpqamwaqPclTKS4uLgMujfQXs7Ozjh06BCA3sIep06dsjkuh3ttHSBZ7yuzptVqER4eLtmLcOLEiVi3bp1Yhdb6PUbzjC9cuFD8nRWqS0opKSlBbm4u8vPzJY9bB25SPeqISF4YuBGR7AQFBcHT0xM9PT24cOECXr9+DQ8PD3GgM5TIyEhxcHrs2DHJmfKmpiZcvnwZ3759Q1dXl5j1UqvVYuBYVFQkVuGz9vTpU7FYir2EZVZfvnxBa2trv+O/fv3CwYMHxdn1wQbQf4NwfWP9uQOJiYkB0Ftd88iRI5Ll541GI8rLy9He3g6NRjOi5WJCs+O+Xr16JWZH1q5dKz5PDg4OWL9+PYDeAbMwqdDXmTNn8OnTJ5jNZsybN2/Y12UPoS1AcXEx2tvb4ebmZlNuXsi4vX//XiyTL1WRVbjXra2tyMnJkfys6upqMXiKjIy0K/Ntj/DwcLFCa1lZmZg5BORxr729vcUCO8LkTl8/f/7E27dvYTAYJNt4dHV1idlM64JHo3nG3d3dxWBQr9dLBpX19fWoqakBMHALFbPZLP57oKqZRCQfDNyISHYUCgVWr14N4P/eRWvWrJHsxyZl6tSp0Gq1AHoHLzExMbh79y4sFgvMZjPKy8uxefNmtLe3Q6FQ4ODBgzYZp/3790Oj0aCzsxObNm3C9evX8fHjR3z48AHFxcXYtm3bsJdmCYOs379/Y9euXXjw4AE+ffqE5uZmVFZWYsOGDaitrRXPlyr+8DcJ1eVevnwJk8mEtra2cQ3i/Pz8EBcXB6A3ExQfH4/a2lp8/vwZTU1NKCoqwo4dO9Dd3Q1HR0fs379/xJ+1d+9eXLp0CWazGRaLBTdu3MDmzZvR2dkJjUbTr8dWYmKimJ3Yt28fMjMz0dDQgLa2NtTX1yMtLU0M+oKCgsQy+n+asFzy+fPnAGyzbUDv4N66L+CECRMkS8evXLlS3Hel1+uxd+9eGI1GtLe34/Xr1zh79ixSU1MB9Gb6Bur1NlKHDh2Cs7Mzenp6kJ6ebtPKYrzvtVqtFlsJPHv2TPIcodT+x48fsWXLFty9exfv3r2DxWLBw4cPsXv3brx58wZA7x5WwWif8bS0NDg5OaGtrQ2xsbEwGAxobW3F27dvUVpaip07d6K7uxseHh42FXqtCZlaNzc3m9YARCRPLE5CRLK0Zs0aFBYWigHSUE23+9q2bRs6OjqQn5+PhoYGJCcn9ztHqVTi8OHDWL58uc3/u7q6orCwEDt27EBrayvS09Ntjjs6OiI1NRXZ2dl2X8+GDRtQVVWFuro6mEwmyYGUj48P/P39cfPmTbx79w7d3d1/LLMxlJCQEBQUFKCzs1PMcly5csVmz9RYO3DgALq6ulBWVoZHjx5J9qJycXFBTk7OiDMtYWFhePToEbKzs/t9n56enrh48WK/UvJTpkyBXq9HQkICGhsbodfrodfr+733ggULkJeXZ/eEw3AFBQVh0qRJYiVUqe9qyZIl4l69wMDAAcviZ2VlQafToaamBlVVVeK+UmsBAQE4d+7ciAthDGT69OlISUnByZMn8ebNG+Tm5opBihzudWhoKEwmk01rBWvBwcHYt28fcnJyBvxbM2HCBGi1WjFLKhjNM+7r64sLFy4gJSUFZrNZcvLCw8MDBQUFA1bBFQK30NBQ6R+eiGSFGTcikqXAwECxIISXlxfmz58/rNcrFApotVpUVFQgOjoaXl5ecHJygkqlgre3N2JjY1FZWYno6GjJ1/v5+cFgMCApKQlz5syBWq2Gq6srVq1ahdLS0n7ZjaGoVCro9XrodDr4+/tDrVbDwcEBrq6uCA4ORkZGBsrLy8UZ+R8/fthk4P625cuXIyMjA97e3lAqlXB1dYXFYhmzz5eiVCpx4sQJFBcXY926dZgxYwZUKhWcnJzg6+uLrVu3oqqqatAGxEMJCAhARUUFIiMj4ebmBrVajblz50Kr1cJgMAwYEPr4+MBgMCA9PR3BwcHQaDRwcHCARqNBSEgIjh8/jqtXr464f5g9HBwcbAbcUs+kdTA32H1ycXFBfn4+zp8/j/DwcEybNg1KpRLu7u5YunQpMjMzcfXqVbvaVYxEfHy8mNkqLCy0yW6N970WJo2am5vFPah97dy5EyUlJYiIiMDMmTOhUqng6OiIWbNmISoqCqWlpdi9e3e/1432GQ8NDcWdO3ewfft2+Pr6wtnZGY6OjvD19UVCQgIMBoPYs6+v79+/w2QyAbB//zARjS9Fz98se0VERCRDwmB2z549khkSImtxcXF4/PgxUlJSkJSUNN6X80dcu3YNGRkZmD17Nm7fvj1ouwMikgdm3IiIiIgGkZiYCACoqKgY84qvf0tpaSmA3mwhgzaifwMDNyIiIqJBLFu2DAsWLEBTUxPu3bs33pczai9evEB9fT2mT5+OyMjI8b4cIrITAzciIiKiIeh0OigUigF7ov1LcnNzAQCpqaljVgCJiEaPgRsRERHREEJCQrBx40YYjUbJ3n//isePH6O6uhorVqxgto3oH8PAjYiIiMgOOp0OXl5eyMzMRFdX13hfzoicPn0akydPxtGjR8f7UohomFhVkoiIiIiISOaYcSMiIiIiIpI5Bm5EREREREQyx8CNiIiIiIhI5hi4ERERERERyRwDNyIiIiIiIplj4EZERERERCRzDNyIiIiIiIhkjoEbERERERGRzP0H1K+jhwVSXuEAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Define pre-settings\n",
    "figwidth = 10\n",
    "figheight = 6\n",
    "xlab_size = 20\n",
    "ylab_size = 20\n",
    "figfont_size = 24\n",
    "\n",
    "# Define consistent color maps\n",
    "my_cmap_sns_light = [(0.9921568627450981, 0.8156862745098039, 0.6352941176470588), (0.7764705882352941, 0.8588235294117647, 0.9372549019607843)]        # light blue, light orange\n",
    "my_cmap_sns_dark = [(0.9019607843137255, 0.3333333333333333, 0.050980392156862744), (0.19215686274509805, 0.5098039215686274, 0.7411764705882353)]      # dark blue, dark orange\n",
    "my_cmap_dark = sns.color_palette(my_cmap_sns_dark, as_cmap=True)\n",
    "my_cmap_light = sns.color_palette(my_cmap_sns_light, as_cmap=True)\n",
    "\n",
    "\n",
    "# Create empty figure and plot the individual datapoints\n",
    "fig, ax = plt.subplots(figsize=(figwidth,figheight))\n",
    "\n",
    "\n",
    "# 1. Violinplots\n",
    "axsns = sns.violinplot(x='median', y='mode', data=dfq1a,\n",
    "                       hue=True, hue_order=[False, True], split=True,   # half violinplots https://stackoverflow.com/questions/53872439/half-not-split-violin-plots-in-seaborn\n",
    "                       inner='quartile',\n",
    "                       cut=0,               # 0 means ending sharp at end points\n",
    "                       width=.7,\n",
    "                       orient = 'h',\n",
    "                       zorder=20,)\n",
    "\n",
    "# change the medium default line to full (https://stackoverflow.com/questions/60638344/quartiles-line-properties-in-seaborn-violinplot)\n",
    "for l in axsns.lines[1::3]:\n",
    "    l.set_linestyle('-')\n",
    "    l.set_linewidth(1.2)\n",
    "    l.set_color('black')\n",
    "    l.set_alpha(0.8)\n",
    "\n",
    "# Apply colorscheme to violinplots https://stackoverflow.com/questions/70442958/seaborn-how-to-apply-custom-color-to-each-seaborn-violinplot\n",
    "from matplotlib.collections import PolyCollection\n",
    "for ind, violin in enumerate(axsns.findobj(PolyCollection)):\n",
    "    violin.set_facecolor(my_cmap_light[ind])\n",
    "\n",
    "\n",
    "# 2. Plot individual points\n",
    "y_tilt = -0.25                                      # Set some delta for the points below the violinplot\n",
    "y_base = np.zeros(dfq1b.values.shape[0]) + y_tilt   # base vector to which to broadcast y-tilt values\n",
    "\n",
    "for i, col in enumerate(modes):\n",
    "    ax.plot(dfq1b[col],\n",
    "            y_base + i,\n",
    "            # 'o',          # circles\n",
    "            '^',            # triangles_up\n",
    "            alpha=1,\n",
    "            zorder=20,      # higher means more visible\n",
    "            markersize=11,\n",
    "            markeredgewidth=0,\n",
    "            # markerfacecolor='none',\n",
    "            markerfacecolor=my_cmap_dark[i],\n",
    "            markeredgecolor=my_cmap_dark[i],)\n",
    "    ax.plot(dfq1b[col],\n",
    "            y_base + i,\n",
    "            # 'o',          # circles\n",
    "            '^',            # triangles_up\n",
    "            markersize=11,\n",
    "            markerfacecolor='white',\n",
    "            markeredgewidth=1,\n",
    "            color ='white',\n",
    "            linewidth = None,\n",
    "            zorder=1,)\n",
    "\n",
    "\n",
    "# 3. Plot gray lines connecting modes\n",
    "for i, idx in enumerate(dfq1b.index):\n",
    "    ax.plot(dfq1b.loc[idx, modes],\n",
    "            [y_tilt, y_tilt+1],\n",
    "            color ='gray', linewidth = 2, linestyle ='-', alpha = .2,\n",
    "            zorder=0)\n",
    "\n",
    "\n",
    "# 4. Plot red line connecting medians\n",
    "ax.plot(np.median(dfq1b, axis=0), [0, 1], color ='red', linewidth = 2, linestyle ='-', alpha = .4)\n",
    "\n",
    "\n",
    "# 5. CI Errorbars\n",
    "for i, mode in enumerate(modes):\n",
    "    plt.errorbar(median_time[mode], i, xerr=np.array([[ci_delta[mode][0], ci_delta[mode][1]]]).T,\n",
    "                 fmt='o', markersize=10,\n",
    "                 # lw = 3,          # if end line for CI\n",
    "                 lw = 5,            # if no ned line for CI\n",
    "                 alpha=1,\n",
    "                 zorder=100,        # higher means more visible\n",
    "                 capsize = 0,      # 10\n",
    "                 # capthick = 4,    # end line for CI\n",
    "                 capthick = 0,      # no end line for CI\n",
    "                 # color = 'black',\n",
    "                 color = my_cmap_dark[i],\n",
    "                 )   # my_cmap[1])\n",
    "    ax.text(median_time[mode],\n",
    "            # i+0.36,\n",
    "            i-0.16,\n",
    "            f'{median_time[mode]:.1f}', horizontalalignment='center',\n",
    "            # color='black',\n",
    "            color= my_cmap_dark[i],\n",
    "            fontsize=figfont_size)\n",
    "    # ax.text(ci[mode].low, i+0.1, f'{ci[mode].low:.1f}', horizontalalignment='center', color='black', fontsize=20)\n",
    "    # ax.text(ci[mode].high, i+0.1, f'{ci[mode].high:.1f}', horizontalalignment='center', color='black', fontsize=20)\n",
    "\n",
    "\n",
    "#Additional settings\n",
    "ax.set_xticks(range(0, 100, 5))\n",
    "ax.set_yticks(range(len(dfq1b.columns)))\n",
    "ax.set_yticklabels(modes, size= ylab_size)\n",
    "ax.set_xlim(0, 40.1)\n",
    "ax.set_ylim(-0.5, 1.5)\n",
    "ax.set_xlabel('Median time per worker (sec)', size = xlab_size)\n",
    "ax.set_ylabel(None)\n",
    "ax.set_yticklabels(['DX', 'SQL', ])\n",
    "# ax.set_title('Median times per worker', size = title_size)\n",
    "sns.despine()\n",
    "ax.legend_.remove()\n",
    "\n",
    "plt.grid(axis = 'x', linewidth = 0.5, color = 'lightgray')\n",
    "\n",
    "if savefig:\n",
    "    plt.savefig(fig_dir + f'/q1_figure1_variant{VARIANT}.pdf', bbox_inches='tight')\n",
    "    plt.savefig(fig_dir + f'/q1_figure1_variant{VARIANT}.svg', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fbc86c76",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "# Question 1. Figure 1b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "cb8b617c",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dfq1c:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">median</th>\n",
       "      <th>ratio median</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mode</th>\n",
       "      <th>RD</th>\n",
       "      <th>SQL</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>worker_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7.018</td>\n",
       "      <td>28.357</td>\n",
       "      <td>0.248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9.144</td>\n",
       "      <td>12.800</td>\n",
       "      <td>0.714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>9.378</td>\n",
       "      <td>16.846</td>\n",
       "      <td>0.557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>9.357</td>\n",
       "      <td>16.023</td>\n",
       "      <td>0.584</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>9.960</td>\n",
       "      <td>13.986</td>\n",
       "      <td>0.712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>5.756</td>\n",
       "      <td>5.691</td>\n",
       "      <td>1.011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>8.377</td>\n",
       "      <td>12.603</td>\n",
       "      <td>0.665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>10.586</td>\n",
       "      <td>12.822</td>\n",
       "      <td>0.826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>17.987</td>\n",
       "      <td>25.491</td>\n",
       "      <td>0.706</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>7.220</td>\n",
       "      <td>9.152</td>\n",
       "      <td>0.789</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>13.366</td>\n",
       "      <td>23.652</td>\n",
       "      <td>0.565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>7.929</td>\n",
       "      <td>13.225</td>\n",
       "      <td>0.600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>5.414</td>\n",
       "      <td>8.546</td>\n",
       "      <td>0.633</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>13.305</td>\n",
       "      <td>17.604</td>\n",
       "      <td>0.756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>5.921</td>\n",
       "      <td>8.445</td>\n",
       "      <td>0.701</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>10.258</td>\n",
       "      <td>8.849</td>\n",
       "      <td>1.159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>11.616</td>\n",
       "      <td>9.700</td>\n",
       "      <td>1.198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>8.540</td>\n",
       "      <td>12.380</td>\n",
       "      <td>0.690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>18.521</td>\n",
       "      <td>28.279</td>\n",
       "      <td>0.655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>12.252</td>\n",
       "      <td>12.471</td>\n",
       "      <td>0.982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>18.601</td>\n",
       "      <td>31.567</td>\n",
       "      <td>0.589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>15.697</td>\n",
       "      <td>30.716</td>\n",
       "      <td>0.511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>14.380</td>\n",
       "      <td>18.619</td>\n",
       "      <td>0.772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>10.471</td>\n",
       "      <td>38.456</td>\n",
       "      <td>0.272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>19.988</td>\n",
       "      <td>42.377</td>\n",
       "      <td>0.472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>20.476</td>\n",
       "      <td>54.861</td>\n",
       "      <td>0.373</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>4.865</td>\n",
       "      <td>6.268</td>\n",
       "      <td>0.776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>17.733</td>\n",
       "      <td>18.866</td>\n",
       "      <td>0.940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>10.912</td>\n",
       "      <td>30.937</td>\n",
       "      <td>0.353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>22.293</td>\n",
       "      <td>25.159</td>\n",
       "      <td>0.886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>11.087</td>\n",
       "      <td>18.552</td>\n",
       "      <td>0.598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>4.360</td>\n",
       "      <td>4.181</td>\n",
       "      <td>1.043</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>7.749</td>\n",
       "      <td>5.184</td>\n",
       "      <td>1.495</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>8.207</td>\n",
       "      <td>10.006</td>\n",
       "      <td>0.820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>4.792</td>\n",
       "      <td>8.986</td>\n",
       "      <td>0.533</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>7.131</td>\n",
       "      <td>6.258</td>\n",
       "      <td>1.140</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>10.635</td>\n",
       "      <td>14.470</td>\n",
       "      <td>0.735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>18.003</td>\n",
       "      <td>4.572</td>\n",
       "      <td>3.937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>3.221</td>\n",
       "      <td>5.837</td>\n",
       "      <td>0.552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>4.337</td>\n",
       "      <td>16.021</td>\n",
       "      <td>0.271</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>17.566</td>\n",
       "      <td>14.229</td>\n",
       "      <td>1.235</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>14.242</td>\n",
       "      <td>50.465</td>\n",
       "      <td>0.282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>154</th>\n",
       "      <td>10.534</td>\n",
       "      <td>14.382</td>\n",
       "      <td>0.732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>6.451</td>\n",
       "      <td>12.857</td>\n",
       "      <td>0.502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>9.951</td>\n",
       "      <td>15.182</td>\n",
       "      <td>0.655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>5.090</td>\n",
       "      <td>9.694</td>\n",
       "      <td>0.525</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>8.392</td>\n",
       "      <td>10.698</td>\n",
       "      <td>0.784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>166</th>\n",
       "      <td>11.477</td>\n",
       "      <td>12.369</td>\n",
       "      <td>0.928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>30.665</td>\n",
       "      <td>36.904</td>\n",
       "      <td>0.831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>4.778</td>\n",
       "      <td>7.235</td>\n",
       "      <td>0.661</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          median        ratio median\n",
       "mode          RD    SQL             \n",
       "worker_id                           \n",
       "0          7.018 28.357        0.248\n",
       "2          9.144 12.800        0.714\n",
       "5          9.378 16.846        0.557\n",
       "6          9.357 16.023        0.584\n",
       "10         9.960 13.986        0.712\n",
       "17         5.756  5.691        1.011\n",
       "21         8.377 12.603        0.665\n",
       "28        10.586 12.822        0.826\n",
       "32        17.987 25.491        0.706\n",
       "39         7.220  9.152        0.789\n",
       "42        13.366 23.652        0.565\n",
       "43         7.929 13.225        0.600\n",
       "50         5.414  8.546        0.633\n",
       "52        13.305 17.604        0.756\n",
       "57         5.921  8.445        0.701\n",
       "58        10.258  8.849        1.159\n",
       "60        11.616  9.700        1.198\n",
       "66         8.540 12.380        0.690\n",
       "72        18.521 28.279        0.655\n",
       "75        12.252 12.471        0.982\n",
       "77        18.601 31.567        0.589\n",
       "80        15.697 30.716        0.511\n",
       "81        14.380 18.619        0.772\n",
       "83        10.471 38.456        0.272\n",
       "87        19.988 42.377        0.472\n",
       "89        20.476 54.861        0.373\n",
       "91         4.865  6.268        0.776\n",
       "92        17.733 18.866        0.940\n",
       "96        10.912 30.937        0.353\n",
       "110       22.293 25.159        0.886\n",
       "115       11.087 18.552        0.598\n",
       "117        4.360  4.181        1.043\n",
       "119        7.749  5.184        1.495\n",
       "121        8.207 10.006        0.820\n",
       "125        4.792  8.986        0.533\n",
       "130        7.131  6.258        1.140\n",
       "136       10.635 14.470        0.735\n",
       "141       18.003  4.572        3.937\n",
       "143        3.221  5.837        0.552\n",
       "146        4.337 16.021        0.271\n",
       "148       17.566 14.229        1.235\n",
       "153       14.242 50.465        0.282\n",
       "154       10.534 14.382        0.732\n",
       "158        6.451 12.857        0.502\n",
       "159        9.951 15.182        0.655\n",
       "162        5.090  9.694        0.525\n",
       "165        8.392 10.698        0.784\n",
       "166       11.477 12.369        0.928\n",
       "168       30.665 36.904        0.831\n",
       "169        4.778  7.235        0.661"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Median ratio: 0.703, 95% CI [0.627, 0.772]\n",
      "Number (fraction) of users faster with RD: 42 (0.840)\n"
     ]
    }
   ],
   "source": [
    "dfq1c = df_filtered_data.groupby(['worker_id', 'mode']).time.agg(['median'])                               # for each worker, calculate median for both modes\n",
    "dfq1c = pd.pivot_table(dfq1c, values=['median'], index=['worker_id'], columns=['mode'])       # pivot to have one row per worker\n",
    "dfq1c['ratio median'] = dfq1c['median','RD'] / dfq1c['median','SQL']                           # add the ratio between medians of the two modes\n",
    "\n",
    "print('dfq1c:')\n",
    "display(dfq1c)\n",
    "\n",
    "sample = np.array(dfq1c['ratio median'])                                                     # extract the sample and then create the boostrapped medians\n",
    "data_ratio = dfq1c['ratio median']\n",
    "\n",
    "median_ratio = np.median(data_ratio)\n",
    "ci_ratio = scipybootstrap((data_ratio,), statistic=np.median, n_resamples=BOOTSTRAPSAMPLES, confidence_level=BOOTSTRAPCONFIDENCE, method=BOOTSTRAPMETHOD, axis=0).confidence_interval        #convert array to sequence\n",
    "ci_ratio_delta = [median_ratio - ci_ratio.low, ci_ratio.high - median_ratio]\n",
    "\n",
    "print(f'Median ratio: {median_ratio:.3f}, 95% CI [{ci_ratio.low:.3f}, {ci_ratio.high:.3f}]')\n",
    "print(f'Number (fraction) of users faster with RD: {np.sum(data_ratio<1.0)} ({np.sum(data_ratio<1.0)/np.sum(data_ratio>0.0):.3f})')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b3c5ea47",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0QAAAFECAYAAAAHhv4HAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB+eElEQVR4nO3deXgTVdsG8DtJs7TpThdogW4QQHYQUMENfFUEFBUUFfjcRQUVBEFRXBAVFEVQwQV9VRQUEfEVRFREQCmK7GsLlKUb3aHN0qbJfH+kGROatE2areT+XRcXaWbmzJmTk8k8c5aRCIIggIiIiIiIKAhJ/Z0BIiIiIiIif2FAREREREREQYsBERERERERBS0GREREREREFLQYEBERERERUdBiQEREREREREGLAREREREREQUtBkRERERERBS0/BoQnTlzxp+7D2gsG8dYLo6xXJxj2TjGcnGOZeMYy6W+tm3bYuLEiWjbtq2/sxKQWGccY7k45s9y8WtAZDAY/Ln7gMaycYzl4hjLxTmWjWMsF+dYNo6xXBzr1KmTv7MQsFhnHGO5OObPcmGXOSIiIiIiCloMiIiIiIiIKGgxICIiIiIioqDFgIiIiIiIiIIWAyIiIiIiIgpaDIiIiIiIiChoMSAiIiIiIqKgxYCIiIiIiIiCFgMiIiIiIiIKWgyIiIiIiIgoaDEgIiIiIiKioMWAiIiIiIiIghYDIiIiIiIiCloMiIiIiIiIKGgxICIiIiIioqDFgIiIiIiIiIJWiL8zQER0ITOZTNDr9TAYDNDr9fVe19TUQBAEAIAgCOK/8/9u6H1H25rNZrv3AUCpVIr/dDodCgsL7d5TKpVQqVTi65CQEEgkEp+WFxERka8xICIiaoAgCKipqXEa0DgLdqz/19TU+PsQGmQNlhwFPlKptF7AdH7wpFAo7IIopVKJiIgIREZGQq1WQyplRwQiIgpsDIiIKGjV1NSgoqICFRUVKC8vF19XVFSgqqpKDGysrS3NoZBKoZLJoJLJoJTJoJJZ/pZLpZBAAgkASCC+tsYn9svs37Nbx8n61r8FAEazGdVmE2pMZtSYzag21eKjb1bBDODWEcNhNAuoNpthrDtes9ksBnrukMlkiIiIQFRUFCIjI+3+t76OiIiATCZzK30iIiJPYEBERBcsvV6PkpISaLVah4GPVqttclpSoC6QsQ1qZFDKzg906l5L/w16lDIZpAHY9UxfU4NF5eUAgLFpaQhVKADUtYqZLUFTjcmEapvXlkDKjBqzqe69uiDL+tpkQlVtLbS1tTCZTGJZOyORSBAeHi4GSOcHTNb/5XK5L4qEiIiCEAMiImqRBEFAVVWVw9Yd63vV1dWNpqOQShEplyNSoUCUXI5IhRyRcgXC5SF1gY1NS04ABjXeIJFIoKwL5OBmIGISBGiNRlQaa1FZa0SV0YhK6991r6tqa2EWBFRWVqKysrLB9MLCwuoFTTExMUhISEBcXByUSqVb+SQiImJAREQBraqqCmfOnMGZM2dQVFSEsrIyMeipra1tdPtQmQyRCjmi5Iq6wKfutUKOKIUCyiAKdHxJJpEgUqFAZF2rkyOCIEBXa0JlbV2A5CBoqjQaUSsI0Ol00Ol0KCgocJhWVFQU4uPjxX8JCQmIj49HeHg4P18iImoQAyIiCgjV1dVi4GP9V1hY2Gi3tvCQEEQpLMFOVF3rjiXokSNCoYCCg/oDlkQigVoeArU8BK1DQx2uIwiWcU22AZL1X3lNDcqqq6GrNeHs2bM4e/Ysjh49are9SqWyC5Cs/2JiYjh2iYiIADAgIiIfq62tRXFxcb3gp7xuLIsj0Qo54lQqxClViFFagx8FIkJCIGPAc0GTSCTiuKx4lcrhOvraWpRV16C02oDS6hqUVlejrLoaZ+tmBzx9+jROnz5tt41MJkOrVq0cBktERBRcGBARkVeYzWaUlZXVa/EpLS11OmubOiQE8Sol4pQqxKtUiFMp0UqlYisPNSg0JATJISFIVofZvV9rNqO8pgalhmoxSLL+X2syoaioCEVFRThw4IDdduHh4WjTpo1d97ukpCSOUyIiukAxICKiZhEEAefOnavX4lNUVASj0ehwG6VUWhfwWIIeSwCkRGgIT0nkOSF19ez8liVBEHDOaESpTYBUarD8rzOZUFVVhezsbGRnZ4vbSCQSxMXFITk5GUlJSeL/DJKIiFo+Xn0QkUtqamqQm5srdkM6ffq00xnCQiQStFIqEaf6t8UnXqVCeEgIB7qT30gkEkQpFIhSKJAeEWG3TF9bax8oVVej2GBApdHS1bO4uBi7d+8W02GQRETU8jEgIiKnzGYzSkpK7IKfwsJCCIJgt54EQIxSadPdzRIERSsUAfn8HSJnQkNC0DYkBG3Varv3tcZaFOr1OKPXo9Bg+b+xIMkaIDFIIiIKbAyIiEik1Wrtgp/c3FwYDIZ660XIQ9AmNAxJYWFoExaK1qGhkDcwzqfGbMaCAwcBAE90vYhjgqjFUctDkCGPQEbkvy1KDJKIiC4MDIiIglRtbS0KCwtx6tQp5Obm4tSpUygrK6u3XohEgtahoWLwkxQWhggXH9YpARBZtw3bi+hCwSCJiOjCwICIKAgIgoCKigq71p/8/HyHDzaNVSqQFPpv8BOnUkHWzG5vcqkUEzp3alYaRC1Bc4Ok1q1bIyUlBampqUhJSUFUVJSfjoSIKHgwICK6ANXW1iI3NxcnTpwQA6Cqqqp666lkMiSFhqJNWBiSwiz/q/iwSiKPakqQVKjTo6q2FgUFBSgoKEBmZiYAIDo6GqmpqWjfvj1SU1ORkJAAKbucEhF5FAMioguAyWTCyZMnkZOTg+PHj+PkyZP1pryWAogPVSEp9N/gJ0ah4GxvRH7gKEiqNBqRp9MhT6tDrk6LIr0BFRUV2L17t9iKpFKpxOAoJSUFbdu2hdzFLqxERGSPARFRC2QymZCfn4/jx4/j+PHjyMnJqdf9LVQmQzu1Gkl1rT+JjUx8QET+FSGXo3NUFDrXdZOrNplQoNeLAVK+Tg+DwYCsrCxkZWUBAGQyGZKSksQAqX379ggPD/fnYRARtTgMiIhaAJPJhIKCAjEAOnHiBGpqauzWUclkaK9Wo71ajXbhasQplQHT+mMSBOwqLQUA9G7VqtljkohagubWe6VMhtTwcKTWBThmQUCRwSAGSLlaHbS1tWK32C1btgAA4uLi7MYhtWrVKmDOBUREgYgBEVEAMpvNYgCUk5ODnJwcVFdX262jlErRLlxdFwSFI14VOAHQ+UyCgI0FhQCAHrGxDIgoKHi63kvrZnxsHRqKvmgFQRBw1mhEnlaL3LqudiXV1SgpKUFJSQn++ecfAIBarbYLkNq0aYOQEP78ExFZ8YxIFADMZjPOnDlj1wXu/Of/KKVStFWr0b4uCIpXqfjQU6IgJpFIEK1QIFqhQNeYGACAvrbWMg6pLkAq0Ouh1Wpx8OBBHDxoeRaYXC5H+/btkZGRgfT0dCQnJ0PGyVSIKIgxICLyA7PZjKKiIrsASK/X262jkErRVh2G9upwtA9XI4EBEBE1IjQkBB0iI9EhMhIAUGs244zeIHaxy9PpYDAacezYMRw7dgwAoFAokJaWhvT0dKSnp6NNmzacyY6IggoDIiIfOXv2LLKzs5GdnY1jx45Bp9PZLZdLpWgbFia2ACWGhjIAIqJmCZFKkawOQ7I6DAPiLc8kK62uximtFqeqtDil1cJQU4MjR47gyJEjACwz2VmDo/T0dCQmJgZsd1wiIk9gQETkJUajESdOnEB2djaysrJQVFRkt1wukSC5rgWonVqN1mGhF+zYmhCJBGPSUsXXRMEgEOu9RCJBnEqFOJUKfVpZxiEVGQxicHRaq4XBYLDrYqdWq5Geno7o6Gio1WpO0kBEFxwGREQeIggCiouLxVagnJyces8CahMairQIy6xRbUJDIQuSbilSiQTtORUwBZmWUO8lEgkSQy3T8veLj4NZEFCo19cFSFWWmey0Wuzbtw8AsGXLFkRGRorjj9LT0xFTN36JiKilYkBE1Ax6vR7Hjh0Tg6CKigq75eEhIXUBUARSw9UI5cxORBTApBJJ3bPLwnAJ4lFrNqPAJkDK1+lx7tw57Nq1C7t27QIAxMbG2nWxi6wbv0RE1FLw6ozIBWazGfn5+cjKykJ2djZOnz4Ns9ksLpdJJGirDkNaeATSIsID6llARESuCpFK0U6tRju1GgORAKPZjHydDiertDhVVYUCvR5lZWUoKyvDjh07AADx8fFicJSRkYGwsDA/HwURUcMYEBE14ty5c2IL0NGjR+tNhhCrVCAtPAKp4eFoF66GIki6wblCEARU1wWOSqmUQSIFhQux3sulUqSEhyMlPBxAIqpNJuRqdXWTNFThjMGA4uJiFBcXY/v27ZBIJEhKSkKHDh3QoUMHpKSk8BlIRBRweFYiOk9tba04GUJ2djYKCwvtlivqLgjSwsORFhGOKIXCTzltOYyCgIUHDwEAnuh6ERQXwIUhUWOCod4rZTJkREYgIzICgOU5SLlaHU5qq3CqSouS6mrk5eUhLy8Pv//+O+RyOdLS0sQAiTPYEVEgYEBEBKCsrAxHjhxBVlYWjh8/Xm8yhNahoUgLD0dqRDiSwsIu2NngiIiaIzQkBB2jItExyjKOqNJoxMmqKpyoqsLJKi20RiOysrKQlZUFAIiIiEBGRoYYIHH8ERH5AwMiCkomkwknT57EkSNHcPjwYRQXF9stV4eEILWuBSg1PBxh7OJBROSyCLkc3WJi0C0mBoIgoKS6GicqLQHSaa0WlZWV2L17N3bv3g0ASEhIQMeOHdGhQwekpaVBwRZ4IvIBXuVR0NBqtcjKysLhw4eRnZ0Ng8EgLpMAaKsOQ3pEBNLCwxGvUrEbBxGRB0kkEsSrVIhXqdAvPg61ZjPydDqcqGtBOqM3oKioCEVFRfjjjz8gk8nQvn17MUBKSkqClGM0icgLGBDRBUsQBBQWFoqtQKdPn4YgCOLyUJkM6RERliAoIhwqmcyPub2wSQF0i44WXxMFA9b7hoXYTNBwJQBdbS1OVWnFAOmc0YicnBzk5ORgw4YNCA0NRUZGhhgg8flHROQpDIjoglJTU4Njx47hyJEjOHLkCM6ePWu3PEGlQnpEBDIiItAmLBRStgL5RIhUihvatfV3Noh8ivXeNWEhIegcHYXO0VEQBAHlNTVicHSqSgu9Xo/9+/dj//79AIBWrVqJY4/S09MRGhrq5yMgopaKARG1eOXl5WIr0PHjx1FbWysuC5FIkBIejoyICKRHhCOS/dGJiAKeRCJBrFKJWKUSfVq1glkQUKDTiwFSvk6H0tJSlJaWYvv27ZBKpWjbti06duyIjh07Ijk5GTK2+hNREzEgohbHZDLh1KlTYivQmTNn7JZHyuXIiLBMA9tOrYacfc6JiFo0qUSCZHUYktVhGJiYgGqTCae0WnGChvKaGpw6dQqnTp3Cr7/+CpVKhQ4dOogBUnRd10UiIkcYEFGLoNPpkJWVhZ07dyIvLw96vV5cJgGQHBaGjEjLeKA4pZITIhARXcCUMhk6RkaiY9003Wet3evqAiSDwWDXvS4uLk4MjtLS0qBUKv2ZfSIKMAyIKCAJgoDi4mIcOnQIhw8fxqlTp+wmRFDJZEgLD0dGpGVWuFBOix3Qas1mfH/6NADgxnbtEMJWOwoCrPe+E6VQoGdsLHrGxsIsCCjU65FT+W/3upKSEpSUlGDbtm2QyWRISUkRA6TWrVtz9jqiIMerSAoYJpMJJ06cwOHDh3Ho0CGUlZXZLY9XKcUJEZLCwjghQgtiBnD0XKX4migYsN77h1QiQVJYGJLCLN3rDCYTTlVVIaeuBems0Yjjx4/j+PHj+Omnn6BWq8XudXw4LFFwYkBEfqXX68UJEbKysuyeDSSTSNBerUaHuq5wUZwQgYiIXKSSyaCJioImyn72upzKKpzSaqHVarFnzx7s2bMHANC6dWux9SglJQVyudzPR0BE3saAiHyupKRE7Ap38uRJmM3/3jsNk8mQHhmBDhERSAkPh5KzBBERkYecP3udyWxGvk33ukK9HoWFhSgsLMSWLVsQEhKCtLQ0MUBKSEjgGFWiCxADIvI666xw1q5wJSUldsvjlEpkREagQ0Qknw1EREQ+I5NK0U6tRju1GlcgEbraWpy06V5XVVuL7OxsZGdnAwAiIyPFrnUZGRl+zj0ReQoDIvIKg8GArKwsHD58GEeOHLGbFU4KoJ1ajQ6RkciIjEA0u8Jd8BRSKZ7q3s3f2SDyKdb7licsJARdoqPRJToagiCgtLpabD06rdXi3Llz+Oeff/DPP/8AAIYNG4Zz584hMTERtbW1COEEP0QtEr+55DFlZWViK1BOTo5dVziVTIb0iHB0iIhEWgS7whERUWCTSCSIU6kQp1KhX3wcas1m5Gp14sNhiwwGtGrVClqtFtdddx1mz54tdq/r0KEDu9cRtSAMiMhtZrMZp0+fFoOgoqIiu+WxSgUyIiLRITICyZwVjoiIWrAQqRSpEeFIjQgHAFQZjThZpRUDJK3RiKysLGRlZQGwdK+zzl6XkZGB8PBwf2afiBrAgIhcUl1djezsbLErnFarFZdJALRVh4lBUCwffEdERBeocLkcXWOi0TXG0r2uxKZ7XW5d97qdO3di586dAICkpCQxQEpJSWH3OqIAwm8jNaq8vByHDx/G4cOHcfz4cZhMJnGZUipFWkQEOvABqdQAsyDgWN3zWDIiI9haSEGB9T54SCQSxKtUiFep0N/avU6nw4nKf7vX5efnIz8/H5s3b4ZcLkdaWpoYILF7HZF/8eqV6jGbzcjLyxOnxi4sLLRbHq1QIKMuCGqrVkPGkzg1olYQsPrUKQDAE10vgoJ1hoIA633wCpFKkRoejtS6bnJaY63Yta6h7nXWf+xeR+RbDIgIgKUr3NGjR8WWoPO7wiWHhdVNjW3pCsc7WURERE2jlofU61534rzZ687vXpeRkYEOHTogJSUFCs7GSuRVDIiCWEVFhV1XuNraWnGZQipFWkQ4OkREIC0iAmHsCkdERNRstt3r+jXSvW7Lli2QyWRo3769GCAlJydDxplaiTyKV7lBxNoVzhoEFRQU2C2PksvFZwO1CwuDTCr1U06JiIiCg7Pudae0VThRpUWl0YicnBzk5OTgl19+gVKpRHp6OjIyMpCRkcHxR0QewIDoAldTUyN2hTty5AgqKyvtlieHhYnjgVqxKxx5iUwiwdC2yeJromDAek/uOL97XXlNDU5WVeFklRantFoYqqtx6NAhHDp0CAAQEREhth5lZGQgKirKz0dA1PIwILoAnT17VmwFOnbsmF1XOLlUirTwcHSIjEA6u8KRj8gkEnSPifF3Noh8ivWemksikSBWqUSsUonerVrBLAg4ozfUtR5VIU+rQ2VlJXbv3o3du3cDAOLi4sTgKD09HaGhof49CKIWgFfDFwCz2Yzc3FwcPnwYWVlZyM/Pt1seKZejQ2QEMiIi0E6tRgi7whEREbU4UokEbcJC0SYsFAPi41FrNiNPpxNbkAr1epSUlKCkpASZmZmQSCRITk4WW5Dat28PuVzu78MgCjgMiFoovV6PrKwsHDlyBFlZWdDpdHbLk0JDLbPCRUYijl3hiIiILjghUilSwsORUjf+yGAy4XSVFieqqnBSW4Wy6hrk5uYiNzcXv//+O0JCQpCamiqOP0pKSoKUN0mJGBC1FIIg4MyZMzhy5AiOHDmCU6dOwWw2i8uts8Jl1M0Kp2ZXOCIioqCiksnQMSoSHaMiAQCVRqPYenSiqgra2locPXoUR48eBQCEhoYiLS0N6enpSEtLQ2JiIgMkCkq8ag5gNTU1OH78uBgEVVRU2C1vpVQiI8IyFihZHcZBuxSwasxmLD50GADwcJfOUPAHl4IA6z35W4Rcjm4xMegWEwNBEFBaXY2TVVqcrKrCKa0Wer0eBw8exMGDBwEAYWFhSEtLE4OkhIQEBkgUFBgQBZjy8nIcPnwYu3fvRkFBgd2ECCESCdqr1UivmxAhmg9qoxak2qZFkyhYsN5ToJBIJIhTqRCnUqFvnGWChkK9HqfqZq/L02qh0+lw4MABHDhwAMC/AVJ6eroYILELPl2IGBD5mclkwsmTJ8VWoKKiIrvlkXI50iMsEyK0D1dDzjs1RERE1ExSiQRJYWFICgvDJYiHSRBQqNPjtFaLU1rLDHbnB0hqtbpeCxIDJLoQMCDyg6qqKnFChOzsbBgMBnGZBHXPBqprBeKECERERORtMokEyeowJKvrAiSzGYV1U3xbWpB00Gq12L9/P/bv3w/g3wDJ2oIUHx/PaxZqkRgQ+YDZbEZBQYHYCpSbmwtBEMTloTIZ0uvGAqVFhEMlk/kxt0SeJwHQTh0mviYKBqz31JLJpFIxQLoUgMlsRoFej1NaLU5XaZGnqx8ghYeH203SYHutQxTIGBB5SVVVFbKzs8V/Wq3WbnmCSoWMiAhkREagdWgopLyjQhcwuVSKO9LT/Z0NIp9ivacLiUwqRVu1Gm3VaiABqDWb7cYg5et0qKqqwr59+7Bv3z4AljFIHTp0ELvZsQWJAhUDIg+xjgWyBkDnPxxVIZUiJVwttgRF8MFoRERE1EKF2ARIl8ESIBXUBUintZYWJJ1Oh71792Lv3r0ALAFSamoqUlJSkJqaijZt2iCEjwmhAMBa2AxlZWXIyspCdnY2jh8/jurqarvliSoV0iLCkRYRgaQwTotNREREF6YQqRTt1Gq0U6sB2AdIp7RVKNDpodPp7Kb5lsvlaNeunRggtWvXDiqVyp+HQUGKAZELrM8Fys7ORlZWFkpLS+2Wh8lkSI0IR1p4BFLDw6GWs3iJiIgo+NgGSAORAJPZjDMGg6X1SKtDrk4Hg9GI48eP4/jx4wAsU4O3adNGDJBSUlIQGRnp5yOhYMAr9gYIgoAzZ86IrUAnTpyAyWQSl0sBJKnDkBZuaQVKVKnYN5bIgVqzGb8XngEAXNk6ESGcPp6CAOs90b9kUqk4zTfiLddYZdXVyNXpkKvVIVerxVmjEfn5+cjPz8e2bdsAALGxsXYBEschkTcwIDqPTqfD0aNHxbFA586ds1seJZcjNSIc6eGW5wIpOSMcUaPMAP6pa1G9vHWifzND5COs90TOSSQStFKp0EqlQs/YWABApdFY13qkRa5Wh2KDAWVlZSgrK8OuXbsAWMYh2QZISUlJHIdEzRb0NchkMiEvL09sBTp/SuwQiQTtw9VIC7dMiR2jUPDOBBEREZGHRcjl6Bwdhc7RUQCAapMJ+dYWJJ0OBXUTNRw6dAiHDh0CYBmH1LZtWzFAat++PcchkcuCOiA6c+YMli5diqqqKrv345RKcTKEtmFh7OZARERE5GNKmQxpERFIi4gAAHEcUm5dK1KeVge90YicnBzk5OQAsLQ8JSQkoF27duK/hIQESHktRw0I6oAoJycHVVVVkEulSK+bDCEtIpxTYhMREREFGNtxSP0RVzcOqUbsYpen06KixogzZ87gzJkz2LFjBwBAqVQiOTkZ7du3F4Ok8PBwPx8NBZKgDois0sPDcVP79v7OBtEFSy6R4KFOGvE1UTBgvSfyLss4JCVaqZTiOKQqoxEFej3ydToU6PQo0OtRXV1tN5sdAMTExKBdu3ZikMRnIgU3fvJE5HUSiQRRCoW/s0HkU6z3RL4XLpejo1yOjnXTdZsFASWGahTodcjX6ZCv06O0uhrl5eUoLy8XHxork8mQlJQktiC1b98e0dHRHDceJBgQEREREdEFSSqRICFUhYTQf2ezqzaZxFakfJ0eBTod9CYTTp8+jdOnT4vbqtVqu252bdu2hVKp9NehkBcxICIiIiKioKGUyZAaHo7UunFEgiCgoqbGEhzpLUFSkcEArVZrN6OdRCJBYmIi2rZti/bt2yM5ORkJCQmQ8REsLR4DIiLyOrMgoNhgAADEq1SQsgsCBQHWe6KWQSKRIEapRIxSia4x0QAsD1Y+ozcgv66rXYFOj3NGIwoLC1FYWChO2BASEoLWrVsjOTkZycnJSEpKQmJiIoOkFoYBERF5Xa0g4NOjxwAAT3S9CApeGFIQYL0narlCpFIkq8OQrA4T36syGlGg0yNfbwmQzuj1qK6tRW5uLnJzc//dti5ISkpKsguSOGlD4OInQ0RERETUiHC5HB2j5OgYZZmwwdrVrlBvwBm9HoV650GSTCYTgySVSoWQkBAGSQGEnwIRERERkYtsu9p1iY4C8G+QdEZvEAOkQr0e1SYT8vLykJeXBwDYsmULZDIZEhMTxVak5ORktG7dmkGSH7DEiYiIiIg8wDZI6mwTJJ01Gu0CpDN6AwwmE/Lz85Gfny9ubw2SrAFScnIyEhMTIZfL/XVIQYEBERF5nQzAZQnx4muiYMB6T0SAJUiKVigQrVCgc5R9kGTb1a7wvCDJOnGDRCJBXFwcWrdujTZt2qB169Zo3bo1oqKi+JwkD2FAREReJ5NKMSgx0d/ZIPIp1nsicsY2SOpkEySdE1uS/h2XpDeZUFxcjOLiYuzbt09MIzQ0VAyOrMFSQkICFHwgtMsYEBERERER+ZlEIkGUQoGo84KkqtpaFBsMKDYYUKS3/F9WXQ29Xo+cnBzk5OTYpdGqVSu7IKl169aIjo5ma1IDGBAREREREQUgiUSCCLkcEXI50iMixPdrzWaUVVejqC5QKjYYUGQwQFdrQklJCUpKSrB//35xfaVSWa81KTExEUql0h+HFXAYEBERERERtSAhUikSQkOREBpq977WWCsGR9ZAqaS6GtXV1Th58iROnjxpt35sbKzYipSYmIiEhAS0atUq6B4sy4CIiLzOaDbji2PHAQB3ZaRDLpX6OUdE3sd6T0S+ppaHQC0PR2pEuPieSRAsrUn6f4OkYoMBVbW1KCsrQ1lZGQ4cOCCuL5PJ0KpVKyQkJCA+Ph4JCQlISEhAXFzcBTvbHQMiIvI6AUCRwSC+JgoGrPdEFAhkEgniVSrEq1R27+tsxyYZDCgxVKO0uhpGkwlFRUUoKiqyW18ikSA2NlYMkKzBUnx8fIvveseAiIiIiIgoyISFhCAlPBwp4f+2JgmCgEqjEaXV1SiprkZpXZBUYjCg2mxGaWkpSktLcejQIbu0oqOjxUDJNmAKPa9LX6BiQERERERERJBIJIhUKBCpUCDNZhIHQRCgra1FabUlQCo1WAKmsupqaGtrUVFRgYqKCmRlZdmlFxERUa/rXUJCAtRqdUDNeseAiIi8TgIgsq7fceCc/oi8i/WeiC4UEokE4XI5wuVyuxYlANDXBUq2LUql1QZUGmtRWVmJyspKHDt2zG4bpVKJuLg4tGrVCnFxcYiLi4PRaERKSoovD0vEgIiIvE4ulWJC507+zgaRT7HeE1EwCA0JQduQELRVq+3erzaZHLQoGVBRY0R1dTXy8vKQl5dnt02/fv18mXURAyIiIiIiIvIopUyGpLAwJIWF2b1fazajoqYG5dU1KKuptvvfXxgQERERERGRT4RIpYhTqRB33qx3/sSHIhARERERUdBiCxEReZ1JELCrtBQA0LtVK8gCaGYZIm9hvSciahkYEBGR15kEARsLCgEAPWJjeWFIQYH1noioZWCXOSIiIiIiCloMiIiIiIiIKGgxICIiIiIioqDFMURE5HUhEgnGpKWKr4mCAes9EVHLwICIiLxOKpGgfXi4v7NB5FOs90RELQO7zBERERERUdBiQEREREREREGLXeaIyOsEQUC12QwAUEqlkHA8BQUB1nsiopaBLURE5HVGQcDCg4ew8OAhGAXB39kh8gnWeyKiloEBERERERERBS0GREREREREFLQYEBERERERUdDipApE5HVSAN2io8XXRMGA9Z6IqGVgQEREXhcileKGdm39nQ0in2K9JyJqGRgQEREFKUEQEAFACcCs00GQyzk1NBERBR0GREREQcZsMEC/ew+qtv+F9xVKAMC5txdBGxMDdf9+CO3VE1KVys+5JCIi8g0GREREQaT66DGUf70SgtFYb5mpvBznftqAyo2/Iea20VB2yPBDDomIiHyL4zyJyOtqzWZ8e/Ikvj15ErVms7+zE7Sqjx5D2ZfLHQZDtgSjEWVfLkf10WM+ytmFifWeiKhlYEBERF5nBnD0XCWOnqsELwv9w2wwoPzrlYAgNG0DQUD51ythNhi8m7ELGOs9EVHLwC5zRERNVPn7ZlRt2ervbLjHbG56MFRHMBpxZt4bgDSw752FXz4IEVde4e9sEBFRC8WACEC12YQS3gUl8hqjTXehUoMB8gC/wHbGZDQCJpO/s+FbghDwx6wzGlEdgOfwC6XeExH5SrSf9suACMCJKi0+zj7q72wQBYXPjx33dxbc1qO0DD39nQmqZ1dpGfYG+Dm8Jdd7IiJfecVP+w3qgKhDhw6Ij4+HVqv1d1bqMZvNkPJuYj0sF8dYLs55smzkcrlH0iHPksvlCAsL81h6/D45xnKpr6ysDGFhYdDpdIiNjfV3dgIO64xjLBfHzH6cfEYiCC52KvegkydPIiUlxV+7D2gsG8dYLo6xXJzzZNkUL3oHJe++65G0yHPiHn0U8ZMmeiw9fp8cY7nU17ZtW0yfPh1z585Fbm6uv7MTcFhnHGO5OObPcmF4SkREREREQSuou8wREblEJoWkhXabE0wmy0xzrpJKIZHJPJ8hT5Lx3h4REbmPARERURPFP/II4h95xN/ZcIvp3DlkX3kVBIOhadNvS6WQKJXo+PsmyCIjvZ9BIiIiP+FtNSKiICCLjETbhQsBicTyryF1y9suWsRgiIiILngMiIiIgkT45YPQ7v33IVGpHAdGde9JVCq0++ADhA8a6J+MEhER+RADIiKiIBJ++SB0/H0T5BMegrxdW7tl8nZtkfj00+i4+XcGQ0REFDQ4hoiIKMjIIiMhHzkS7R97DKaKCpi1OkjVYZBFR0PSWHc6IiKiCwwDIiKiICWRSBASEwPExPg7K0RERH7DLnNERERERBS0GBAREREREVHQYkBERERERERBiwEREREREREFLQZEREREREQUtDjLHBFRgDGeKUL5smWo2rIFxlOnIJjNCElMQFifvogePRphfXq7nXbxondQ8u67AIBDbmzf5bDzrSp/+QUV366GYd8+1FZUQBYZCUW7dogcej2ibh0FWbjazVwTERF5DwMiIqIAUvnLL8if8TTMVVV27xtPnsLZk6dwdvVqxNx5BxJnzoREJvNTLu2ZtVrkTpkC7e+b7d43lZZCX1oK/e7dKPv0MyS/OR+hvXr5J5NEREROMCAiIgoQ2sxM5D7+BGAyAQCUGg3Cr7wSEqUS+j17oN26FRAElH+5HADQetYsl/ehHjgQ0rAwlJeXI6aR5w8JRiNKFi+GUF0NAIj9v/H11zGbkTtpErR/bgMASFQqRPznP1Cmp6G2tAyVGzagtqgIxvx8nHpoAlJXLIcyLc3lfBMREXkLAyIiogBgNhiQ//QzYjDU6uEJiH/sMUgkEnEd7Z9/4vTESRB0OpR/uRwR114H9SUDXNpPWJ/eCOvTG1UnT6JVSkqD6xbMel4MhtSXXYaE6dPrrVP+5XIxGJInJaH9x0uhSE0VlydMmYz8GU+jcsMGmM+eReFzs5Cy7HOX8kxERORNnFSBiCgAVKz8BrUFBQDqgo/HH7cLhqzvt3npJfHv4rff9lp+zq1fj4qvvwYAyOLikDRvLiRS+58MwWhEyftLxL+T3njDLhgCAGlYGJLfeB3Kjh0BALodO1D1xx9eyzcREZGrGBAREQWAs6tXi69bTXjI6XqRw26AIiMDAKDftQs1p097PC+1paUofP4F8e82LzyPkLi4eutVbdkKU3EJACCsf3+nkz1IFAq7Yzr3/feezTAREVEzMCAiIvKz2rIyGA5ZZm+TRkYirG9fp+tKJBKEX3GF+Hflz794PD9nXpsL09mzAICI665DxDXXOFxPu3Wr+Dr8qqsaTDP8iiuAukkgKjf+BsFs9kxmiYiImokBERGRnxkOHAQEAQAQ2q1bo7PHhfbsKb7W793r0bzodu7Cuf/9DwAgCQ1F4oz644asDAcO/JunRmaPk0VEQJlumUzBXFmJmuPHm59ZIiIiD2BARETkZzUnToiv5e3aNbq+PDnJ4baeUDRvnvi61b33Qt6mjdN1bfetaNe20bRDkryXbyIiIncxICIi8rPa4mLxtbxN60bXD0lI+Hfb0hKP5aPqjz+g370bACCLjkbsPfc4XVeoqRG71SEkBCHx8Y2mL7fNd0lps/JKRETkKQyIiIj8zFxVKb6WqFSNri+1WcdcWdXAmq4p+++n4uvY++6FLFztdF2TzYNjpU3IMwBIVKHia9tjJiIi8icGREREfmauqRFfNyW4sA2aBJttm6P66FFot2yx5EGtRswddzS4vu1+mxLEAYBUpRRfmz2UbyIiouZiQERE5GcSSTNOxec9q8hdZZ/+2zoUPepWyMLDG9lv834+zn/GEhERkb8wICIi8jNpWJj4WqiubnR9wWAQX0uUygbWbBpzTQ3OrftR/Dv69tsb3Uaqdi3PAGA2/LueRNH8fBMREXkCAyIiIj+zDS7MekMDa9atYxMQNdqS0wTaP/+EWasFYJnSW5me3ug2dkGcofE8W9bT/7t9RPPzTURE5AkMiIiI/Mxu1riiokbXrz1z5t9tmzC7W2Mqf/n34a6Rw4Y1aRuJVApZXBwAQDAaUVte3ug2Rg/nm4iIyBMYEBER+ZmyQwfxtTEvr9H1jfn54mtFamqz9i2YTKja+Jv4d8Q1Q5q8rV2+c32bbyIiIk9hQERE5GdKjUacHEG/fz8EQWhwff2u3eJrVdeuzdq34dBhmMrKAACKDhmQ2zw8tTGqTp3+zdPePQ2uazp7FjXHcwAA0ogIKFJS3MgtERGR5zEgIiLyM1lkJEL79gEAmEpLYdi3z+m6giCgavNm8e/wywc1a9/6nTvF16Hduru0bfjVV4mvq37/vcF1qzZvAcxmAID6sssgkfLnh4iIAgN/kYiIAkDU8OHi6+J33nG63rkffkBNjqWlRdWtG5QdOzZrv4YD+8XXqu7dXNo2rH9/cSyQdvMW6HfvdrieuaYGpR+8L/4ddfNIl/NJRETkLQyIiIgCQNTNN0Pevj0AS3BROOcVCLW1duto//wTBc+/IP4dP2lis/dryMoWX6s6d3ZpW4lUijibPOROegyGQ4fs1jHrdMh/ciqqs49a9tGjB8KvvLIZOSYiIvKsEH9ngIiIAKlSiTYvvYhTDzwIGI0o//xzaDdvRvg1QyANC4N+715oN28B6sYXRY8e7TCwqPh2NQqeeUb8u8vhQ/XWsWU8dUp8LYuJcTnf0aNHo3L9emj/3Iba4mLk3HY7IoYMgaqTBrXl5ahc/5M4c55UrUbSa6/yoaxERBRQGBAREQUI9SWXoO2Ct5A/42mYKytRc/IkypZ+XG+96NtuQ+vnZzV7f6Yqrfj8IQCQRUS4nIZEIkHbd95B3uQplnFERiMq169H5fr1duuFJCSg7aKFTXrGERERkS8xICIiCiARQ4Yg48d1KPviC1Rt+h3G3FwIBgNkcXEI690L0bePgXpAf4/syzYYAgBpZKRb6UjDwtDu/SWo3LgRZ1d/B/2+fTCVlkKiVEKRloaIIUMQc+cdkLmZPhERkTcxICIiCjAhcXFIePxxJDz+uMvbRt9yM6JvublJ68oTExrtUueKiMGDETF4sMfSIyIi8gVOqkBEREREREGLAREREREREQWtoO8yp93+FwC43Sdfu/0vGA4dgqpLF4dpNDf9QOHucVwox+9J3ioTZ+k2dX/N3d7ZNt5Kt7E8W7+XaJ3Y5PRd0ZzP8fzzRlPKy9P5M+3ZA23hGaf7bOzc1th+rbx9znC1rvnq++eLfXqCv/LW2H4DNV/+Ts/b6bZknvhN8SZf7Yt1wz1BHxCV1D0AUT3gM7e3Nxw+DFXnzg7TaG76gcLd47hQjt+TvFUmztJt6v6au72zbbyVbmN5tn4v8fLsJqfviuZ8juefN5pSXp7On3HZMpQoVU732di5rbH9Wnn7nOFqXfPV988X+/QEf+Wtsf0Gar78nZ63023JPPGb4k2+2hfrhnuCOiDSbv8Lur//Fl+7cyfTur3u77/rpdHc9AOFu8dxoRy/J3mrTEx79qDaQbpN3Z+z9dzJr+02pf/91CvpWu+ANZa27u+/odyzB0hJaTR9VzTnczw/f7Zl5Ky8PJ0/7fa/YN67Dzon+2zs3NaU/dq+561zhqt1zVPl21i+AXh9n57gr3N0U+pnIObL3+l5O92WzBO/Kd4sx3rnCC/2XGDdcE9QjyGyvZN5/l1NV7dv7G930g8U7h7HhXL8nuStMjEuW9boPhraX1O2aWp+7bZ5913vpPvOO01Oz7ZsPKU5n2O984RtGTkpL1c1lr/GPqPGzm1N2a+r29bLVxO2c7mueah8G8zH+XXTS/v0BH+do12qnwGUL3+n5+10WzJP/KZ4k6/2xbrhvqANiM6/k2m9C+ru9uen0dz0A4W7x3GhHL8neatMrHf7z0+3qftztp47+T1/G3NlpVfS1f39d5PyDADmvfs8Wvea8zk6zJ9NGTkqL0/nr7HPyLZFw5W8ODo2V47D1XJ1p655onybkm9v79MT/HWOdrV+Bkq+/J2et9NtyTz1m+KtcnS0L9OePT7ZT7DXDVcEbUDkyTuZjt5vbvqBwt3juFCO35O8VSbO0m3q/pq7vavLvZFuQ2k3dfumas7n2NzWJHe3ceWuoW2Lhit5aWoLpCvrNCfNhuqDK/lqjC8+U2/w1znanfoZCPnyd3reTrcl89RvirfK0VG63u650NB75FhQBkSevpN5fhqO7rC6kn6gcLecmlu+FyJvlUlD6TZlf83dvil58UW6DaXdWPquaM7n2NTjcDVdV/LXlDzYtmg0NS+Npevpc0Zz6por+WqMLz5Tb/DXOdrd+unvfPk7PW+n25K5Uya+LEdn+/J2zwWrYK4brgrKgMgbdzLtlju5w9rU9AOFu+XU3PK9EHmrTJrbguCpFgh30/JFup5Mp7ktFp7epyvrNqXFxN30m3vedHWZt1v8vL2tv8+D/jpHN6d++jNf/k7P2+m2ZO6UiS/L0Vf7Yt1ovqALiLx1J9OWszusTUk/ULhbTs0t3wuRt8rEnbvTtvtr7vaeyIsv0m0ofVc053NsznF4ouXamo4/PyNPnTM8WScayldjfPGZeoO/ztHNrZ/+zJcnWmndTc/b6bZk7pSJL8vRV/ti3fCMoAuIvHkn05N58DdvjhNqCcfvSd4qk+beafTknUpvteL48m6dJ7b1deuZJ/fRVM35jDxxzvDGcfr6++eJ7b2530ApY2+k4U6anmqldSc9b6fbkrlTJr4sR1/ti3XDM4IqIPL2ncymCvRo3d1yam75Xoi8VSbNrYvNbS2wza+3WnE83RJwfvquaM7n6K3WM0/vw528uLrf5p4zvHWcvv7+ubNPT/DXOTpQfzs9XR7+Pt/zt7U+d85XnihHX+2LdcNzgiog8sWdTG/kxdfcLSdP30W7EHirTAKh/Kx58FYrji9n/PHkNr5qPfPGPpqqOZ9Rc84Z/hpH0px1fZGON/YXqL93/krLHy1J3l63JXOnTHxZjr7aF+uG5wRNQOSrO5nuph8o3C2n5pbvhchbZeLL1oCG6P52PqNiIKZrm743WwN80XrWnPx5Ki/ufkbNOWd48zj98f3z5XnQX+foQP3t9HR59JDJAuJ8z9/W+tw5XzWnHH31mbFueFbQBES+vJPpTvqBwt1y8kcf/EAXDP3EG5pRMRDTFdP38l01b7eeNfaeLzTnM/J2a4+7/PH989fYL29t4400vJGmp8tjvELhlTzwt7U+t8rEjfOVu+Xoq8+MdcOzgiIg8tdsXE1NP1A0p5yaU74XoubWOU+n6y0NzagYiOlaebs1wJutZ95suXZFcz4jb7f2uMsf3z9fnAe9dT7y1n4b4698OdtvTF4eespCPJaelb8+t0Dmbpm4c75ypxx99ZmxbnheUARE/pyNqynpBwp/5CWQjt+TvNUn+EItL3/wdmuA11rPvNxyHez88f3z9ufoqzEKntzWm2l7etv0v9wP+rxV3y7Uc0Mgj7tzZ313t2Xd8DyJIAiCv3Z+8uRJpKSk+Gv3AY1l4xjLxTGWi3MsG8dYLs6xbBxjudTXtm1bTJ8+HXPnzkVubq6/sxNwWGccY7k45s9yCYoWIiIiIiIiIkcYEBERERERUdBiQEREREREREGLAREREREREQUtBkRERERERBS0GBAREREREVHQYkBERERERERBiwEREREREREFLQZEREREREQUtBgQERERERFR0GJAREREREREQYsBERERERERBS0GREREREREFLQYEBERERERUdBiQEREREREREGLAREREREREQUtBkRERERERBS0GBAREREREVHQYkBERERERERBiwEREREREREFLQZEREREREQUtBgQERERERFR0GJAREREREREQYsBERERERERBS0GREREREREFLQYEBERERERUdBiQEREREREREGLAREREREREQUtBkRERERERBS0GBAREREREVHQYkBERERERERBiwEREREREREFLQZEREREREQUtBgQERERERFR0GJAREREREREQYsBERERERERBS2JIAiCP3Z85MgR6HQ6hISE+GP3Aa+2tpZl4wDLxTGWi3MsG8dYLs6xbBxjudR36tQpREdHo6KiAu3bt/d3dgIO64xjLBfHamtrERYWhk6dOvl83379NFgZnGPZOMZycYzl4hzLxjGWi3MsG8dYLvVZg6DIyEg/5yQwsc44xnJxzJ/l4rcWIiIiIiIiIn/jGCIiIiIiIgpaDIiIiIiIiChoMSAiIiIiIqKgxYCIiIiIiIiCFgMiIiIiIiIKWgyIiIiIiIgoaDEgIiIiIiKioMWAiIiIiIiIghYDIiIiIiIiCloMiIiIiIiIKGgxICIiIiIioqDFgIiIiIiIiIJWiLsbCoKAH374AatWrcKhQ4eg0+kQHx+Pfv364a677kKPHj2alTFBEPDrr79izZo12Lt3L8rKyiCXy5GcnIyBAwdi/PjxSEpKatY+vMXbZWM2m/HDDz9gzZo1OHDgAKqqqtCqVSt06dIFI0eOxHXXXQeJROKho/Ecb5eLMy+++CK+/PJLAMCRI0e8so/m8Ha5DB48GHl5eU1ad+vWrYiPj2/W/jzJF3XGYDBgxYoV+Omnn5CTkyN+n/r27YuxY8eiT58+HjgSz/JGueTm5mLIkCFu5SeQvle+qDM7d+7EF198gX/++QclJSVQKpVISUnB4MGDMXbsWERHRzf/QDzMF+Xy66+/YtWqVdi7dy8qKioQGRmJHj16YPTo0W7XLX+w/mZMnDgRkyZNanZ6eXl5+OSTT7Blyxbk5+cjNDQUKSkpGDZsGMaMGQOVSuWBXPuGp8vmfLm5uRgxYgR0Ol1AnVca4+lyyczMxMqVK7F7926UlJRAIpEgMTERAwYMwLhx49CxY0cP5No3PF02mzZtwtdffy2eZ6Kjo9GxY0cMHz4cN954I+RyuVvpSgRBEFzdyGAw4PHHH8emTZscLpfJZHjiiSfw4IMPupWps2fPYvLkyfjjjz+crqNQKDB79myMHDnSrX14iy/K5tFHH8Xff//tdJ0BAwbgrbfeQqtWrdzahzd4u1ycyczMxN133w1rNQ+0E6y3y6WqqgoXX3wxmvo1D6SAyBd15vDhw5g0aRJOnTrlcLlEIsGECRPwxBNPuL0PT/NWubgbEEkkEhw+fNjl7bzBF3Vm3rx5WLp0qdPl8fHxWLRoEXr37u32PjzN2+Wi0+kwdepU/Prrr07XGTRoEObPnx+QwaKtbdu24d5774XZbPbIBdzvv/+OyZMnQ6vVOlyu0WiwePFitG3btln78QVPl835amtr8cADD+DPP/8EEHi/1854slxqamrw9NNP44cffnC6jlQqxeTJkz1+veQNni6bqVOn4qeffnK6TpcuXbBgwQKkpqa6nL5bLUQzZ84UT6wZGRm47bbbEBcXhwMHDmDFihXQ6XSYP38+EhMTcdNNN7mUttlstrvgT0pKwq233or09HRotVps3boVGzZsQE1NDWbMmIHIyEgMHjzYncPwCm+XzcMPP4x//vkHAJCcnIxRo0ahXbt2KCgowHfffYdjx45h+/btePTRR/HZZ59BoVB4+hDd4s1ycUar1WLmzJlNDgb8wdvlcuTIEfH4J0yYgO7duze4flRUlMv78BZvl83p06dxzz33oKysDABw6aWX4j//+Q+io6Oxc+dOrFy5EtXV1Vi8eDFat26NMWPGePLw3OatcmnVqhXefffdJq373nvv4cCBAwDglbvE7vJ2nVm6dKkYDMnlctx2223o3r07zp07h59//hl///03iouL8eCDD+Lbb79Fu3btPHl4bvNmuQiCgEmTJmHr1q0AAKVSidtuuw09e/aEVqvFunXrsH37dmzduhX33XcfvvzySyiVSk8fokfs378fEydOhNls9kh6R44cwaRJk1BdXQ2ZTIZRo0bh4osvhlarxZo1a7Br1y5kZWXhkUcewVdffYXQ0FCP7NcbPF025xMEATNmzBCDoZbC0+XyzDPPiMFQTEwMRo8ejU6dOqG2thZ//fUXvv/+exiNRsyfPx9qtRp33XWXR/brDZ4um6effloMhmJiYnD77bcjPT0dZWVlWL9+PXbv3o1Dhw7hwQcfxDfffIPIyEjXdiC4aOvWrYJGoxE0Go0wfvx4wWAw2C0/evSo0L9/f0Gj0Qj9+/cXKisrXUr/m2++sUtfq9XWW+fXX38VLrroIkGj0QiXX355vTz4i7fLZuXKlWL648aNE/R6vd3ympoa4eGHHxbXWbVqVbOPyRO8XS7OvPDCC+J+rf8CiS/K5YsvvhD3cfz4cU9l3et8UTb33XefuI/333+/3vJdu3YJXbt2FfdRXV3t9vF4ir++S7a+//57MQ8PPPCAYDabPb4Pd3i7bCoqKoRevXoJGo1G6Nevn3D48GG75WazWXjrrbfEPEyZMqXZx+QJ3i6XFStWiOkPHDhQyMrKqrfOokWLxHXeeOONZh2Pt2zatEm4+OKL7X4vFi5c2Kw0x4wZI2g0GqFLly7Cpk2b7JaZzWbhpZdeEve1ZMmSZu3Lm7xRNraqqqqESZMmBfTvtSOeLpdt27aJ6QwfPlwoKSmpt86ePXuE3r17CxqNRujZs6dQVFTUnEPwGk+XzZ9//immM3ToUKG8vNxuudlsFl588cVm7cvlSRU+/vhjAEBISAhefvnlend6MjIy8NxzzwEAKioqsHLlSpfS/+qrrwBY7jK98cYbCAsLq7eOtZ82AJw5cwabN2929TC8wttls2LFCgCW7g3z5s2r1+9YLpfj+eefF/9et26dy8fgDd4uF0cyMzOxfPlyAEBERESz0/MGX5SLtcuBdYxDS+Htsvnrr7+wZcsWAMAtt9zisOtBr169xFahiooKbNu2zeXj8DR/fJdsnTlzBi+99BIAyx26V155JWDGK3q7bP744w/odDoAwP33349OnTrZLZdIJHjsscfQvn17AJbxNLW1tW4diyd5u1ys6QPA/PnzHY5tmDhxIi677DIAwOeff46KigqX9uFNNTU1WLhwISZMmIBz5855LN1du3Zh586dAICRI0fiyiuvtFsukUgwc+ZMZGRkALCUo9Fo9Nj+PcFbZWPrwIEDuP322xvsBhVovFUu1utfwNI119Gwhx49eohduPV6PdauXeux/XuCt8rGev0LAHPmzKnX9VYikWDGjBlQq9UA3Lv+dSkgqqioEJszL7/8cqfdAW644Qbxg1y/fn2T06+qqsLevXsBAJdddlmDYxmuv/568fWePXuavA9v8XbZmEwmJCUlIS0tDX379kXr1q0drpeYmIjY2FgAQH5+viuH4BXeLhdHbLvK3XTTTejSpUuz0vMGX5WLNSDq0KEDpNKWMamkL8rG2iVBqVRi8uTJTtcbNWoUxowZg4cfftjvY/L88V0635w5c8QfuenTpyMuLs6j6bvLF2WTm5srvu7cubPDdaRSKbp16wbAcrFSXl7u0j48zdvlcvr0aZw4cQIA0LdvXwwYMMDpuuPGjQNgKZdffvmlyfvwpj///BNDhw7Fu+++C7PZjLCwMNxzzz0eSdv2guzOO+90uI5UKhWXVVRUIDMz0yP79gRvlg1gGdc2c+ZMjBo1CtnZ2QCAK664Al27dvXYPrzBm+Vi/a5qNJoGr1sC7frXyptlExMTA41Gg3bt2jkdn6lQKJCeng7Avetfl66QduzYIfYFvOSSS5wnKpWiX79+ACwf1tmzZ5uUfnl5Obp37474+HjxoJyxHevgrTsXrvB22chkMixcuBDr16/H559/7nQ9rVaLyspKAAiIixVvl4sjb7zxBnJzcxEfH49nnnnG7XS8yRflIgiC+EOj0WiakVvf8kXZWC/ILrvsMiQkJDhdr3PnznjxxRfxxBNPiBe6/uKP75Ktbdu2iXdx+/bti5tvvtkj6XqCL8omPDxcfH369Gmn61l/iGUymd/H5Hm7XI4dOya+trYAOdOzZ0/xtXUcrL99//33YqDbrVs3rFy5EldffbVH0raOg46KisJFF13kdD3bz8Xaah0IvFk2AFBSUoJvvvkGZrMZcrkcU6ZMwZIlS8Q7/IHKW+Wi1+uRlpaG1q1bi62Gzti2jgTC9a+VN+vMCy+8gP/9738N3kwRBAFnzpwB4N71r0uTKlgvroDGL7A6dOgAwJLBrKws8WTbkHbt2jW5ud72RBwIs9Z4u2ya6p133hGb3W3vIviLr8vFtqvc888/HxB1wxFflEteXh6qqqoAQOzGUlVVhcOHD0Or1SIhIQEajQYymcydQ/Aab5dNUVERSktLAVi6xVkZjUYcPXoUxcXFiI2NRefOnRES4vaTCTzO3+eYN954Q3w9Y8aMZqfnSb4oG9tpqZctW4ZRo0bV6372119/ib0c+vfv7/dJbbxdLrYXYw3dWADsb2La5svfYmNjMXHiRIwZMwYymUw8NzSHyWQSr1EyMjIabJ1PS0uDTCaDyWQKmNkarbxRNucbPHgwnnrqKaSlpXk8bW/xRrmEhobadQtryNGjR8XXgXaN44s648wXX3yBoqIiAO5d/7r0a2/7LJPk5OQG17Xt0pWXl+fRi34AWLVqlfja9s6Tv/irbEwmE0pKSrBv3z589tln2L59OwDLj/GoUaPcTtdTfFkutl3lhg4div/85z+uZdaHfFEutlOWSqVSTJw4ERs3boTJZBLfj46OxtixY/HAAw8EzLMwvF02tj8mbdu2hdFoxJIlS/Dll1+KM84BlrJ54IEHcM899wRE0OjP8+8vv/yC/fv3AwCuueYarz0zzF2+KJtu3brhiiuuwObNm3H8+HGMHTsWTz75JLp27QqtVouNGzfirbfegtlsRnh4OKZNm+bewXiQt8vFdoyvwWBocF3rzRkA4l1cf7vrrrvwwgsvePzcV1JSgpqaGgCNl7tMJkN8fDwKCwsDopu7lbfKxio6OhrfffddQHZpb4i3y6UpAu3618rXZWM2m1FeXo7Dhw9j+fLl+PnnnwFYbu64MyW5SwGR7cVCTExMg+vaRq2eHkC5ZcsW/PbbbwAsU8UOHDjQo+m7w19lM3LkSGRlZYl/SyQS3HXXXZgyZYrf704Cvi0Xa1e5mJgYzJo1y+XtfckX5WIbEL322msO16moqMA777yDzZs34/333xfHn/mTt8umsLBQfK1QKDB27Fjs3r273noVFRV4/fXXsXv3brz11ltuP+zNU/x5/v3vf/8rvr7//vubnZ6n+apsFixYgClTpmDTpk3Yu3cv/u///q/eOn369MGzzz4bEGMhvF0uts/OsU7D7szBgwfF19Zu3f7W2GMI3OVKuQOW1rPCwsKAmmzCW2VjFR4e3uKCIcD75dKYrKwscfIFhUKB6667zq/5seXrsnnkkUfEWMBq2LBhmDlzplvdlV0aQ2R7B6ix5wjYXow3dufIFQUFBZg+fbr498MPPxwQd7b9VTYFBQV2fwuCgMzMTKcP4PM1X5WLbVe55557LiAu7Bvii3KxDYjkcjkefPBBrFu3Dvv27cOWLVvw8ssvixOX7N27F5MnT/baMyZc4e2ysX1A4rx587B792507NgR77//Pnbu3Im///4bb775Jtq0aQMA+Pnnn7FgwQIXjsA7/HWOOXz4sDgeonfv3gH1wFErX5WNWq3Ga6+9hiuuuMLhcolEgrS0tIDpxuLtctFoNOK59qeffhK7qzhi2x3I2npyodLr9eLrpjxzybqOJ6+V6MJTWVmJxx9/XBwWcddddyExMdHPufKf869/AWD37t348ccf3Xr+pEsBke0Uoo21Ptgu99TUo2VlZbj33nvFPokDBw4Up9/2N3+UTW1tLf7v//4Pb775Jl555RUMHz4cUqkUR48exZQpUwLiIs4X5WLbVW7IkCEYNmyY6xn1MV+Ui7VbilKpxH//+188+eSTyMjIgEKhQEJCAkaPHo1Vq1aJXWUyMzMDYqp2b5eNdepkwDI4vkuXLlixYgWuuuoqqNVqREZGYtiwYVi+fLl4sfff//7XrvuRP/jr/PvZZ5+JrwOxdQjwXdmsW7cOV111FTZv3ozExERMnDgRb731Fl5++WXxTu2qVaswYsQIcSyRP3m7XKRSqfhgSL1ej8cee8zhIO/ly5fbTascKFO1e4ttt+Sm9NSwrhMI07RTYNLr9ZgwYQKOHz8OwHIzoqEZUoPBsGHD8Prrr2Pu3Lm4/fbboVAokJeXh9mzZ9s1nDSVSwGRbUtMY/Pl294B8kTXreLiYowfP16sDCkpKXjjjTcC5sTqj7IJCQnBpEmTMGzYMNx6662YP38+Pv74Y/Fu0+LFi7Fjxw630/cEX5SLtatcVFQUXnjhBZfz6A++KJcVK1bgn3/+wU8//YSLL77Y4TqJiYl49tlnxb+//vrrJqfvLd4um/PvHM2ePdtuBjGrNm3a4LHHHgNguVD57rvvmpS+t/jjHKPVasUpypOSkjB48GC30/ImX5TNr7/+iqlTp8JgMGDAgAFYt24dJk2ahBtuuAGjR4/GwoUL8dFHH0GlUkGr1eL+++9HcXGx6wfjQb4ol/vvv1+csGHXrl244YYb8Pbbb2PdunX4+uuv8cADD+CFF15AfHy8ONtaIHTn9ibbVqGmPFvIWvb+7pZLgUmr1eLBBx8Ur+diYmKwaNGiJrU+XsgefPBB3HjjjRg5ciReeuklrFy5UmydX7NmDb7//nuX0nMpILIdQFldXd3gurYn1+Z+aKdOncJdd90lzkzTpk0bfPLJJwHVLcpfZXO+Sy+9VLyIAyyzIfmTt8vFtqvcjBkzGp3pKFD4qr6Eh4eLXb+cGTx4sBgQ7Nq1y+/d5rxdNqGhoeLr1NTUBvs92/bPtj5k0V/8cY7ZsmWLuK9rr702YJ9l5e2yEQQBr776KkwmE6KiorBw4UKHQfSgQYPEyRTOnj1r99BSf/BFnVGpVPjoo4/EZzMVFxfjvffew+TJk/Hcc89h8+bNSE5OxocffihOqRwZGenKYbQ4tlNHN1butusEQvd/CixlZWW4++678ddffwGwfHeWLl2K1NRU/2YsAHXu3NnupnhDj6hxxKVfN9uTWGOD/2yXNydw2bt3L8aMGYOTJ08CsMzY8vnnnzc6c4uv+aNsnBk9erR44eLvizhvlottV7nLL78ct9xyi7vZ9LlAqi8ymQwpKSkALBdF/h7Y6+2ysb2QbWyq19jYWPGOU0PjI3zBH3Vm48aN4utAGrx7Pm+XzaFDh8RnDw0bNqzBMUK33XabGHT//vvvTUrfW3xVZxITE/HNN9/g2WefRa9evaBWqxEWFoYuXbrgySefxJo1a9ClSxdxsoFAeEaeN0VERIivm/JMJ+s6gXSTl/zv1KlTGDNmjNj9Njo6Gp988klATNgSqK677jrx/HzgwIEm3ZCwcmmWOduItKCgwOlTrwH7mZySkpJc2Y1oy5YteOyxx8Q+/x06dMDHH38ckIPIfF02DYmKikKrVq1QXFxsN9uNP3izXPbv3y8+BGzLli3o1KlTo9tY10lOTra72PO1QKovQGB11fB22djOjNUU1u49/m4583WdMZlM4gV9YmJiQE6mYOXtsjlx4oT4urHn+SgUCmRkZNidn/zFl3VGLpdj3LhxGDdunMPl1dXVYjm2pGfOuCMhIQFqtRparbbRqbRNJpPYtdJb53dqeQ4cOIAHHnhAHDOfkJCApUuXtqiHrPuDVCpFamoqdu/eDZPJhIqKiibHDC61ENk+PbexB6tZl0skEvGhkK7YuHEjHn74YTEY6tmzJ5YtWxaQwRDg/bI5duwYnn76aYwdO9ZuDnpnrFGxv5/67Ms605J4u1xOnz6NTz/9FG+++WaTZhy0TsAgl8vdmq7Sk7xdNrY/KI1NlGA9oQLw+7nH19+l/fv3i8c+ePDggBmv6Ygvy6YpdxytLfT+fn5VIJ1/d+7cKU420K1bN4+nH2isZW/7EHlHjh8/LpYLL3YJgDilvzUYSk1NxZdffhm09aOkpASzZs3C//3f/2Hx4sWNrm97jnblGtilFqLevXtDLpfDaDRi+/bt4uwy5zOZTOI0rZ07d3a5v/Bff/1lN7XgwIED8c4779j1hw403i4bk8mEb7/9FoCly8+tt97qdN2cnBxxph9/34nzZrl07NgR7777bqPrLViwQPyxt67v777a3q4vRUVFeOWVVwBYLmavuuoqp+vm5OSI01d2797d7xdx3i6bmJgYdOzYEdnZ2cjOzkZhYaHdQyltHThwQBxb4e8fI1+df6127dolvg6kh/854u2ysQ2GrQ+odaa2tlac/Mffd/x9UWdeeOEFnDhxAhkZGXjuueecrmc7y9ygQYOanH5L1a9fP+zduxelpaXIzs52GmRmZmbabUPB7dixY7j//vvFZ3V16dIFS5cuRatWrfycM/9RqVT4+uuvIQgCysrK8PDDDztdt6qqSnz4ekJCgsOxns64PIbokksuAWBpwXHWFLx27Vqxq9bQoUNd2QVKS0vx+OOPixchV199NZYsWRLQwRDg/bLp2LGjODh+y5YtYn92Rz788EPx9TXXXNPkfXiDN8slNjYW11xzTaP/bB+MZ33P3z/I3q4v3bp1E08EW7ZsabDbhu0dl5tuuqnJ+/AWX5xnrFOzC4Jg9305n+2kJK7uw9N8US629u3bJ74O9D7r3i6b7t27i9+nX3/91eHzL6zWrFmDqqoqAPD7Q8N9UWeOHTuGbdu2YfXq1XZT2tsqLCzE6tWrAQC9evVChw4dXNpHS2Q75s7Z4G6TyYQvv/wSgGXckb9/l8i/qqurMWnSJHFMWffu3fHZZ58FdTAEWBoBrF22s7KyGpw9+dNPPxUbU1y9/nV5yqC7774bgGUqySlTpognfqujR49izpw5ACxNVaNHj3Yp/dmzZ4sn5h49euDtt99uMVN0erNsJBKJ+FT02tpaTJs2rV76gKUyWLvUxcfH44477nDnUDzK23WmpfJmuSiVStx2221i+k8++aTD+rJ06VKsWbMGgKVZ/uabb3bnUDzO23VmzJgx4gDm5cuXi2Vg67vvvhPf79OnT0C0kvjyu2S9yyaXy+26XgUqb5aNQqEQn3mn0+nwxBNPOByfuXPnTrFlVi6Xi+dsf/J2nbn++usBWCa5efPNN+stLy0txaOPPio+dNR2FtQLWc+ePcWLuJUrV+LHH3+0Wy4IAl5++WWxNfHOO+/0e88F8q+FCxeKXSzbtm2LDz/88IKfkbGp7rnnHvH1zJkzHU5y9NNPP+G9994DYGlVeuCBB1zah0td5gBLU/d1112Hn376Cbt27cJNN92EO+64A23atMHBgwexfPly8UnwTz31VL1ZU7Zv347x48cDAPr372935+TYsWNYv369+Pf111+PLVu2NJqn6Ohop89Z8SVvlg0AjB07Fr/88gt27NiBXbt2Yfjw4bj99tuRmpqKsrIyrF+/XpyaUS6X46233vL7GCLA++XSUnm7XB555BH89ttvyMnJwc6dOzFs2DCxvpSXl2PdunXinRa1Wo033ngjYJ5r4O2yiYmJwUsvvYTHH38cJpMJTz31FH788UcMHjwYCoUCv/32m3guUiqVmD17tg+OunG+/C6dOnUKgOX86u9ulE3h7bKZMGECNm/ejIMHD2L37t244YYbcPPNN6Nz584wmUzIzMzE2rVrxYdrzpw5MyBmQ/V2uYwePRqfffYZTpw4gc8//xyHDh3C0KFDERUVhezsbHzzzTfiWIixY8f6vdXMU3JzczFkyBAAzifpee6553D77bfDaDRi8uTJ2LhxIwYNGgSDwYDvvvtOnAU2NTUVDz30kE/z701NKZtg1FC5VFRUiK2FgKWF8Z9//mk0TZVKdUG0LDZWZ6699loMGzYMa9euxYkTJzBixAiMGTMGnTp1QlVVFX777TdxG4lEgjlz5rjcZdnlgAgA5s6dC61Wi61btyI3Nxevv/663XKJRIKJEydizJgxLqX73Xff2T00cd68eU3aLpAukr1VNoAlyFm8eDEmT56MrVu3oqCgAAsWLKi3XqtWrbBgwYKA6o/szXJpybxZLhEREfjkk08wceJE7N+/H4WFhXj77bfrrdemTRvMnz+/wefx+IO368x//vMfzJ8/H88++6x4Qv3tt9/s1omOjsbixYsDqouPL75LVVVVYvcn2ymEA503yyY0NBRLly7F5MmTkZmZifLycofPGZLL5XjmmWcConXeypvlolAo8P777+O+++5Dbm4uduzYUa9Li0Qiwd13342nnnqqWcfR0nTt2hULFizAtGnToNPp8P3339d7WGRKSgo++uijgLh5Sf6zfv16uy6nS5cuxdKlSxvdLpgCztdeew1yuRzfffcdKioqsGTJknrrqNVqzJkzx63u4m4FRKGhofjoo4/wv//9D9999x0OHTqEyspKREdHo2/fvhg/fjz69u3rcrpZWVnuZCegeKtsrCIjI/HRRx/hl19+wapVq7Bv3z6cPXsWarUa6enpGDJkCO64446AO7l6u1xaKm+XS5s2bfDVV1/hhx9+wNq1a3HgwAGcO3cO4eHhSE1NxbXXXovbb7894OoL4Js6M3ToUPTt2xdffPEFNm7ciLy8PEgkErRr1w7XXHMN7rzzzoB7NogvysXaYgC0rIdoertsYmNj8d///he//vorvv/+e+zZswelpaVQKBRITk7GwIEDMW7cuIBoGbLl7XJJTU3FmjVrsGzZMmzYsAE5OTkwGo1ISEhA//79ceedd6JHjx4ePKKW45prrsHatWvxySefYPPmzSgsLIREIkFaWhquu+46jBs3LiDPv+RbF8L1r7cpFArMnTsXI0eOxNdff41//vkHZWVlUKlUaN++Pa666iqMHTvW7d9siWDbJENERERERBREXJ5UgYiIiIiI6ELBgIiIiIiIiIIWAyIiIiIiIgpaDIiIiIiIiChoMSAiIiIiIqKgxYCIiIiIiIiCFgMiIiIiIiIKWgyIiIiIiIgoaDEgIiIiIiKioMWAiIiIiIiIghYDIgpKJpPJ31kgcshsNvs7C0REREGFARF51eDBg9GpU6cG/3Xt2hX9+/fHTTfdhOeffx7Hjx/3Wn6Ki4sxefJk/PPPP/WWWfOzaNEir+3fG37//XeMGzcO/fv3R/fu3XH11Vfju+++83e2vM76eX377bd2748bNw6dOnXCuHHj/JQz95w8eRL33HMP8vPz7d7Pzc11eqxETXEh1aF58+ahU6dO2Lp1q/jet99+2+BvTLdu3TBgwACMHDkSr7zyCg4cOOA0/bKyMgwcOBCdOnVCjx49cPTo0UbzVFNTg1GjRqFTp07o379/ve9wUy1cuBCdOnXCp59+6vK2Bw4cwGuvvYabbroJAwYMQLdu3TBo0CDceuuteOONN3D48GGX0jtx4gQWLFiAMWPGYMCAAejatSv69OmDESNG4Pnnn8eOHTsaTWPRokXiZ5Cbm+vS/o8dO4auXbvisccec2k7IncxICK/q62txdmzZ3H48GGsWLECN954o1d+tMvKynDDDTdg3bp1EATB4+n7w59//omHHnoIf/31F86ePYuamhrk5+cjJibG31kjFxw+fBgjRozAn3/+6e+sEAWsv//+G5988gmuuuoqDBo0qMnbGY1GVFRU4NChQ/j0009x6623YtasWaipqam3bmxsLF555RUAQHV1NaZOnepwPVvz5s3Dvn37AACvvPIKkpKSXDiqf/3+++8AgCuvvLLJ2xiNRjz99NO49dZb8cknn+Dw4cOoqKiA0WhEcXEx9u/fjw8//BA333wzZs6cCaPR2GB6586dwzPPPIOhQ4di8eLF2LVrFyoqKlBbWwutVousrCysWLECd911F8aNG4cTJ064dayNycjIwB133IGffvopKG7wkf+F+DsDFBz69u2LDz/80OGympoa5OXlYd26dfj0009hNBoxa9YsdOnSBV26dPFYHnQ6Hc6dO+d0efv27QEAUVFRHtunt3377bcQBAHh4eFYtGgRunTpgpqamqAOiBISEtC+fXskJCT4OytNdvbsWVRXVztcJpfLxboZHh7uy2wRBQzr74JEIsG0adOcrvfBBx/g4osvFv8WBAE1NTUoKyvD/v378d///heHDh3CV199hfLycrz99tuQSu3vDV955ZUYO3Ysli1bhkOHDmHBggV46qmnHO5vw4YN+PzzzwFYWqevueYat46vtLQUBw4cQPv27ZGamtrk7WbOnIk1a9ZAKpXi1ltvxYgRI5Ceno7Q0FCcO3cO+/fvx2effYYdO3bgm2++gV6vx5tvvukwrYKCAtxzzz3IyckBAHTt2hVjxoxB37590apVK5SWliI7OxsrVqzAtm3b8Ndff+GWW27Be++9h0suucSt427Io48+itWrV+PVV1/FlVdeGdS/a+R9DIjIJ2QyGdRqtcNlarUaMTEx6NatGzQaDaZPnw6j0YiPP/4Yr7/+us/y+PPPP/tsX55SUlICABg4cCAuu+wyP+cmMMyfP9/fWfCoxMTEFlk3iTxp2bJlOH78OG666SZ06NDB6Xoqlcrhb01sbCw6dOiAYcOG4amnnsK6deuwYcMGvP/++3j44YfrrT9t2jRkZmbi6NGj+OSTT3DllVdiwIABduucPn0azzzzDABL8OAsaGqKzZs3QxAEl1qHDhw4gDVr1gAAnn32Wdx11112yyMiIpCcnIxrr70WEyZMwKZNm7B27VrceeeddkEjABgMBjz00EPIycmBVCrF1KlTce+990IikYjrREdHIyMjA9dffz02bNiAadOmQavV4pFHHsHKlSuRkZHh9vE7EhMTg3HjxmHx4sVYuHAhnn/+eY+mT2SLXeYooIwcORKtW7cGgCb1UQ521skhwsLC/JwTIiLv0Ol0+OCDDwAA48ePb1Zacrkcr732GtLS0gAAS5cuxdmzZ+utp1KpMH/+fMjlcpjNZkyfPt2uh0FNTQ0ef/xxVFZWQq1WY8GCBVAoFG7na/PmzQCAyy+/vMnb/PbbbwAsgc8dd9zhdD2JRII5c+aIwc26devqrfPBBx/gyJEjAIAnn3wS9913n10wdL5rr70W77zzDgBAq9Vi1qxZTc63K+68806EhIRg5cqVbo/NImoKBkQUcBITEwFYJkBwpKamBitXrsSECRNwxRVXoEePHujVqxeuvvpqPP7443aDba06deqEIUOGiH+PHz8enTp1wowZM+zWaWhShdOnT2POnDm44YYb0KtXL/Tu3RvDhg3DK6+8goKCguYcMkpKSvDWW29h5MiR6N27N3r06IFrr70Ws2bNcjio1zpZxV9//QUAWL16tcuTQlgnH1i0aBGMRiM++eQT3HTTTejVqxcGDBiAe++9F5mZmeL62dnZmDp1KgYNGoRu3bphyJAhmDt3LrRardN96HQ6fPjhh7jtttvQr18/dO/eHUOGDMGzzz6LY8eONZi/mpoafP311xgzZgz69++Pvn37YuzYsY22ljQ2qUJeXh7mz5+PUaNG4ZJLLkHXrl3Rr18/3HTTTZg7dy4KCwsdbmct3+3bt6OyshJvvfUWhg4dip49e6Jfv34YP3481q5d22Dezmcd7G57kTdkyBC7z7GhAfHWevDtt99Cq9Vi4cKFGDp0KHr06IGBAwdi4sSJOHjwoLj+zp078cgjj+DSSy9F9+7dxXECDY0rKC8vx4IFC3DTTTehT58+6NmzJ66//nrMmTPH7Xo/Y8YMdOrUCZMnT4YgCPjiiy/Eujdo0CDce++92LhxY6Pp7Nq1C1OnTsVVV12F7t27o3///hg7diyWL1/u9JisZbZy5Ups3boVN954I7p164aBAwfiySefbHB/X331lfhZODp2o9GI3r17o1OnTrj99tsdpvHLL7+Ik8mcfyHuzjnGOnD9jjvuQFlZGR5//HH07t0bffr0wS233CJe5DZk/fr1uOiii8R0HH2nfV3WtlavXo2ysjJ07twZ3bp1a/J2ziiVSkycOBEAUFlZ6XSMSufOnTFlyhQAlu5kL730krjs9ddfFydnmD17ttit1R0mkwl//PEHVCpVvVaohlh/I41GY6OzU8bFxWHIkCHo3bs3YmNj7ZZptVpxIocuXbrgvvvua9L+L7/8cowYMQKA5Qam9ffIkxISEnDFFVfAaDS6NdkEUVOxyxwFlJqaGpw8eRIAHI4BOX36NO6//36HAzn1ej3y8/Oxfv16TJw4EZMmTfJYvlauXInZs2fXG+dx9OhRHD16FCtWrMCcOXPEHwdXbNq0CdOmTas3vunkyZM4efIkvvnmG/GOnTfo9XqMHz8eO3futHvvjz/+QGZmJt59912YTCY8+eSTMBgM4jq5ubn4+OOPsXPnTixbtgxyudwu3SNHjmDChAkOZ01buXIlvv32Wzz99NMOA5eysjJMmDABe/bssXv/77//xt9//43777/frWNduXIlXnzxxXoXcOfOncO5c+dw+PBhfPPNN/j0009x0UUXOUwjNzcX06dPt7tANRgM2L59O7Zv347MzEzMnj3brfy5q6SkBLfeeqvY9x+wDAj/+eefsXXrVixbtgx79+7Fyy+/bDfl/PHjx7FgwQIcOnQICxcurJduZmYmHnvssXoX7jk5OcjJycHXX3+NefPm4brrrnM771OnTsUPP/wg/q3X61FcXIw//vgDt9xyC15++WXIZDK7bcxmM+bNm4dPPvnE7v2amhqxjnz99ddYsmSJeIPlfLt377arCyUlJU679VrZdmfatm0bbrnlFrvle/bsgU6nAwDs378fOp2uXuuttSWgT58+duMVm3uOqampwf333283g9qpU6eQkpIidq115Pfff8fUqVNhMpnQu3dvfPjhh3bl4K+ytrVixQoAllYJTxkyZAgUCgVqamqwbds2/N///Z/D9e655x5s3rwZ27Ztw//+9z9cf/31CAsLE8cN3XbbbRg2bFiz8rJr1y6cPXsWV1xxBVQqVZO3a9u2LQDL+WfRokWYPHlyg+u/++67Dt/fuHEjqqqqAAB33XVXgy1D57vrrrvwv//9DwCwZs0a9O/fv8nbNtV1112HjRs3YvXq1ZgyZQqUSqXH90HEFiIKKJ9++ikqKioAwK5FB7DcRZs4cSJOnDiBsLAwPP300/jxxx+RmZmJH374AU8//TSio6MBAO+99x5Onz4tbrtz5067u/cffPABdu7caXfHz5mffvoJzz33HKqrq9G+fXvMnz8fW7ZswdatW/HGG2+gbdu2qK6uxrRp08RZgppq165dmDRpEs6dO4e4uDi89NJL+O2337Bt2zYsXrwYnTt3hslkwrx587B8+XJxu7Vr12Lnzp3o27cvAGDEiBHYuXMndu7ciYceesilPCxbtgw7d+7EmDFj8MMPP2Dz5s2YPXs2FAoFTCYTXnzxRUydOhVt27bFkiVLsG3bNqxduxbXX389AMvFzvr16+3SLCoqEqeQjo2NxfPPP4+NGzciMzMTn332GQYOHAiTyYSXX37ZYavK448/jj179kAqleKhhx7CTz/9hG3btuG9995Deno6PvroI5eOEQD27t2L5557DkajEd26dcP777+P3377DX/88QdWrFiBkSNHArAER6+99prTdGbPni3eid+wYYP4WVkvTr7++mv8/fffTcpTcnIydu7cKXYHAv79bF35HN955x2cOnUKDz/8MDZs2IDffvsNU6ZMgUQigV6vx5QpUzB79mz06NEDn332GTIzM7Fq1Srx4uWnn36yC4gBICsrCw899BDOnj2Ltm3bYt68eeKF4QcffIBu3brBYDBgypQpDqexb4pNmzbhhx9+QJcuXfDJJ58gMzMTK1euFAOPb7/9VuyWY2vhwoXiBfq1116LL7/8Etu3b8cvv/yCGTNmIDw8HAcPHsRDDz3kdLKKb775Bq1atcLHH3+MP//8E++9916jU7W3bt0anTt3BgCHMwLatqjW1tbWK1Pg34Do6quvFt/zxDlm//79OHDgAB5//HFs2bIFa9aswezZsxu8wP7rr7/w2GOPwWg0omfPnvjoo4/qTdzhr7K2Onr0KLKysgC41p2sMaGhoeKYl127djldTyKRYO7cueJvy+zZszFjxgwIggCNRoOZM2c2Oy/WOnHFFVe4tN2IESPEG1FLlizByJEj8cknnzTa+n4+23rrSgsVAPTq1QtxcXEAgO3bt7u0bVNdfvnlkEgkOHv2LGfiJO8RiLzo6quvFjQajXDHHXcIVVVV9f5VVlYK+fn5wtatW4UZM2YInTp1EjQajTBo0CChpKTELq1NmzYJGo1G0Gg0wpo1axzu7+effxbXWbFihd2y06dPi8syMzPrbWtdtnDhQvG96upq4fLLLxc0Go1w7bXXCuXl5fW2Ky0tFYYMGSJoNBrhiiuuEKqrq5tcPrfccoug0WiEAQMGCKdPn663XKfTCaNHjxY0Go3Qs2dPobS01G752LFjBY1GI0yfPr3J+zx/W41GI7zyyiv1lr/wwgvi8kGDBtU79pqaGrFsZsyYYbds2rRpgkajEfr16yecPHmyXtomk0l4+OGHBY1GI1x66aWCwWAQl23YsEHc77Jly+ptW15eLpa3RqMRVq1a5fC4xo4da/f+Y489Jmg0GuGSSy4RKioqHJbJI488Img0GqFz586CXq+3W2bdn0ajETZu3Fhv2+zsbHH5Sy+95DB9ZzIzM8Vtz68HtvX2/GO1fr80Go3w6aef1kv3gQceEJffeOON9epmRUWF0L17d0Gj0QiLFi2yW3bnnXcKGo1GGDx4sFBWVlYvbYPBIIwaNUrQaDTCiBEjXDre6dOni/kaPny4UFVVZbfcZDIJDz30kKDRaIQePXoIZ86cEZfl5OQInTt3FjQajTB79myH6e/Zs0fo0qWLoNFohI8//thumW2Z/fHHHy7lWxAE4c033xQ0Go0wcODAesusZdavXz9Bo9EIb775pt3yrKwscd/Hjx8XBKH555iFCxeKaT755JMO8+yoDu3Zs0fo3bu3oNFohFtvvVU4d+5cve38XdaCIAhLly4VNBqN0K1bN8FoNDpcZ9WqVQ2e252xnoM0Go3TtK1+/PFHu3NAz549haNHj7p0LM7ceOONgkajcXiubMzq1avFz8j23yWXXCJMnDhR+PTTT4Vjx441mMbtt98ulrE77rjjDkGj0QidOnWyO5fb1k1Hv2+uuOaaawSNRiO8+OKLzUqHyBm2EJFP/PPPP+jTp0+9f3379sVVV12Fe++9V5xC+qKLLsLnn3+OVq1a2aWhVqsxfvx4DBs2DDfccIPD/dje3SorK2t2vn///XecOXMGgGXcg/Uuoa3Y2FhxLFJhYaE40LUxBw8exP79+wEAEydOFFsYbIWGhoqtWHq93mvPY3jggQfqvWc7C9Ho0aPrHbtcLhf78xcVFYnvnz17Vhy0O3bsWId966VSKaZPnw7AMt3sr7/+Ki6zdr9ITU2tN2sSYJnpqLGuIY706dMHo0aNwqOPPup0anVri4nZbHY40BqwjCWyvbtv1aFDB7Rr1w4AXH4IYXOp1Wrceeed9d63/QzHjx9fb9B3VFSUOLjc9jPMzs4WJzV55JFHHE53q1Qqxc/hyJEj9bo3NtWzzz5br/uUVCrFzJkzIZFIYDAY7MaNffXVVzCbzQgNDXVaD3r06CGeI77++muH60RHR+PSSy91Ob/W1qvi4mJkZ2eL7+t0OrFV0zrA/fw75taWgNTUVLHcPXmOGTp0aJOO4ciRI3jggQeg1WrRtWtXfPzxx4iIiKi3nr/LGoBYrzIyMhAS4tle/rbdGa09E5y5/vrr7bqDDR061COzqp05cwaHDx9GamqqW+OQRo4ciS+++KLeIyrKysqwYcMGzJkzB0OHDsXw4cMdTqYAWMYJAkBkZKTrBwCILUSCIDT4aIvm6NixIwC4fZ4hagwDIgoIycnJuOOOO/Dhhx/i22+/dfgchosvvhgzZ87Em2++6fCH8ezZs3aDOm3HSrjLml5oaGiD3RmuvvpqhIaGAkCTu0vZ5rWhMRidO3cWL568MWg1OTlZ/EGzZTvw1tl4Gmv3GtsHF+7atUscK9C5c2dotVqH/+Li4hAfHw8Adl2urBeRDXWPGTx4cL1nhzTm//7v/zBnzhyMHTvW4fITJ07YdTWpra11uF7Pnj2d7sMaxOv1epfy1lxdu3Z1+J2w/Qy7du3qcFtHn6FtPdNoNE4/w86dO4vje9zpNhcfH++0i067du3Ei6Bt27bVy1t6ejoAOM2b9XM6fvy4eMFnq3Pnzi6NlbDq1auXGCDa5mvHjh0wGo3o1KmTWHf3799vVxccdZfz5DnG2ffU1smTJ3HfffehoqICYWFhWLp0qdMLYX+XtTVN2zx4km2dbyx/mZmZdjOffv/999i9e3ez8+DO7HLn69OnD7777jusWrUKjzzyCHr16lVvTGd2djYmT56MiRMnNvpwVlfZlt354/08xfr5246TJPIkTqpAPtG/f39xECpg+SHKysrC22+/jc2bN6OoqAjR0dEYNGhQoz9MZrMZu3btwr59+3DixAmcPn0aOTk5yM/PhyAI4nq2r91lnXEsNTW1wRO9TCZDamoqDh061OSpQa1pR0ZGioGBMxkZGcjJyWn2bHaOOHvYnW3A4eju8fnrWNmO3WrqxBbW49Lr9eKd2pSUFKfrh4aGok2bNsjLy2tS+rYqKyvx559/IisrC6dOncLp06dx7Nixenc2ndWf82dosmVtgfFE3XNFUz5DZw91bewzHDVqVJPy4E7dtAY8zqSkpCArK8tu5j9r69uBAwfQp0+fJu2nsLCwXhk19Dk2RCqV4vLLL8f333+PP//8U5wh0BocDRgwAN27dxcH7O/evRuXXnoptFqtGDTaBkSePMc05Zjef/99sX7qdDr8+OOPDlsXAf+XtTU9wDsPzLZOJAA4P8cBltaWadOmwWw2Izk5GYWFhaitrcW0adPw3XffuTRBxPmsAZErzx9yplu3bujWrRsef/xx6HQ67N69G3/++Sd+/vlncSKin3/+GXPnzsWzzz4rbhcTE4MTJ0643bpjbU2XSqUOWzg9wfr5a7VaVFVV8SHV5HFsISK/UCgU4sD24cOHw2g0YvHixXjxxRcb3C4zMxPXX3897rzzTrz66qtYvnw5tm7diry8PCQnJ2PMmDEezaf1B7Mpz/mx3r21zjLlz7Rd4elnGNleZLi6je0PcmOzLTV0AeOI2WzG22+/jYEDB+Kxxx7DO++8g++//x67du2CXq9H7969m3RR4uluO54QSJ+hKxr7DK11oLKysln7cbRNc2aquuqqqwBYWmqsLdHWgekDBgyAUqlEr169APzb4rlt2zYYjUZERkaKk6HY5s0T54GmHJMgCMjIyBC7u86fP1/ssne+QChr67F64wLYGvgnJiY6fYaQIAh46qmnUFRUBIlEgnnz5uGee+4BYJnF7+WXX3Z7/0ajEX/++SdCQ0M9PjtbWFgYLrvsMkydOhXr16/H66+/LtafFStW2M08aG19qampcTiDa2Osj4ZISkpyueW+qWzPFQ096oHIXYH3y05BRSqV4pVXXsGRI0eQnZ2NFStWICMjw+HD9/bu3Yv7778fRqMRarUa//nPf9C9e3d06NABGo0GsbGxqK2tFado9QTrRUpTAhHrSdr6o+PPtP3JNo/r1q1zqZ+97d3Fxrqd2XZ3aYpXX30Vn332GQDLBYD1GSkZGRno2LEjFAoFVq5c6fJMgRci22B07969XpvmtrHP0PrdsG1xUKlUqKqqwg033IC33nrLK/lqzKBBgyCTyVBVVYW9e/eKLTcymQz9+vUDYAmM/vrrL7Hb2ZYtWwBYZhKzDap9fR7o0KEDPv30U3Gq9qqqKjz//PNYsmRJvXUDoaytPQY8faFdVlaGU6dOAbCMg3Jm6dKl4md399134+KLL0b37t2xefNmZGVl4dtvv8WVV14pzrrpin/++QdVVVW46qqrXP6OHTt2DGvXrkVJSQkeffRRp1OeA5YyvPHGG1FZWYmXXnoJRqMRBw8eFLtoXnLJJVi1ahUAy+yJjrqsO5OVlSU+D+myyy5z6RhcYdsF3t3ul0QNYQsR+Z1SqcQbb7wh9nl+/fXXxWlWbS1YsABGoxERERFYs2YN5s6di7Fjx+KSSy4Ru2Q46r/eHElJSQAs40saGpNUW1sr3llLTk52Ke1z5845fQitlXVsi3WbQNamTRvxdWNd2s7vWqZUKsVxONaxA46YTCaXumgVFBRg2bJlAID//Oc/+OGHHzBt2jTceOON6Nq1q3h32NP1p6WyrWeNTRDRnO6Btl3zHLGOF7D9Tlnz5mrd8qSoqCj07t0bgOUCcseOHRAEAV26dBHvZF9yySUALAGlwWAQu0ZZW5esvHmOceS+++5DXFwcOnfuLN54+u233xwOuA+EsrZ2R/P0uDzb6f4HDx7scJ29e/diwYIFACwPLLU+pFWpVOL1118Xf7NmzZrl9IHODXF3um3A8t1599138dVXXzV5umvbVijbKdKHDBki1tvly5c7fcjrF198gX379tm9Z/t8KneCwqayPTc3p4sikTMMiCggdO7cWXzYZk1NDWbNmlXvpGx9VsRll10mzuZ1PttBzudv785dJessXXq9XvzxcmTTpk3iD4z1QqmpaQOW55A4c+jQIfFhtU3tx+9Pffv2Fe/m2s4ed768vDz07t0b11xzjdhyA/x7cbBp0yanF4h//fWXSxdIe/bsEevDqFGjnI7VsK0/vhwHFGh3PG3rZkOf4c6dO9GzZ09cd911+PHHH13ez7Fjx5wGXDk5OWJQbBtEWPN24MCBBi9CZ82ahQEDBoitIJ5mzdO2bdvEsUG2F5w9e/ZEaGgojEYjVq5cifz8fISEhNS7+PXmOaYxkyZNEm9gvPzyy/VmWguEsm7dujUAOO3W5w6dTodPP/0UgCW4dTSpTWVlJSZPngyj0QiVSoX58+fbdavr3LkznnjiCQCWMTRPPfWU00DCmeYERD169BBbGr/88ssmna+svyOApaXQSq1W4+677wZgafF5//3362177tw5vPrqqxg1ahTuvPNOnDhxAn/88QfWrFkDwHLeHzhwoMvH0VTWuhkVFcWAiLyCAREFjEceeURsqt+1a5fdg0iBf2evycnJcfjDU1BQgDfeeEP8+/yZdGy7qTR1lp2rr75anPDgtddeczgVc3l5OebOnQvAMnjY2d3G83Xr1k2c+eudd95xeBfWYDCI46qUSiVuvPHGJqXtT3FxceKg8VWrVjmcfcxsNuPVV1+FXq/H6dOnxfEMAHDrrbcCsHye7733Xr1tq6ur7T7nprD97K393c+3atUqu4f+udolrzlsAzRPzwDljh49eojT+H744YcOxxUYDAa89tprqK6uRl5eXoPdjpwRBAHz5s2rdzFXW1srjs2Ijo62m4TgtttuE9d58cUXHQbNe/bswerVq1FRUYHo6GivjD+xBkS7d+8Wu1RZW4UAy7T01rFC1nrcp0+fepMDePMc05iwsDBxcH1paWm9BxIHQllbu9w21prYVDU1NXj66afF9CZNmuTwAnvmzJlisD5jxgyHXX/vvfdeMQjevn27Sw+Mzs/PR3Z2NtLS0pze4GtIbGwshg8fDsDye/niiy82eM4qLy8XW7sGDBggzlxq9eCDD4q/R2+//TY++OADu++l0WjEmDFjEBUVhX/++QcjR47EpEmTYDKZEBER0aSHnDeH9ffRG7MNEgEMiCiAKBQKvPDCC+Lfb775pt1dwUGDBgGw3MGaOnUqDh06hPLychw7dgwfffQRRo4cabf++QMvbS9E1q9fj4qKCqfPmrHN06xZswBYurSMHj0aa9euRXFxMYqLi7Fu3TrcdtttYl/0V155xaX+/c8//zzkcjnKy8tx++234+uvv0ZBQQHKysqwadMm3HnnnWLL2PTp08W7pYFu+vTpCA8Ph9FoxH333YfFixfjxIkTKCsrw44dOzBhwgTx2TLDhw+3a/nq168fbrrpJgCWQHHWrFnIzs5GeXk5tm3bhrFjx2L//v0uTe/at29fcVzMO++8gy+++AK5ubkoKSnB33//jenTp+OZZ56x28aXA3dtx06tXbsW586d80qrhitmzZqFkJAQnDt3DrfffjuWLVuG3NxclJaWYuvWrbj77rvFZ4Lcd999bnfj+umnn/Dwww9j7969qKiowM6dO3H//fdj69atAP6tS1ZdunQRZ0XbuHEjxo8fj61bt4pjQj7//HM88MADMBqNUCqVmDZtWjNLwrGOHTsiOTkZRqMRR48eRUhIiN1kCcC/z0WzPhPN0fOrvH2Oacw111wjBlirV6/GH3/8IS4LhLK2toZlZ2c36WaBwWCwmxK8qqoKRUVFOHjwIL744gvcdNNNWL9+PQDg2muvdTgN//Lly8VW+8GDB4vPlTqfVCrFvHnzxGnLFy5ciAMHDjTpuKxjFd1pHbKaOXOmeONi+fLlGDFiBJYuXYqDBw+itLQUpaWl2LdvH5YsWYLhw4cjOzsb0dHReP755+ulpVAosHjxYnTs2BGCIGD+/Pm49dZbsXLlShw/fhwhISG4++67MW3aNKhUKuj1evEc+eSTT9q1ODmybt06rFy5ssF/DT1nzzrFeUvoJUEtEydVoIBy6aWX4sYbb8T333+PqqoqvPzyy1i0aBEAYOrUqdixYweKi4uxdu1auz7gVldffTVKSkqwb98+u+4BgGWAcK9evbB7927xBHz+dOCOXHvttXj55Zfx4osv4uTJk2I/clthYWF46aWXHF7wNKRnz5545513MGXKFBQXF+O5556rt05ISAimTJni8CGlgSolJQUfffQRJk6ciJKSEixYsEC8O2nr6quvdjhL00svvQSdToeff/4ZX331Fb766iu75XfccQcyMzOb/EyKmJgYzJgxAy+++CL0er3Du5kKhQL33nuvOLj85MmTbrV6uCMlJQVt2rRBQUEBFi1ahEWLFuHmm2+ud8fel/r06YOFCxdi6tSpqKiowOzZszF79ux6640ePRqPPfaYW/uIj49HRkYGfvvtt3oPG5VIJJg2bRpuueWWets988wzqKmpwTfffIMdO3bgvvvuq7eOWq3Gm2++ic6dO7uVt6a48sor8eWXXwKwPOfp/NYR2xYjwHFABHj3HNMUzz33HDIzM6HT6TBr1iz88MMPYtDl77K2dsOqqanBnj177LpzOvLggw82mqb14blPP/10ve6qR44cwauvvgrAUj/nzJnTYFpt2rTBrFmzMHXqVBiNRjz55JNYvXp1o0GrNSBqznTbkZGR+Pjjj/Hiiy9i/fr1OHHiBObNm+d0fY1Gg7lz5zqd6CYxMRErVqzAvHnz8M033+DAgQN203Ofzzq1/Jw5c1BQUID777/f6TOt5s+f3+jxREREYOTIkfXeP378uHhToTnPayJqCFuIKODMmDFDbM3ZsGGDOIahbdu2WL16NcaNG4f27dtDLpdDoVCgTZs2GDJkCN59910sWbJEvNu5Y8cO8SRq9dZbb2Hw4MGIiIiAUqls8jTWo0ePxo8//ohx48YhPT0doaGhCA8PR+fOnfHwww9j3bp1GDFihFvHe9VVV2HDhg148MEH0alTJ4SFhSEsLAwdOnTA+PHj8f333zu8CAl0vXv3xvr16zFlyhT06tULUVFRCAkJQWxsLK688kosWLAAS5YscXjhoFKpsHDhQixYsECcNCM0NBTdunXDnDlz7FoSm+qOO+7AJ598gquuugoxMTGQyWRQq9XQaDQYN24c/ve//2Hy5Mni84+sLVi+EBISgiVLlmDAgAHi5x8IU8sOGTIEGzZswEMPPYQuXbogPDwccrkcCQkJuO666/Dxxx/j5ZdfdvthjAqFAkuXLsW0adOQkZEBpVKJ5ORkDB8+HKtWrXJa7+VyOebMmYNly5Zh+PDhSE5OhkKhgEqlQseOHXHvvfdi3bp19SYw8DTb9B1Nm2wbJKWmptbrpmTLm+eYxiQlJYnPDMvNzcXbb78tLvN3Waenp4sPnLV2TXSVQqFAXFwc+vbtiwkTJuCHH37ArFmz6j28VK/X44knnkB1dTUkEgnmzp3bpGcojRgxQuy+lpOTIwZUztTU1GD79u0IDQ0VZyV0V2xsLN5++22sXLkSDz30EHr16oU2bdpApVIhLCwM7dq1w3XXXYc333wTq1evbvThveHh4XjppZewbt06TJgwAb169UJkZKTd+XL06NFYunQpMjMzMW7cOJhMJrz//vu48sor7bode4p1psaGHuRM1FwSwddPECQioqA2Y8YMrF69GsnJydi4caO/s0MBbuXKlXj22WfRvn17n96ooKbZu3cvPvjgA2zduhUbN25s1oN4HRkzZgx27dqFSZMmYeLEiR5Nm8iKLUREREQUsEaOHImkpCScOnUKf//9t7+zQ+fp0aMH3nnnHWzatMnjwdDRo0exa9cuqNVqjBs3zqNpE9liQEREREQBSy6X44EHHgCARsd8kv/YTg7jKStXrgQA3HnnnfVmaCTyJAZEREREFNBGjRqF9PR0/Pzzz8jOzvZ3dsgHioqK8NVXXyE2Nhb33nuvv7NDFzgGRERERBTQFAoFXnvtNUgkEoczVtKFZ+HChdDr9Xj++ec93hWP6HwMiIiIiCjg9ezZE/fffz9++eUXr8xmRoHj2LFj+PbbbzF06FBcf/31/s4OBQHOMkdEREREREGLLURERERERBS0GBAREREREVHQYkBERERERERBiwEREREREREFLQZEREREREQUtBgQERERERFR0GJAREREREREQYsBERERERERBa3/B/PhRHA6JM42AAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 1000x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Define figure settings\n",
    "figwidth = 10\n",
    "figheight = 3\n",
    "xlab_size = 20\n",
    "ylab_size = 20\n",
    "figfont_size = 24\n",
    "\n",
    "# Define consistent color maps\n",
    "my_cmap_sns_dark = [(0.8392156862745098, 0.15294117647058825, 0.1568627450980392)]\n",
    "my_cmap_sns_light = [(0.984313725490196, 0.6039215686274509, 0.6)]\n",
    "my_cmap_dark = sns.color_palette(my_cmap_sns_dark, as_cmap=True)\n",
    "my_cmap_light = sns.color_palette(my_cmap_sns_light, as_cmap=True)\n",
    "\n",
    "# Create data frame for split violinplot\n",
    "dfvp = pd.DataFrame()\n",
    "dfvp[\"values\"] = sample\n",
    "dfvp[\"all\"] = \"\"                                        # attribute that is shared by all entries\n",
    "# print(dfvp)\n",
    "\n",
    "# Create empty figure and plot the individual datapoints\n",
    "fig, ax = plt.subplots(figsize=(figwidth,figheight))\n",
    "\n",
    "\n",
    "# 1. Violinplot\n",
    "axsns = sns.violinplot(x='values', y='all',    # y='all' just need to group both types together\n",
    "                       data=dfvp,\n",
    "                       hue = True, hue_order = [False, True],\n",
    "                       split = True, inner = 'quartile',\n",
    "                       cut=0,                  # 0 means ending sharp at end points\n",
    "                       width=.6, scale = 'width',\n",
    "                       # dodge = False,        # When using ``hue`` nesting, setting this to ``True`` will separate the strips for different hue levels along the categorical axis.\n",
    "                       orient = 'h',\n",
    "                       color=my_cmap_light[0],)\n",
    "\n",
    "# change the medium default linke to full\n",
    "for l in axsns.lines[1::3]:\n",
    "    l.set_linestyle('-')\n",
    "    l.set_linewidth(1.5)\n",
    "    l.set_color('black')\n",
    "    l.set_alpha(0.8)\n",
    "\n",
    "\n",
    "# 2. Plot individual points\n",
    "y_tilt = -0.13                                   # Set some delta for the points below the violinplot\n",
    "y_base = np.zeros(len(data_ratio)) + y_tilt     # base vector to which to broadcast y-tilt values\n",
    "\n",
    "ax.plot(data_ratio, y_base,\n",
    "        # 'o',\n",
    "        '^',\n",
    "        alpha=1,\n",
    "        zorder=20,      # higher means more visible\n",
    "        markersize=11,\n",
    "        markeredgewidth=0,\n",
    "        # markerfacecolor='none',\n",
    "        markerfacecolor=my_cmap_dark[0],\n",
    "        markeredgecolor=my_cmap_dark[0],\n",
    "        )\n",
    "\n",
    "\n",
    "# 3. CI Errorbars & show numbers\n",
    "axeb = plt.errorbar(median_ratio, 0, xerr=np.array([[ci_ratio_delta[0], ci_ratio_delta[1]]]).T,\n",
    "                    fmt='o',\n",
    "                    markersize=10, alpha=1,\n",
    "                    # lw = 3,\n",
    "                    lw = 5,\n",
    "                    zorder=100,      # higher means more visible\n",
    "                    capsize = 0,     # 10\n",
    "                    # capthick = 4,\n",
    "                    capthick = 0,\n",
    "                    # color = 'black',\n",
    "                    color = my_cmap_dark[0],\n",
    "                    )\n",
    "\n",
    "med = np.median(sample)\n",
    "# ax.text(med, 0.32, f'{100*med:.1f}%', horizontalalignment='center', color='black', fontsize=20)\n",
    "# ax.text(med, 0.32, f'{med:.2f}', horizontalalignment='center', color='black', fontsize=20)\n",
    "ax.text(med, -0.1, f'{med:.2f}', horizontalalignment='center',\n",
    "        # color='black',\n",
    "        color = my_cmap_dark[0],\n",
    "        fontsize=figfont_size)\n",
    "# ax.text(ci_ratio.low, 0.04, f'{100*ci_ratio.low:.1f}%', horizontalalignment='center', color='black', fontsize=20)\n",
    "# ax.text(ci_ratio.high, 0.04, f'{100*ci_ratio.high:.1f}%', horizontalalignment='center', color='black', fontsize=20)\n",
    "\n",
    "\n",
    "# 4. vertical bar for x-axis = 1\n",
    "plt.plot([1, 1], [-10, 10], color = 'black', zorder = 0, linewidth = 2)\n",
    "\n",
    "\n",
    "# Additional settings\n",
    "# ax.set_ylim(-0.2, 0.4)\n",
    "ax.set_xticks(np.linspace(0, 2, num=21))\n",
    "ax.set_ylim(-0.25, 0.35)\n",
    "ax.set_ylabel(None)         # remove the 'all'\n",
    "ax.set_xlim(0.2, 1.25)\n",
    "if VARIANT == 1:\n",
    "    ax.set_xlim(0.2, 1.301)\n",
    "if VARIANT == 3:\n",
    "    ax.set_xlim(0.499, 1.205)\n",
    "# ax.set_xlabel('Ratio of median time per worker (RD / SQL)', size = xlab_size)\n",
    "ax.set_xlabel('Ratio of median time per worker (DX / SQL)', size = xlab_size)\n",
    "sns.despine(left=True)               # remove bounding box\n",
    "plt.grid(axis = 'x', linewidth = 0.5, color = 'lightgray')\n",
    "ax.legend_.remove()\n",
    "\n",
    "if savefig:\n",
    "    plt.savefig(fig_dir + f'/q1_figure2_variant{VARIANT}.pdf', bbox_inches='tight')\n",
    "    plt.savefig(fig_dir + f'/q1_figure2_variant{VARIANT}.svg', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c17c56d9",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "# Question 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "d81ad1bc",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "df6:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th>mode</th>\n",
       "      <th colspan=\"2\" halign=\"left\">RD</th>\n",
       "      <th colspan=\"2\" halign=\"left\">SQL</th>\n",
       "      <th>RD</th>\n",
       "      <th>SQL</th>\n",
       "      <th>H1</th>\n",
       "      <th>H2</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>section</th>\n",
       "      <th>H1</th>\n",
       "      <th>H2</th>\n",
       "      <th>H1</th>\n",
       "      <th>H2</th>\n",
       "      <th>ratio</th>\n",
       "      <th>ratio</th>\n",
       "      <th>ratio</th>\n",
       "      <th>ratio</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>worker_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7.991</td>\n",
       "      <td>5.639</td>\n",
       "      <td>54.237</td>\n",
       "      <td>6.925</td>\n",
       "      <td>0.706</td>\n",
       "      <td>0.128</td>\n",
       "      <td>0.147</td>\n",
       "      <td>0.814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9.747</td>\n",
       "      <td>8.585</td>\n",
       "      <td>13.806</td>\n",
       "      <td>11.911</td>\n",
       "      <td>0.881</td>\n",
       "      <td>0.863</td>\n",
       "      <td>0.706</td>\n",
       "      <td>0.721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>12.148</td>\n",
       "      <td>7.399</td>\n",
       "      <td>22.152</td>\n",
       "      <td>16.363</td>\n",
       "      <td>0.609</td>\n",
       "      <td>0.739</td>\n",
       "      <td>0.548</td>\n",
       "      <td>0.452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>11.375</td>\n",
       "      <td>7.215</td>\n",
       "      <td>17.726</td>\n",
       "      <td>14.008</td>\n",
       "      <td>0.634</td>\n",
       "      <td>0.790</td>\n",
       "      <td>0.642</td>\n",
       "      <td>0.515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>14.007</td>\n",
       "      <td>6.982</td>\n",
       "      <td>26.168</td>\n",
       "      <td>9.383</td>\n",
       "      <td>0.498</td>\n",
       "      <td>0.359</td>\n",
       "      <td>0.535</td>\n",
       "      <td>0.744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>6.242</td>\n",
       "      <td>4.905</td>\n",
       "      <td>7.510</td>\n",
       "      <td>4.858</td>\n",
       "      <td>0.786</td>\n",
       "      <td>0.647</td>\n",
       "      <td>0.831</td>\n",
       "      <td>1.009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>8.774</td>\n",
       "      <td>7.048</td>\n",
       "      <td>13.952</td>\n",
       "      <td>12.367</td>\n",
       "      <td>0.803</td>\n",
       "      <td>0.886</td>\n",
       "      <td>0.629</td>\n",
       "      <td>0.570</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>12.962</td>\n",
       "      <td>9.658</td>\n",
       "      <td>14.064</td>\n",
       "      <td>9.983</td>\n",
       "      <td>0.745</td>\n",
       "      <td>0.710</td>\n",
       "      <td>0.922</td>\n",
       "      <td>0.967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>15.100</td>\n",
       "      <td>19.551</td>\n",
       "      <td>23.410</td>\n",
       "      <td>30.148</td>\n",
       "      <td>1.295</td>\n",
       "      <td>1.288</td>\n",
       "      <td>0.645</td>\n",
       "      <td>0.648</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>8.720</td>\n",
       "      <td>6.993</td>\n",
       "      <td>19.712</td>\n",
       "      <td>7.722</td>\n",
       "      <td>0.802</td>\n",
       "      <td>0.392</td>\n",
       "      <td>0.442</td>\n",
       "      <td>0.906</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>15.617</td>\n",
       "      <td>11.213</td>\n",
       "      <td>55.898</td>\n",
       "      <td>15.980</td>\n",
       "      <td>0.718</td>\n",
       "      <td>0.286</td>\n",
       "      <td>0.279</td>\n",
       "      <td>0.702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>14.982</td>\n",
       "      <td>6.216</td>\n",
       "      <td>65.841</td>\n",
       "      <td>8.636</td>\n",
       "      <td>0.415</td>\n",
       "      <td>0.131</td>\n",
       "      <td>0.228</td>\n",
       "      <td>0.720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>10.453</td>\n",
       "      <td>3.932</td>\n",
       "      <td>35.611</td>\n",
       "      <td>7.218</td>\n",
       "      <td>0.376</td>\n",
       "      <td>0.203</td>\n",
       "      <td>0.294</td>\n",
       "      <td>0.545</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>19.684</td>\n",
       "      <td>10.204</td>\n",
       "      <td>20.617</td>\n",
       "      <td>15.413</td>\n",
       "      <td>0.518</td>\n",
       "      <td>0.748</td>\n",
       "      <td>0.955</td>\n",
       "      <td>0.662</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>11.396</td>\n",
       "      <td>5.369</td>\n",
       "      <td>25.916</td>\n",
       "      <td>5.913</td>\n",
       "      <td>0.471</td>\n",
       "      <td>0.228</td>\n",
       "      <td>0.440</td>\n",
       "      <td>0.908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>14.386</td>\n",
       "      <td>7.908</td>\n",
       "      <td>10.149</td>\n",
       "      <td>8.407</td>\n",
       "      <td>0.550</td>\n",
       "      <td>0.828</td>\n",
       "      <td>1.417</td>\n",
       "      <td>0.941</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>15.075</td>\n",
       "      <td>10.655</td>\n",
       "      <td>10.811</td>\n",
       "      <td>9.302</td>\n",
       "      <td>0.707</td>\n",
       "      <td>0.860</td>\n",
       "      <td>1.394</td>\n",
       "      <td>1.146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>9.150</td>\n",
       "      <td>8.181</td>\n",
       "      <td>14.344</td>\n",
       "      <td>11.183</td>\n",
       "      <td>0.894</td>\n",
       "      <td>0.780</td>\n",
       "      <td>0.638</td>\n",
       "      <td>0.732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>26.365</td>\n",
       "      <td>13.165</td>\n",
       "      <td>40.867</td>\n",
       "      <td>17.230</td>\n",
       "      <td>0.499</td>\n",
       "      <td>0.422</td>\n",
       "      <td>0.645</td>\n",
       "      <td>0.764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>22.348</td>\n",
       "      <td>10.909</td>\n",
       "      <td>22.399</td>\n",
       "      <td>11.322</td>\n",
       "      <td>0.488</td>\n",
       "      <td>0.505</td>\n",
       "      <td>0.998</td>\n",
       "      <td>0.964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>12.448</td>\n",
       "      <td>25.276</td>\n",
       "      <td>28.640</td>\n",
       "      <td>36.115</td>\n",
       "      <td>2.031</td>\n",
       "      <td>1.261</td>\n",
       "      <td>0.435</td>\n",
       "      <td>0.700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>12.041</td>\n",
       "      <td>35.852</td>\n",
       "      <td>43.927</td>\n",
       "      <td>24.125</td>\n",
       "      <td>2.977</td>\n",
       "      <td>0.549</td>\n",
       "      <td>0.274</td>\n",
       "      <td>1.486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>14.380</td>\n",
       "      <td>13.649</td>\n",
       "      <td>20.355</td>\n",
       "      <td>15.739</td>\n",
       "      <td>0.949</td>\n",
       "      <td>0.773</td>\n",
       "      <td>0.706</td>\n",
       "      <td>0.867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>21.773</td>\n",
       "      <td>9.941</td>\n",
       "      <td>71.220</td>\n",
       "      <td>13.771</td>\n",
       "      <td>0.457</td>\n",
       "      <td>0.193</td>\n",
       "      <td>0.306</td>\n",
       "      <td>0.722</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>22.406</td>\n",
       "      <td>15.235</td>\n",
       "      <td>82.963</td>\n",
       "      <td>20.297</td>\n",
       "      <td>0.680</td>\n",
       "      <td>0.245</td>\n",
       "      <td>0.270</td>\n",
       "      <td>0.751</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>20.476</td>\n",
       "      <td>20.556</td>\n",
       "      <td>67.059</td>\n",
       "      <td>13.220</td>\n",
       "      <td>1.004</td>\n",
       "      <td>0.197</td>\n",
       "      <td>0.305</td>\n",
       "      <td>1.555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>4.927</td>\n",
       "      <td>4.822</td>\n",
       "      <td>7.377</td>\n",
       "      <td>5.995</td>\n",
       "      <td>0.979</td>\n",
       "      <td>0.813</td>\n",
       "      <td>0.668</td>\n",
       "      <td>0.804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>18.061</td>\n",
       "      <td>17.733</td>\n",
       "      <td>23.764</td>\n",
       "      <td>18.866</td>\n",
       "      <td>0.982</td>\n",
       "      <td>0.794</td>\n",
       "      <td>0.760</td>\n",
       "      <td>0.940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>23.182</td>\n",
       "      <td>5.483</td>\n",
       "      <td>63.761</td>\n",
       "      <td>19.377</td>\n",
       "      <td>0.237</td>\n",
       "      <td>0.304</td>\n",
       "      <td>0.364</td>\n",
       "      <td>0.283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>29.262</td>\n",
       "      <td>20.494</td>\n",
       "      <td>36.913</td>\n",
       "      <td>23.960</td>\n",
       "      <td>0.700</td>\n",
       "      <td>0.649</td>\n",
       "      <td>0.793</td>\n",
       "      <td>0.855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>13.155</td>\n",
       "      <td>9.254</td>\n",
       "      <td>29.754</td>\n",
       "      <td>12.825</td>\n",
       "      <td>0.703</td>\n",
       "      <td>0.431</td>\n",
       "      <td>0.442</td>\n",
       "      <td>0.722</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>4.472</td>\n",
       "      <td>4.218</td>\n",
       "      <td>4.181</td>\n",
       "      <td>4.517</td>\n",
       "      <td>0.943</td>\n",
       "      <td>1.080</td>\n",
       "      <td>1.069</td>\n",
       "      <td>0.934</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>7.749</td>\n",
       "      <td>7.720</td>\n",
       "      <td>4.866</td>\n",
       "      <td>5.247</td>\n",
       "      <td>0.996</td>\n",
       "      <td>1.079</td>\n",
       "      <td>1.593</td>\n",
       "      <td>1.471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>9.007</td>\n",
       "      <td>7.050</td>\n",
       "      <td>11.855</td>\n",
       "      <td>9.004</td>\n",
       "      <td>0.783</td>\n",
       "      <td>0.759</td>\n",
       "      <td>0.760</td>\n",
       "      <td>0.783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>10.337</td>\n",
       "      <td>3.229</td>\n",
       "      <td>18.303</td>\n",
       "      <td>6.014</td>\n",
       "      <td>0.312</td>\n",
       "      <td>0.329</td>\n",
       "      <td>0.565</td>\n",
       "      <td>0.537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>7.131</td>\n",
       "      <td>6.832</td>\n",
       "      <td>6.951</td>\n",
       "      <td>6.143</td>\n",
       "      <td>0.958</td>\n",
       "      <td>0.884</td>\n",
       "      <td>1.026</td>\n",
       "      <td>1.112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>10.238</td>\n",
       "      <td>10.995</td>\n",
       "      <td>11.290</td>\n",
       "      <td>15.410</td>\n",
       "      <td>1.074</td>\n",
       "      <td>1.365</td>\n",
       "      <td>0.907</td>\n",
       "      <td>0.713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>21.690</td>\n",
       "      <td>6.064</td>\n",
       "      <td>7.111</td>\n",
       "      <td>3.978</td>\n",
       "      <td>0.280</td>\n",
       "      <td>0.559</td>\n",
       "      <td>3.050</td>\n",
       "      <td>1.524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>3.986</td>\n",
       "      <td>3.192</td>\n",
       "      <td>8.069</td>\n",
       "      <td>4.089</td>\n",
       "      <td>0.801</td>\n",
       "      <td>0.507</td>\n",
       "      <td>0.494</td>\n",
       "      <td>0.781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>5.294</td>\n",
       "      <td>3.913</td>\n",
       "      <td>17.922</td>\n",
       "      <td>8.953</td>\n",
       "      <td>0.739</td>\n",
       "      <td>0.500</td>\n",
       "      <td>0.295</td>\n",
       "      <td>0.437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>25.429</td>\n",
       "      <td>13.747</td>\n",
       "      <td>12.210</td>\n",
       "      <td>14.229</td>\n",
       "      <td>0.541</td>\n",
       "      <td>1.165</td>\n",
       "      <td>2.083</td>\n",
       "      <td>0.966</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>19.672</td>\n",
       "      <td>12.869</td>\n",
       "      <td>62.773</td>\n",
       "      <td>23.929</td>\n",
       "      <td>0.654</td>\n",
       "      <td>0.381</td>\n",
       "      <td>0.313</td>\n",
       "      <td>0.538</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>154</th>\n",
       "      <td>15.026</td>\n",
       "      <td>7.610</td>\n",
       "      <td>18.871</td>\n",
       "      <td>9.690</td>\n",
       "      <td>0.506</td>\n",
       "      <td>0.513</td>\n",
       "      <td>0.796</td>\n",
       "      <td>0.785</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>7.914</td>\n",
       "      <td>4.925</td>\n",
       "      <td>24.032</td>\n",
       "      <td>10.136</td>\n",
       "      <td>0.622</td>\n",
       "      <td>0.422</td>\n",
       "      <td>0.329</td>\n",
       "      <td>0.486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>10.122</td>\n",
       "      <td>7.675</td>\n",
       "      <td>15.822</td>\n",
       "      <td>13.043</td>\n",
       "      <td>0.758</td>\n",
       "      <td>0.824</td>\n",
       "      <td>0.640</td>\n",
       "      <td>0.588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>7.441</td>\n",
       "      <td>3.383</td>\n",
       "      <td>11.116</td>\n",
       "      <td>8.642</td>\n",
       "      <td>0.455</td>\n",
       "      <td>0.777</td>\n",
       "      <td>0.669</td>\n",
       "      <td>0.391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>8.486</td>\n",
       "      <td>8.392</td>\n",
       "      <td>11.799</td>\n",
       "      <td>9.719</td>\n",
       "      <td>0.989</td>\n",
       "      <td>0.824</td>\n",
       "      <td>0.719</td>\n",
       "      <td>0.864</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>166</th>\n",
       "      <td>17.133</td>\n",
       "      <td>10.287</td>\n",
       "      <td>12.759</td>\n",
       "      <td>10.253</td>\n",
       "      <td>0.600</td>\n",
       "      <td>0.804</td>\n",
       "      <td>1.343</td>\n",
       "      <td>1.003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>24.126</td>\n",
       "      <td>30.665</td>\n",
       "      <td>39.257</td>\n",
       "      <td>35.037</td>\n",
       "      <td>1.271</td>\n",
       "      <td>0.893</td>\n",
       "      <td>0.615</td>\n",
       "      <td>0.875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>5.588</td>\n",
       "      <td>4.401</td>\n",
       "      <td>7.989</td>\n",
       "      <td>5.506</td>\n",
       "      <td>0.788</td>\n",
       "      <td>0.689</td>\n",
       "      <td>0.700</td>\n",
       "      <td>0.799</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "mode          RD           SQL           RD   SQL    H1    H2\n",
       "section       H1     H2     H1     H2 ratio ratio ratio ratio\n",
       "worker_id                                                    \n",
       "0          7.991  5.639 54.237  6.925 0.706 0.128 0.147 0.814\n",
       "2          9.747  8.585 13.806 11.911 0.881 0.863 0.706 0.721\n",
       "5         12.148  7.399 22.152 16.363 0.609 0.739 0.548 0.452\n",
       "6         11.375  7.215 17.726 14.008 0.634 0.790 0.642 0.515\n",
       "10        14.007  6.982 26.168  9.383 0.498 0.359 0.535 0.744\n",
       "17         6.242  4.905  7.510  4.858 0.786 0.647 0.831 1.009\n",
       "21         8.774  7.048 13.952 12.367 0.803 0.886 0.629 0.570\n",
       "28        12.962  9.658 14.064  9.983 0.745 0.710 0.922 0.967\n",
       "32        15.100 19.551 23.410 30.148 1.295 1.288 0.645 0.648\n",
       "39         8.720  6.993 19.712  7.722 0.802 0.392 0.442 0.906\n",
       "42        15.617 11.213 55.898 15.980 0.718 0.286 0.279 0.702\n",
       "43        14.982  6.216 65.841  8.636 0.415 0.131 0.228 0.720\n",
       "50        10.453  3.932 35.611  7.218 0.376 0.203 0.294 0.545\n",
       "52        19.684 10.204 20.617 15.413 0.518 0.748 0.955 0.662\n",
       "57        11.396  5.369 25.916  5.913 0.471 0.228 0.440 0.908\n",
       "58        14.386  7.908 10.149  8.407 0.550 0.828 1.417 0.941\n",
       "60        15.075 10.655 10.811  9.302 0.707 0.860 1.394 1.146\n",
       "66         9.150  8.181 14.344 11.183 0.894 0.780 0.638 0.732\n",
       "72        26.365 13.165 40.867 17.230 0.499 0.422 0.645 0.764\n",
       "75        22.348 10.909 22.399 11.322 0.488 0.505 0.998 0.964\n",
       "77        12.448 25.276 28.640 36.115 2.031 1.261 0.435 0.700\n",
       "80        12.041 35.852 43.927 24.125 2.977 0.549 0.274 1.486\n",
       "81        14.380 13.649 20.355 15.739 0.949 0.773 0.706 0.867\n",
       "83        21.773  9.941 71.220 13.771 0.457 0.193 0.306 0.722\n",
       "87        22.406 15.235 82.963 20.297 0.680 0.245 0.270 0.751\n",
       "89        20.476 20.556 67.059 13.220 1.004 0.197 0.305 1.555\n",
       "91         4.927  4.822  7.377  5.995 0.979 0.813 0.668 0.804\n",
       "92        18.061 17.733 23.764 18.866 0.982 0.794 0.760 0.940\n",
       "96        23.182  5.483 63.761 19.377 0.237 0.304 0.364 0.283\n",
       "110       29.262 20.494 36.913 23.960 0.700 0.649 0.793 0.855\n",
       "115       13.155  9.254 29.754 12.825 0.703 0.431 0.442 0.722\n",
       "117        4.472  4.218  4.181  4.517 0.943 1.080 1.069 0.934\n",
       "119        7.749  7.720  4.866  5.247 0.996 1.079 1.593 1.471\n",
       "121        9.007  7.050 11.855  9.004 0.783 0.759 0.760 0.783\n",
       "125       10.337  3.229 18.303  6.014 0.312 0.329 0.565 0.537\n",
       "130        7.131  6.832  6.951  6.143 0.958 0.884 1.026 1.112\n",
       "136       10.238 10.995 11.290 15.410 1.074 1.365 0.907 0.713\n",
       "141       21.690  6.064  7.111  3.978 0.280 0.559 3.050 1.524\n",
       "143        3.986  3.192  8.069  4.089 0.801 0.507 0.494 0.781\n",
       "146        5.294  3.913 17.922  8.953 0.739 0.500 0.295 0.437\n",
       "148       25.429 13.747 12.210 14.229 0.541 1.165 2.083 0.966\n",
       "153       19.672 12.869 62.773 23.929 0.654 0.381 0.313 0.538\n",
       "154       15.026  7.610 18.871  9.690 0.506 0.513 0.796 0.785\n",
       "158        7.914  4.925 24.032 10.136 0.622 0.422 0.329 0.486\n",
       "159       10.122  7.675 15.822 13.043 0.758 0.824 0.640 0.588\n",
       "162        7.441  3.383 11.116  8.642 0.455 0.777 0.669 0.391\n",
       "165        8.486  8.392 11.799  9.719 0.989 0.824 0.719 0.864\n",
       "166       17.133 10.287 12.759 10.253 0.600 0.804 1.343 1.003\n",
       "168       24.126 30.665 39.257 35.037 1.271 0.893 0.615 0.875\n",
       "169        5.588  4.401  7.989  5.506 0.788 0.689 0.700 0.799"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RD, H1: 12.298, 95% CI [10.180, 14.982]\n",
      "RD, H2: 7.814, 95% CI [7.022, 9.941]\n",
      "RD, ratio: 0.712, 95% CI [0.634, 0.793]\n",
      "SQL, H1: 19.291, 95% CI [14.064, 23.764]\n",
      "SQL, H2: 10.718, 95% CI [9.299, 13.495]\n",
      "SQL, ratio: 0.700, 95% CI [0.507, 0.785]\n",
      "H1: 0.643, 95% CI [0.548, 0.713]\n",
      "H2: 0.782, 95% CI [0.721, 0.864]\n"
     ]
    }
   ],
   "source": [
    "# Create df6, df7\n",
    "df0 = df_filtered_data[['worker_id', 'question', 'time', 'mode']].copy()\n",
    "df0['H1'] = np.where(df0['question'].between(1, 16, inclusive='both'), 'H1', '')\n",
    "df0['H2'] = np.where(df0['question'].between(17, 32, inclusive='both'), 'H2', '')\n",
    "# display(df0)\n",
    "\n",
    "df1 = df0[['worker_id','question','time','mode','H1']].rename(columns={'H1': 'section'}) # Two sections: 1st half (H1) and 2nd half (H2)\n",
    "df2 = df0[['worker_id','question','time','mode','H2']].rename(columns={'H2': 'section'})\n",
    "df4 = pd.concat([df1, df2])\n",
    "df4 = df4.loc[df4['section'] != '']\n",
    "df4 = df4.reset_index(inplace=False, drop=True)\n",
    "# display(df4)\n",
    "\n",
    "df5 = df4.groupby(['worker_id', 'mode', 'section']).time.agg(['median'])                # for each worker, calculate median for both modes and section\n",
    "df5.reset_index(inplace=True)\n",
    "# display(df5)\n",
    "\n",
    "# pivot to have one row per worker\n",
    "df6 = pd.pivot_table(df5, values=['median'], index=['worker_id'], columns=['mode', 'section'])\n",
    "df6=df6.droplevel(0, axis=1)\n",
    "\n",
    "# relative improvements per user per mode\n",
    "df6['RD', 'ratio'] = df6['RD', 'H2'] / df6['RD', 'H1']\n",
    "df6['SQL', 'ratio'] = df6['SQL', 'H2'] / df6['SQL', 'H1']\n",
    "# relative improvements of RD over SQL per user half\n",
    "df6['H1', 'ratio'] = df6['RD', 'H1'] / df6['SQL', 'H1']\n",
    "df6['H2', 'ratio'] = df6['RD', 'H2'] / df6['SQL', 'H2']\n",
    "print('df6:')\n",
    "display(df6)\n",
    "\n",
    "# Median of median task time for each mode and section\n",
    "modes = ['RD', 'SQL']\n",
    "sections = ['H1', 'H2', 'ratio']\n",
    "median_time = {}\n",
    "ci = {}\n",
    "ci_delta = {}\n",
    "for mode in modes:\n",
    "    for section in sections:\n",
    "        column = (mode, section)\n",
    "        median_time[column] = np.median(df6[column])\n",
    "        ci[column] = scipybootstrap((df6[column],), statistic=np.median,\n",
    "                                    n_resamples=BOOTSTRAPSAMPLES,\n",
    "                                    confidence_level=BOOTSTRAPCONFIDENCE,\n",
    "                                    method='percentile',\n",
    "                                    axis=0).confidence_interval        #convert array to sequence\n",
    "        ci_delta[column] = [median_time[column] - ci[column].low, ci[column].high - median_time[column]]\n",
    "        print(f'{mode}, {section}: {median_time[column]:.3f}, 95% CI [{ci[column].low:.3f}, {ci[column].high:.3f}]')\n",
    "\n",
    "for half in ['H1', 'H2']:\n",
    "    column = half\n",
    "    median_time[column] = np.median(df6[column])\n",
    "    ci[column] = scipybootstrap((df6[column],), statistic=np.median,\n",
    "                                n_resamples=BOOTSTRAPSAMPLES,\n",
    "                                confidence_level=BOOTSTRAPCONFIDENCE,\n",
    "                                method='percentile',\n",
    "                                axis=0).confidence_interval        #convert array to sequence\n",
    "    ci_delta[column] = [median_time[column] - ci[column].low, ci[column].high - median_time[column]]\n",
    "    print(f'{column}: {median_time[column]:.3f}, 95% CI [{ci[column].low[0]:.3f}, {ci[column].high[0]:.3f}]')\n",
    "\n",
    "# uses df5 to make df7 (used for later plot)\n",
    "modes = ['SQL', 'RD']\n",
    "sections = ['H1', 'H2']\n",
    "df7 = df5.loc[df5['section'].isin(sections)]\n",
    "# display(df7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "cc21c930",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot (uses df6, df7)\n",
    "\n",
    "# Define pre-settings\n",
    "figwidth = 10\n",
    "figheight = 6\n",
    "if VARIANT == 1:\n",
    "    figheight = 8\n",
    "xlab_size = 20\n",
    "ylab_size = 20\n",
    "figfont_size = 24\n",
    "\n",
    "# Define consistent color maps\n",
    "my_cmap_sns_light = [(0.7764705882352941, 0.8588235294117647, 0.9372549019607843), (0.9921568627450981, 0.8156862745098039, 0.6352941176470588)]        # light orange, light blue\n",
    "my_cmap_sns_dark = [(0.19215686274509805, 0.5098039215686274, 0.7411764705882353), (0.9019607843137255, 0.3333333333333333, 0.050980392156862744)]      # dark orange, dark blue\n",
    "my_cmap_dark = sns.color_palette(my_cmap_sns_dark, as_cmap=True)\n",
    "my_cmap_light = sns.color_palette(my_cmap_sns_light, as_cmap=True)\n",
    "\n",
    "\n",
    "# Create empty figure and plot the individual datapoints\n",
    "fig, ax = plt.subplots(figsize=(figwidth,figheight))\n",
    "\n",
    "\n",
    "# 1. Violinplots\n",
    "axsns = sns.violinplot(x='median', y='section', data=df7,\n",
    "                       hue='mode',\n",
    "                       hue_order=['SQL', 'RD'],\n",
    "                       split=True,   # half violinplots https://stackoverflow.com/questions/53872439/half-not-split-violin-plots-in-seaborn\n",
    "                       inner='quartile',\n",
    "                       cut=0,               # 0 means ending sharp at end points\n",
    "                       width=.4,\n",
    "                       orient = 'h',\n",
    "                       zorder=20,\n",
    "                       palette = my_cmap_light,)\n",
    "\n",
    "# change the medium default line to full (https://stackoverflow.com/questions/60638344/quartiles-line-properties-in-seaborn-violinplot)\n",
    "for l in axsns.lines[1::3]:\n",
    "    l.set_linestyle('-')\n",
    "    l.set_linewidth(1.2)\n",
    "    l.set_color('black')\n",
    "    # l.set_alpha(0.8)\n",
    "\n",
    "\n",
    "# 2. Plot individual points\n",
    "y_base = np.zeros(df6.values.shape[0])  # base vector to which to broadcast y-tilt values\n",
    "y_tilt_mode = [0.3, 0.38]\n",
    "y_tilt_section = [0, 1]\n",
    "for i, mode in enumerate(modes):\n",
    "    for j, section in enumerate(sections):\n",
    "        column = (mode, section)\n",
    "        ax.plot(df6[column],\n",
    "                y_base + y_tilt_mode[i] + y_tilt_section[j],\n",
    "                # 'o',\n",
    "                # '|',\n",
    "                '^',\n",
    "                alpha=1,\n",
    "                zorder=20,      # higher means more visible\n",
    "                markersize=11,\n",
    "                markeredgewidth=0,\n",
    "                # markerfacecolor='none',\n",
    "                markerfacecolor=my_cmap_sns_dark[i],\n",
    "                markeredgecolor=my_cmap_sns_dark[i],)\n",
    "        ax.plot(df6[column],          # white background\n",
    "                y_base + y_tilt_mode[i] + y_tilt_section[j],\n",
    "                # 'o',\n",
    "                # '|',\n",
    "                '^',\n",
    "                markersize=11,\n",
    "                markeredgewidth=1,\n",
    "                markerfacecolor='white',\n",
    "                color ='white',\n",
    "                linewidth = None,\n",
    "                zorder=1,)\n",
    "\n",
    "\n",
    "# # 3. Plot lines connecting points\n",
    "# for idx in df6.index:\n",
    "#     for i, mode in enumerate(modes):\n",
    "#         for j in range(len(sections)-1):\n",
    "#             start = (mode, sections[j])\n",
    "#             end = (mode, sections[j+1])\n",
    "#             ax.plot(df6.loc[idx, [start, end]],\n",
    "#                     [y_tilt_mode[i] + y_tilt_section[j], y_tilt_mode[i] + y_tilt_section[j+1]],\n",
    "#                     color=my_cmap_sns_dark[i], linewidth=2, linestyle='-', alpha=.2, zorder=0)\n",
    "\n",
    "\n",
    "# 4. CI Errorbars & numbers\n",
    "y_tilt_mode = [0.5, 0.55]\n",
    "# y_tilt_section_bar = [0.23, 0.8]\n",
    "# y_tilt_section_number = [0.19, 0.89]\n",
    "for i, mode in enumerate(modes):\n",
    "    for j, section in enumerate(sections):\n",
    "        column = (mode, section)\n",
    "        plt.errorbar(median_time[column], y_tilt_mode[i]+y_tilt_section[j],\n",
    "                     xerr=np.array([[ci_delta[column][0] ,ci_delta[column][1]]]).T,\n",
    "                     fmt='o', markersize=10,\n",
    "                     lw = 3, alpha=1,\n",
    "                     zorder=100,        # higher means more visible\n",
    "                     #             capsize = 10, capthick = 4,\n",
    "                     capsize = 0,\n",
    "                     color = my_cmap_sns_dark[i]    # 'black'\n",
    "                     )\n",
    "        ax.text(median_time[column], y_tilt_mode[i]+y_tilt_section[j] + 0.18, f'{median_time[column]:.1f}',\n",
    "                horizontalalignment='center', color = my_cmap_sns_dark[i],\n",
    "                fontsize=figfont_size)\n",
    "\n",
    "\n",
    "# 5. Plot red line connecting medians\n",
    "for i, mode in enumerate(modes):\n",
    "    ax.plot([median_time[(mode, 'H1')], median_time[(mode, 'H2')]],\n",
    "            [y_tilt_mode[i]+y_tilt_section[0], y_tilt_mode[i]+y_tilt_section[1]],\n",
    "            color=my_cmap_sns_dark[i], linewidth = 3, linestyle ='-',\n",
    "            alpha=.3,\n",
    "            zorder=0)\n",
    "\n",
    "\n",
    "# #Additional settings\n",
    "ax.set_xticks(range(0, 100, 5))\n",
    "ax.set_xlabel('Median time per worker and halves (sec)', size = xlab_size)\n",
    "ax.set_ylabel(None)\n",
    "ax.set_xlim(0, 40.1)\n",
    "ax.set_ylim(1.82, -0.25)\n",
    "leg = plt. legend(loc='lower right',\n",
    "                  borderaxespad= 0.2,\n",
    "                  frameon = True,\n",
    "                  labelspacing = 0.1)\n",
    "leg.get_frame().set_alpha(1)\n",
    "leg.get_frame().set_linewidth(0.0)\n",
    "for text, text2 in zip(leg.get_texts(), ['SQL', 'DX']):\n",
    "    text.set_text(text2)\n",
    "\n",
    "plt.grid(axis = 'x', linewidth = 0.5, color = 'lightgray')\n",
    "sns.despine()               # remove bounding box\n",
    "\n",
    "if savefig:\n",
    "    plt.savefig(fig_dir + f'/q2_figure_variant{VARIANT}.pdf', bbox_inches='tight')\n",
    "    plt.savefig(fig_dir + f'/q2_figure_variant{VARIANT}.svg', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6cd2ab3b",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "# Question 3: four patterns\n",
    "\n",
    "(1) calculate the median time per pattern (4) across the two modes (2). Thus 8 values.\n",
    "(2) show repeated measure violin plot figure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "da5ecbe0",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "df9:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th>mode</th>\n",
       "      <th colspan=\"4\" halign=\"left\">RD</th>\n",
       "      <th colspan=\"4\" halign=\"left\">SQL</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pattern</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>worker_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10.743</td>\n",
       "      <td>8.359</td>\n",
       "      <td>6.698</td>\n",
       "      <td>6.008</td>\n",
       "      <td>14.905</td>\n",
       "      <td>36.797</td>\n",
       "      <td>59.405</td>\n",
       "      <td>20.626</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5.763</td>\n",
       "      <td>8.801</td>\n",
       "      <td>13.316</td>\n",
       "      <td>12.110</td>\n",
       "      <td>21.611</td>\n",
       "      <td>13.085</td>\n",
       "      <td>13.359</td>\n",
       "      <td>9.433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>4.341</td>\n",
       "      <td>11.566</td>\n",
       "      <td>12.451</td>\n",
       "      <td>5.909</td>\n",
       "      <td>16.144</td>\n",
       "      <td>17.267</td>\n",
       "      <td>22.487</td>\n",
       "      <td>15.869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>4.598</td>\n",
       "      <td>12.098</td>\n",
       "      <td>10.607</td>\n",
       "      <td>9.812</td>\n",
       "      <td>17.572</td>\n",
       "      <td>14.300</td>\n",
       "      <td>30.809</td>\n",
       "      <td>11.163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11.838</td>\n",
       "      <td>13.354</td>\n",
       "      <td>9.617</td>\n",
       "      <td>7.479</td>\n",
       "      <td>22.340</td>\n",
       "      <td>12.156</td>\n",
       "      <td>29.614</td>\n",
       "      <td>9.487</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>5.732</td>\n",
       "      <td>6.175</td>\n",
       "      <td>6.480</td>\n",
       "      <td>4.373</td>\n",
       "      <td>5.058</td>\n",
       "      <td>6.283</td>\n",
       "      <td>6.177</td>\n",
       "      <td>6.156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>9.734</td>\n",
       "      <td>23.887</td>\n",
       "      <td>6.872</td>\n",
       "      <td>7.048</td>\n",
       "      <td>15.326</td>\n",
       "      <td>12.707</td>\n",
       "      <td>13.424</td>\n",
       "      <td>12.530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>7.678</td>\n",
       "      <td>7.417</td>\n",
       "      <td>17.119</td>\n",
       "      <td>14.781</td>\n",
       "      <td>10.409</td>\n",
       "      <td>13.274</td>\n",
       "      <td>44.967</td>\n",
       "      <td>13.194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>13.519</td>\n",
       "      <td>34.927</td>\n",
       "      <td>11.123</td>\n",
       "      <td>16.664</td>\n",
       "      <td>26.592</td>\n",
       "      <td>37.562</td>\n",
       "      <td>17.396</td>\n",
       "      <td>24.245</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>4.110</td>\n",
       "      <td>8.950</td>\n",
       "      <td>9.537</td>\n",
       "      <td>6.993</td>\n",
       "      <td>15.761</td>\n",
       "      <td>8.436</td>\n",
       "      <td>8.331</td>\n",
       "      <td>11.702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>9.339</td>\n",
       "      <td>15.576</td>\n",
       "      <td>12.989</td>\n",
       "      <td>11.909</td>\n",
       "      <td>51.422</td>\n",
       "      <td>23.992</td>\n",
       "      <td>20.569</td>\n",
       "      <td>42.055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>39.382</td>\n",
       "      <td>8.934</td>\n",
       "      <td>7.427</td>\n",
       "      <td>7.880</td>\n",
       "      <td>12.249</td>\n",
       "      <td>10.734</td>\n",
       "      <td>45.501</td>\n",
       "      <td>10.773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>9.282</td>\n",
       "      <td>8.398</td>\n",
       "      <td>5.043</td>\n",
       "      <td>4.728</td>\n",
       "      <td>28.998</td>\n",
       "      <td>10.566</td>\n",
       "      <td>10.888</td>\n",
       "      <td>6.618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>14.657</td>\n",
       "      <td>12.528</td>\n",
       "      <td>13.305</td>\n",
       "      <td>17.442</td>\n",
       "      <td>14.966</td>\n",
       "      <td>12.585</td>\n",
       "      <td>26.383</td>\n",
       "      <td>55.478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>5.646</td>\n",
       "      <td>6.652</td>\n",
       "      <td>18.717</td>\n",
       "      <td>7.854</td>\n",
       "      <td>7.042</td>\n",
       "      <td>9.678</td>\n",
       "      <td>10.959</td>\n",
       "      <td>14.056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>20.224</td>\n",
       "      <td>10.258</td>\n",
       "      <td>14.566</td>\n",
       "      <td>9.842</td>\n",
       "      <td>14.597</td>\n",
       "      <td>7.352</td>\n",
       "      <td>8.555</td>\n",
       "      <td>7.309</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>10.117</td>\n",
       "      <td>16.352</td>\n",
       "      <td>10.577</td>\n",
       "      <td>11.575</td>\n",
       "      <td>9.776</td>\n",
       "      <td>10.035</td>\n",
       "      <td>10.053</td>\n",
       "      <td>9.177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>9.035</td>\n",
       "      <td>9.623</td>\n",
       "      <td>8.552</td>\n",
       "      <td>7.316</td>\n",
       "      <td>9.283</td>\n",
       "      <td>13.579</td>\n",
       "      <td>12.380</td>\n",
       "      <td>11.319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>24.194</td>\n",
       "      <td>30.939</td>\n",
       "      <td>20.398</td>\n",
       "      <td>14.492</td>\n",
       "      <td>31.044</td>\n",
       "      <td>37.365</td>\n",
       "      <td>28.279</td>\n",
       "      <td>15.788</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>19.223</td>\n",
       "      <td>13.901</td>\n",
       "      <td>10.663</td>\n",
       "      <td>23.811</td>\n",
       "      <td>14.259</td>\n",
       "      <td>22.319</td>\n",
       "      <td>11.861</td>\n",
       "      <td>12.069</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>10.902</td>\n",
       "      <td>29.985</td>\n",
       "      <td>20.251</td>\n",
       "      <td>25.949</td>\n",
       "      <td>28.447</td>\n",
       "      <td>36.307</td>\n",
       "      <td>22.822</td>\n",
       "      <td>30.817</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>21.343</td>\n",
       "      <td>15.039</td>\n",
       "      <td>16.091</td>\n",
       "      <td>27.936</td>\n",
       "      <td>61.486</td>\n",
       "      <td>27.852</td>\n",
       "      <td>31.846</td>\n",
       "      <td>29.410</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>8.212</td>\n",
       "      <td>11.478</td>\n",
       "      <td>29.560</td>\n",
       "      <td>15.845</td>\n",
       "      <td>44.715</td>\n",
       "      <td>11.319</td>\n",
       "      <td>21.013</td>\n",
       "      <td>11.835</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>7.583</td>\n",
       "      <td>11.575</td>\n",
       "      <td>20.982</td>\n",
       "      <td>10.124</td>\n",
       "      <td>38.788</td>\n",
       "      <td>44.747</td>\n",
       "      <td>42.904</td>\n",
       "      <td>25.588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>16.300</td>\n",
       "      <td>19.988</td>\n",
       "      <td>22.802</td>\n",
       "      <td>24.525</td>\n",
       "      <td>32.172</td>\n",
       "      <td>55.120</td>\n",
       "      <td>51.922</td>\n",
       "      <td>76.430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>13.842</td>\n",
       "      <td>19.970</td>\n",
       "      <td>25.956</td>\n",
       "      <td>21.399</td>\n",
       "      <td>50.622</td>\n",
       "      <td>56.785</td>\n",
       "      <td>58.758</td>\n",
       "      <td>42.281</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>4.175</td>\n",
       "      <td>7.650</td>\n",
       "      <td>6.031</td>\n",
       "      <td>4.373</td>\n",
       "      <td>4.236</td>\n",
       "      <td>9.535</td>\n",
       "      <td>12.401</td>\n",
       "      <td>3.961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>13.043</td>\n",
       "      <td>29.417</td>\n",
       "      <td>15.970</td>\n",
       "      <td>20.207</td>\n",
       "      <td>20.310</td>\n",
       "      <td>28.922</td>\n",
       "      <td>11.677</td>\n",
       "      <td>29.362</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>4.561</td>\n",
       "      <td>15.671</td>\n",
       "      <td>13.224</td>\n",
       "      <td>6.077</td>\n",
       "      <td>34.899</td>\n",
       "      <td>23.767</td>\n",
       "      <td>45.007</td>\n",
       "      <td>57.634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>25.987</td>\n",
       "      <td>21.697</td>\n",
       "      <td>25.157</td>\n",
       "      <td>16.268</td>\n",
       "      <td>25.159</td>\n",
       "      <td>24.812</td>\n",
       "      <td>46.842</td>\n",
       "      <td>22.615</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>10.288</td>\n",
       "      <td>10.971</td>\n",
       "      <td>10.514</td>\n",
       "      <td>12.494</td>\n",
       "      <td>19.202</td>\n",
       "      <td>21.965</td>\n",
       "      <td>25.297</td>\n",
       "      <td>13.241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>4.501</td>\n",
       "      <td>5.051</td>\n",
       "      <td>4.184</td>\n",
       "      <td>4.068</td>\n",
       "      <td>4.128</td>\n",
       "      <td>3.881</td>\n",
       "      <td>5.637</td>\n",
       "      <td>4.769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>3.929</td>\n",
       "      <td>11.073</td>\n",
       "      <td>9.399</td>\n",
       "      <td>8.395</td>\n",
       "      <td>5.261</td>\n",
       "      <td>4.867</td>\n",
       "      <td>4.465</td>\n",
       "      <td>5.851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>8.557</td>\n",
       "      <td>12.180</td>\n",
       "      <td>7.963</td>\n",
       "      <td>7.330</td>\n",
       "      <td>9.715</td>\n",
       "      <td>14.059</td>\n",
       "      <td>10.139</td>\n",
       "      <td>6.388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>6.665</td>\n",
       "      <td>6.969</td>\n",
       "      <td>7.196</td>\n",
       "      <td>4.274</td>\n",
       "      <td>12.033</td>\n",
       "      <td>13.067</td>\n",
       "      <td>8.127</td>\n",
       "      <td>11.806</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>6.500</td>\n",
       "      <td>7.206</td>\n",
       "      <td>6.030</td>\n",
       "      <td>10.046</td>\n",
       "      <td>5.974</td>\n",
       "      <td>6.951</td>\n",
       "      <td>7.649</td>\n",
       "      <td>6.117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>9.207</td>\n",
       "      <td>11.666</td>\n",
       "      <td>12.430</td>\n",
       "      <td>7.110</td>\n",
       "      <td>17.006</td>\n",
       "      <td>20.848</td>\n",
       "      <td>10.535</td>\n",
       "      <td>12.446</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>4.412</td>\n",
       "      <td>18.003</td>\n",
       "      <td>30.385</td>\n",
       "      <td>5.393</td>\n",
       "      <td>5.205</td>\n",
       "      <td>6.478</td>\n",
       "      <td>4.095</td>\n",
       "      <td>4.957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>2.710</td>\n",
       "      <td>4.406</td>\n",
       "      <td>3.758</td>\n",
       "      <td>2.865</td>\n",
       "      <td>5.851</td>\n",
       "      <td>6.253</td>\n",
       "      <td>7.363</td>\n",
       "      <td>3.783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>3.819</td>\n",
       "      <td>7.717</td>\n",
       "      <td>11.749</td>\n",
       "      <td>4.819</td>\n",
       "      <td>13.605</td>\n",
       "      <td>17.785</td>\n",
       "      <td>15.294</td>\n",
       "      <td>11.594</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>38.515</td>\n",
       "      <td>25.681</td>\n",
       "      <td>17.514</td>\n",
       "      <td>9.155</td>\n",
       "      <td>29.904</td>\n",
       "      <td>14.229</td>\n",
       "      <td>11.193</td>\n",
       "      <td>14.480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>13.610</td>\n",
       "      <td>12.074</td>\n",
       "      <td>14.245</td>\n",
       "      <td>21.149</td>\n",
       "      <td>46.125</td>\n",
       "      <td>63.980</td>\n",
       "      <td>21.224</td>\n",
       "      <td>54.346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>154</th>\n",
       "      <td>13.622</td>\n",
       "      <td>15.070</td>\n",
       "      <td>11.265</td>\n",
       "      <td>6.736</td>\n",
       "      <td>14.081</td>\n",
       "      <td>13.113</td>\n",
       "      <td>16.783</td>\n",
       "      <td>10.037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>6.357</td>\n",
       "      <td>8.718</td>\n",
       "      <td>6.011</td>\n",
       "      <td>3.740</td>\n",
       "      <td>18.930</td>\n",
       "      <td>8.051</td>\n",
       "      <td>9.223</td>\n",
       "      <td>36.023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>5.360</td>\n",
       "      <td>11.519</td>\n",
       "      <td>10.209</td>\n",
       "      <td>6.990</td>\n",
       "      <td>13.451</td>\n",
       "      <td>14.365</td>\n",
       "      <td>14.837</td>\n",
       "      <td>19.510</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>3.383</td>\n",
       "      <td>12.178</td>\n",
       "      <td>4.913</td>\n",
       "      <td>4.809</td>\n",
       "      <td>7.378</td>\n",
       "      <td>11.100</td>\n",
       "      <td>9.827</td>\n",
       "      <td>8.625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>10.000</td>\n",
       "      <td>8.908</td>\n",
       "      <td>8.761</td>\n",
       "      <td>4.774</td>\n",
       "      <td>14.120</td>\n",
       "      <td>10.698</td>\n",
       "      <td>10.992</td>\n",
       "      <td>10.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>166</th>\n",
       "      <td>10.038</td>\n",
       "      <td>18.571</td>\n",
       "      <td>7.737</td>\n",
       "      <td>11.921</td>\n",
       "      <td>12.090</td>\n",
       "      <td>13.363</td>\n",
       "      <td>16.262</td>\n",
       "      <td>9.098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>34.071</td>\n",
       "      <td>33.947</td>\n",
       "      <td>23.260</td>\n",
       "      <td>27.477</td>\n",
       "      <td>37.389</td>\n",
       "      <td>35.180</td>\n",
       "      <td>36.904</td>\n",
       "      <td>31.392</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>4.458</td>\n",
       "      <td>4.822</td>\n",
       "      <td>4.658</td>\n",
       "      <td>4.772</td>\n",
       "      <td>6.428</td>\n",
       "      <td>7.386</td>\n",
       "      <td>7.333</td>\n",
       "      <td>5.377</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "mode          RD                         SQL                     \n",
       "pattern        1      2      3      4      1      2      3      4\n",
       "worker_id                                                        \n",
       "0         10.743  8.359  6.698  6.008 14.905 36.797 59.405 20.626\n",
       "2          5.763  8.801 13.316 12.110 21.611 13.085 13.359  9.433\n",
       "5          4.341 11.566 12.451  5.909 16.144 17.267 22.487 15.869\n",
       "6          4.598 12.098 10.607  9.812 17.572 14.300 30.809 11.163\n",
       "10        11.838 13.354  9.617  7.479 22.340 12.156 29.614  9.487\n",
       "17         5.732  6.175  6.480  4.373  5.058  6.283  6.177  6.156\n",
       "21         9.734 23.887  6.872  7.048 15.326 12.707 13.424 12.530\n",
       "28         7.678  7.417 17.119 14.781 10.409 13.274 44.967 13.194\n",
       "32        13.519 34.927 11.123 16.664 26.592 37.562 17.396 24.245\n",
       "39         4.110  8.950  9.537  6.993 15.761  8.436  8.331 11.702\n",
       "42         9.339 15.576 12.989 11.909 51.422 23.992 20.569 42.055\n",
       "43        39.382  8.934  7.427  7.880 12.249 10.734 45.501 10.773\n",
       "50         9.282  8.398  5.043  4.728 28.998 10.566 10.888  6.618\n",
       "52        14.657 12.528 13.305 17.442 14.966 12.585 26.383 55.478\n",
       "57         5.646  6.652 18.717  7.854  7.042  9.678 10.959 14.056\n",
       "58        20.224 10.258 14.566  9.842 14.597  7.352  8.555  7.309\n",
       "60        10.117 16.352 10.577 11.575  9.776 10.035 10.053  9.177\n",
       "66         9.035  9.623  8.552  7.316  9.283 13.579 12.380 11.319\n",
       "72        24.194 30.939 20.398 14.492 31.044 37.365 28.279 15.788\n",
       "75        19.223 13.901 10.663 23.811 14.259 22.319 11.861 12.069\n",
       "77        10.902 29.985 20.251 25.949 28.447 36.307 22.822 30.817\n",
       "80        21.343 15.039 16.091 27.936 61.486 27.852 31.846 29.410\n",
       "81         8.212 11.478 29.560 15.845 44.715 11.319 21.013 11.835\n",
       "83         7.583 11.575 20.982 10.124 38.788 44.747 42.904 25.588\n",
       "87        16.300 19.988 22.802 24.525 32.172 55.120 51.922 76.430\n",
       "89        13.842 19.970 25.956 21.399 50.622 56.785 58.758 42.281\n",
       "91         4.175  7.650  6.031  4.373  4.236  9.535 12.401  3.961\n",
       "92        13.043 29.417 15.970 20.207 20.310 28.922 11.677 29.362\n",
       "96         4.561 15.671 13.224  6.077 34.899 23.767 45.007 57.634\n",
       "110       25.987 21.697 25.157 16.268 25.159 24.812 46.842 22.615\n",
       "115       10.288 10.971 10.514 12.494 19.202 21.965 25.297 13.241\n",
       "117        4.501  5.051  4.184  4.068  4.128  3.881  5.637  4.769\n",
       "119        3.929 11.073  9.399  8.395  5.261  4.867  4.465  5.851\n",
       "121        8.557 12.180  7.963  7.330  9.715 14.059 10.139  6.388\n",
       "125        6.665  6.969  7.196  4.274 12.033 13.067  8.127 11.806\n",
       "130        6.500  7.206  6.030 10.046  5.974  6.951  7.649  6.117\n",
       "136        9.207 11.666 12.430  7.110 17.006 20.848 10.535 12.446\n",
       "141        4.412 18.003 30.385  5.393  5.205  6.478  4.095  4.957\n",
       "143        2.710  4.406  3.758  2.865  5.851  6.253  7.363  3.783\n",
       "146        3.819  7.717 11.749  4.819 13.605 17.785 15.294 11.594\n",
       "148       38.515 25.681 17.514  9.155 29.904 14.229 11.193 14.480\n",
       "153       13.610 12.074 14.245 21.149 46.125 63.980 21.224 54.346\n",
       "154       13.622 15.070 11.265  6.736 14.081 13.113 16.783 10.037\n",
       "158        6.357  8.718  6.011  3.740 18.930  8.051  9.223 36.023\n",
       "159        5.360 11.519 10.209  6.990 13.451 14.365 14.837 19.510\n",
       "162        3.383 12.178  4.913  4.809  7.378 11.100  9.827  8.625\n",
       "165       10.000  8.908  8.761  4.774 14.120 10.698 10.992 10.002\n",
       "166       10.038 18.571  7.737 11.921 12.090 13.363 16.262  9.098\n",
       "168       34.071 33.947 23.260 27.477 37.389 35.180 36.904 31.392\n",
       "169        4.458  4.822  4.658  4.772  6.428  7.386  7.333  5.377"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RD, 1: 9.245, 95% CI [6.665, 10.471]\n",
      "RD, 2: 11.620, 95% CI [10.258, 13.354]\n",
      "RD, 3: 10.893, 95% CI [9.508, 13.224]\n",
      "RD, 4: 8.137, 95% CI [7.021, 11.575]\n",
      "SQL, 1: 15.146, 95% CI [13.605, 19.756]\n",
      "SQL, 2: 13.319, 95% CI [11.951, 17.526]\n",
      "SQL, 3: 14.130, 95% CI [11.193, 21.013]\n",
      "SQL, 4: 11.952, 95% CI [10.600, 14.922]\n",
      "ratio, 1: 0.639, 95% CI [0.486, 0.779]\n",
      "ratio, 2: 0.830, 95% CI [0.698, 0.965]\n",
      "ratio, 3: 0.662, 95% CI [0.533, 0.768]\n",
      "ratio, 4: 0.712, 95% CI [0.598, 0.859]\n"
     ]
    }
   ],
   "source": [
    "# Create df8, df9\n",
    "df0 = df_filtered_data[['worker_id', 'pattern', 'time', 'mode']]\n",
    "# print(df)\n",
    "\n",
    "df8 = df0.groupby(['worker_id', 'mode', 'pattern']).time.agg(['median'])         # for each worker, calculate median for both modes\n",
    "df8.reset_index(inplace=True)\n",
    "# print('df8:')\n",
    "# display(df8)\n",
    "\n",
    "# Pivot to have one row per worker\n",
    "df9 = pd.pivot_table(df8, values=['median'], index=['worker_id'], columns=['mode', 'pattern'])\n",
    "df9=df9.droplevel(0, axis=1)\n",
    "print('df9:')\n",
    "display(df9)\n",
    "\n",
    "# Median of median task time for each mode and section (for error plots)\n",
    "modes = ['RD', 'SQL']\n",
    "patterns = [1, 2, 3, 4]\n",
    "median_time = {}\n",
    "ci = {}\n",
    "ci_delta = {}\n",
    "for mode in modes:\n",
    "    for pattern in patterns:\n",
    "        column = (mode, pattern)\n",
    "        median_time[column] = np.median(df9[column])\n",
    "        ci[column] = scipybootstrap((df9[column],), statistic=np.median,\n",
    "                                    n_resamples=BOOTSTRAPSAMPLES,\n",
    "                                    confidence_level=BOOTSTRAPCONFIDENCE,\n",
    "                                    method='percentile',\n",
    "                                    axis=0).confidence_interval        #convert array to sequence\n",
    "        ci_delta[column] = [median_time[column] - ci[column].low, ci[column].high - median_time[column]]\n",
    "        print(f'{mode}, {pattern}: {median_time[column]:.3f}, 95% CI [{ci[column].low:.3f}, {ci[column].high:.3f}]')\n",
    "\n",
    "# Median ratio RD/SQL per pattern\n",
    "for pattern in patterns:\n",
    "    column = ('ratio', pattern)\n",
    "    df9['ratio', pattern] = df9['RD', pattern] / df9['SQL', pattern]\n",
    "    median_time[column] = np.median(df9[column])\n",
    "    ci[column] = scipybootstrap((df9[column],), statistic=np.median,\n",
    "                                n_resamples=BOOTSTRAPSAMPLES,\n",
    "                                confidence_level=BOOTSTRAPCONFIDENCE,\n",
    "                                method='percentile',\n",
    "                                axis=0).confidence_interval        #convert array to sequence\n",
    "    ci_delta[column] = [median_time[column] - ci[column].low, ci[column].high - median_time[column]]\n",
    "    print(f'ratio, {pattern}: {median_time[column]:.3f}, 95% CI [{ci[column].low:.3f}, {ci[column].high:.3f}]')\n",
    "\n",
    "\n",
    "# print(median_time)\n",
    "# print(ci)\n",
    "# print(ci_delta)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "29006886",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA10AAAMSCAYAAACRda7mAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdd5gkV30v/G9VdQ6Tw+ZdrbKQtMqrAAYEskS0DAgsAeJiuAgjWyIYA74Xm2teX3QJBgMCC4MBmSAQiGAkhKwASCisEqtVXG2ajbM7O6lzV3z/6K6a6p7OXdXx+3m2nsrnnDpVPdu/PlWnBMMwDBAREREREZErxHYXgIiIiIiIqJcx6CIiIiIiInIRgy4iIiIiIiIXMegiIiIiIiJyEYMuIiIiIiIiFzHoIiIiIiIichGDLiIiIiIiIhf1XdD1wgsv4IUXXmh3MYiIiIiIqE/0XdAlyzLS6XS7i9F1Dh8+3O4idCXWW/1YZ/VjnTWG9VY/1lljWG/1Y53Vj3XWmFbVW98FXQBgGEa7i9B1MplMu4vQlVhv9WOd1Y911hjWW/1YZ41hvdWPdVY/1lljWlVvfRl0ERERERERtQqDLiIiIiIiIhcx6CIiIiIiInIRgy4iIiIiIiIXMegiIiIiIiJyEYMuIiIiIiIiFzHoIiIiIiIichGDLiIiIiIiIhcx6CIiIiIiInIRgy4iIiIiIiIXMegiIiIiIiJyEYMuIiIiIiIiFzHoIiIiIiIichGDLiIiIiIiIhcx6CIiIiIiInIRgy4iIiIiIiIXMegiIiIiIiJyEYMuIiIiIiIiFzHoIiIiIiIichGDLiIiIiIiIhd52l0Aon6l6zo0TYOqqhXH5nalxua0YRjQdR2GYVjT5rw5BlB2bCcIgjW2T5tjURStsX3avkwURUiSZI3t0/ZlHo8HHo+nYN48HjNPIiIiom7HoIuoBMMwoCgKZFlGNpuFLMsF0+Y6+7h4maqqSCQSkCQJiqJAVVVrbAZUVNoTTzwBSZKwefNmKzAzB6/XWzBdavD5fCXnfT6fNZjzkiS1+3CJiIioxzHoop5iGAZkWUYmk7GGbDaLdDqNbDa7bDDX2wczwCrVCuQmURQhShJEUSoaixAEEYKtFUkQRIiiBEEUrHVmy1RuWJqH1WIlINd4tNSCVDwP2Fu/DOQmjfy//NhYalEzDB2Gbixbpus6DLO1TdegG0Z+XrOW67oGXdPyYx2GoVtl2L17NwBg06ZN7lV4niRJBcGY3+8vOe/3+62heN4+MIgjIiKiYgy6qOMYhmEFSqlUCul02hoymUzBuHhZNpuFruvVM6mD5PHC4/XB4/XC48kNktcLSfLk1nm8kDy5acnjgSiIeObxBwAAZ73sUvj8fkiSB6Lkye+Tu51OlDwQJQmSKFlBUz+zB2N/2PIkDMPAqy5/JzRNg6ap0DUVen5a0zTomgpNVXPzqgpNVWzTKlRVsS3LtTBqimItN4NLTdOsa8kJHo/HCsACgUDJwCwQCFjr7NvYl4siH7klIiLqFQy6yFWyLFvBU/FQHFSZ05lMpunASRBEeH0+eH1+eHx+eL35aa8PXp8vH0T54PUuTVuBlTmdD6LqDYZURbGCrpVrN8Lj9TZ1LP1CEEVIoggJHqu1KDI47EpeuWfdNCtYU/PBmKrIUBUlv0zOL7NPy7ZBKRhrmgoA1u2jyWSyqTKarWnFAVqp6Xg8Dl3XC5ax1Y2IiKhzMOiimmmahlQqhWQyWTA2h1Lzqqo2nJ/k8cDrC8Dn98Pr88PrC8Dr8+fnc9Nevx8+c9rnhycfaElS/cES9Q9BEHKtjpIH8AccSVPXdWiqAkXOQlVkKIoMVZYLphUlC1Up3iZbsF7PP+tn3uYaj8cbLpPP56sasJVqdbMvY+BGRETUPAZdfUxRFCSTSStYSiQSVrBkD6ySySQSiQRkWW4oH0EQ4QsE4PObQxA+fwBev31ZIB9gBXKBlD+Q+0JM1CVEUYSYD/6boWtaQTBmBWhm0Cbbl2WXtrWtN1vdzMCtGV6vd1lAVuq2yGq3VPJ2SSIi6mf8VttHFhYW8POf/xyzs7NIpVKNfRkTBPj8Afj9QfgCQfgC+el8IJULrvLr8sGUx+vrq1YnQRSwYu1Ga5qoHqIkwS8F4Q8EG06jeuCWLR3E5bdRlSy0fCu12TNnIpFo6rjsHZOUG+zrS3ViYk5LktRXf1OIiKj7MejqIy+88AJ27txZsEwQRfgDuS94Pn8Q/kAoH0zZly0FV15/gF92qpAkDzZf/Pp2F4P6mCOBm67lb3mUrVY11QrKbAFb8W2USuEtlbpeeLtks8EbkGtVrNTjZKmh+JUBPp8P8/PzGBwcLHi9AP++ERGRG+oOuj7+8Y/jZz/7WeVEPR4Eg0FMTExg06ZNuOSSS3DxxRfXlc/WrVvxF3/xF1i5ciXuvffeeotJJZi9tY2vXIvTz38l/IFg37VCEVFtRFGyfoBphqapuUBMznVGosgyVHVpXs3Pm8FcYYcmckFHJmbrm67r1ishnFb8nrfid75VGorfHVdq3lwmssdSIqK+4kpLl6qqiMfjiMfj2LlzJ2677Tacd955uPHGGzEwMFB1/2Qyif/9v/+3411/U47PH0BkYKjdxSCiPmB2WOIPhJpOy9D1pR4l1cJeJjWlsMfJ3GsCcoGaqsjWawSW9su/TkBRrNY4YOl2SrcJgrAsECt++XelQZIkeL1eSJJUdn2pafs8Az8iotZpKui6/vrrccIJJyxbrmkaEokEnnvuOfzkJz9BOp3Gli1bcO211+Lmm2+u+EdelmV84AMfwPbt25spGhER9RhBFK2eSp1k6DpU+/velr3nrdS74GzTJca6fT/NXLYU3Jkvcm+2o5Nm5HrxlKwgrHi6eCxJErLZLAYGBgqWVUqj3PpaljMoJKJe0lTQdfbZZ2Pz5s0Vt3nb296Gq666CrFYDFu2bMGvf/1rvPa1ry257ezsLK6//no8+uijzRSLqK00TcXWh+4DAGy64JXshZGowwmiCK+Ye2+fmwreD2e+7NsepGkadPPF3wXj3EvBFSWLnc88CQBYe+zJMAw9v70GTS/cXtfN/ZamdU2DYegF5THfK5fNZl099kaVCsbcHERRLAgEzflq25sDg0QiKsf1b4PHH388rrnmGnzuc58DAPzyl78sGXQ99NBD+PjHP47p6Wm3i0TkKkM3sG/ncwCA0ze/AuBrjogIRe+Ha4CqKFbQdfrmVzT04nVD1/MBmhmM6UtBmb48YDN0M6DLb2cL4HLzujVt6Lq1j7lc0wvndd2en55LX8vta7/N06Tly9Et7EGYfVrXdavnzXJBW6kAz1xWvK54eantalkmiuKyZYIgMHgkckFLfoK/9NJLraDrmWeeKVg3NzeHT37yk7j77rutZW9+85vxwAMP4PDhw60oHhERUV8QRBEeUQQ89QdsbjMMw2oNNGzBXHHAZhQFbtY2BfstLTeKts3trxekbZRKx5ZPrlVx+TbF9Px+rXgu0E32gKw4MCu3rNT6avubt5CWWr64uIiDBw+W3MY+b94may4rXldpXG5dpWkGpNSolgRdk5OT1vT8/HzBuhdffNEKuMLhMD7xiU/giiuuqLu3QyIiIupe9i+43WBZkGgFeSWCRMMe3OlLwaB9n2UBn77USmhuYyytL7WPNTaK9tN1GIZeYrvSHZZ1WwtjK9USlBXPV1pebttGhlQqhUgk4lh69iDTrTLXO1TLu546Npe36lpvSdAVj8et6Wg0umy9KIr4sz/7M3zoQx8qCNCIiIgohy9e7yzdFiSWYgaOZtCYCyJzwd7S9FKAtix4M3QYurG0Lh9gmsuW5nUrn6V0jKXl+WUwjKL982PbtgXpWets00Zx2vZ1RsE6wNzXKBuAlqsz9rDdWz796U+7nkdLgq6HH37Ymj7llFMK1m3YsAG/+c1vsG7dulYUhch1kseD1155jTVNROQEvnidnGa1ZIgipNZ8JexoZkCFksGajlycZptGPoADrCARMAoCPBgGDBSnm3tvam5fAzBseZfYNreNAcNA6W1saSzLD7C2WyobCrezbVO4XT5dLKVbsayVymE/7pLrzXxsdVtUBvOW3sJzYa+fonovWtdurn/C4vE4vvzlL1vzr3vd6wrWs2Wr9RRFRmx+tt3FICpLyXejzeuUiIi6i3n7WrvLQaUUt3oCrQvIXAm6stksDh8+jEceeQQ33XQT9u3bBwA46aST8IY3vMGNLHvOs88+i6eeesrRNGdnc19gjxyYwpEDU46m3Q+eeOIJAMBZZ53V5pL0viMHc9fnfb/8fptLQkRERL3u0gs3uZ5HU0HX1VdfXfO269atw1e+8hV4G+jith/deeedyzodofbavXs3AAZdrXDMMce0uwhEROSiUChUsSdATdMgSXznSj1YZ43pmY401q5di8svvxxXX301BgYG3M6uZ5gXwOYTJzEQ6sxAVdN03PvUQQDAxaevgiR178PEtXjgniAA4JIz17S5JO3VivPe73VMVMwwAN0woOn5QdOXpvODrueW2bfT84Nm5Mf59bqRe6eglu+EwFym5zsVMOd7gSAAYv52LwGCbV6AKOaeaxIFc7y03j5duK1te7FwX/vysumUyEsQCsvRi7emGQag6Qbue+oAAODDH/4w/H5/2e2npqawfv36VhWvJ7DOGjM11Zq7v5oKuq6//nqccMIJy5b7/X5Eo1GsW7cOIyMjzWTR905aM4TJ4VC7i1GSrGrWl+/TN47B5+ntX1eGI7n/HM48drzNJWmvfjvvRCZdN6DqOlRNh6oZULTctGabVrXcNppm5OeXlqn5QMlcpun5deZyc7/ioErTofZKBNQGhgFo1jMb3VOPZpBoD8QEQYAA+/xSIAn7clsigi29gvRRObLLd8GQ7+AgPwHkO43Id6YA2Do6yP0wYK3Lj3XbtkT9rKmg6+yzz8bmzZudKgsREVHNjHyLjj3gUdTc9NIyczoX3JjrzUCocHujaF974KR3VMuPIOReCOvxeCBJUtnB/lLacuNKL6u1v8tGFEUsLCxgdHS06vt8Ko3LsT/Mbu/lrNJ08WB2Ra7bugsvNV1qbL3Xqsp8uWWV0jTvXrGvq8bszU3P9+TWS9asWQOfz9fuYhC1FPsHpYaJgoDjVg1a09QfeN6pGjMYMgMYRc0HMvmxYguA7MvtAU8uGDLKLF9a1y4ej8cavF5vwbjSdLllkiTB6/VagZS5rHjaHLfr3VC8fakxxfVWLQisNF0cZNrTK+yVDcvGxdPmfLmAuFwwXSrYLvWy4HJjv99fNQgn6jUMuqhhHknEmy7c2O5iUIvxvPcWTdchKzpkVYOs6pCV3FjRcssUVc8tV3Uo+fncunzQpOpQtKXl5rjVr0QRRRFer9caPB4PfD6fFeDYAyNzOpVKYWxsbNl+labNIKibX4hL7dcLL1Ymovow6CIi6lKKpiMra8goan6sIbts0K1pWdWsACubH2su3zNnBkNmAOTz+axAxj5dz1AqoGqkxy622hARUasw6CIiajPdMJCRVaSzGlJZFWlZRdoa54KqtKwhk1WRVjRkZBUZ2dmAyQyC/H5/wbh4MLerNm3Os/tiIiIiBl1ERK7QDQOprIpkRkEyY45z06msilR2aTqdVRt+TF4QBPh8PoTDYQQCgYLB7/cXTJvz5rQ9sGJwRERE5B4GXdQwVdNx1xP7AAB/etZaeHr8PV2Uw/Oeu60vnpIRTyuIp2XEUwoSGQWJtIJ4OjedzCh1P9cUCAQQCoUQDoetcTAYRDAYRCgUKhgHg0EEAgH4fD7s27ePt8kRERF1MAZd1DDdMPD01BwA4NV8mW3f6PXzbhgGMoqGxaSMWEpeGqdy43hKRlqu7e31giAgHA4jEokgEokgGo0iHA5bY3OIRCIIhUJsbSIiIupRDLqIqO/ouoHFlIyFRBYLySwWEjIWkllrmaxW74rc6/VicHAQg4ODGBgYsIZoNGqNw+EwAykiIiKqP+i64YYbcMMNN7hRlgL33nuv63kQUe/SDQOxlIz5eBZz8QzmE9nckJQRS2arvug2EolgaGho2WAGWoFAgO+ZISIiopqwpYuIupqsapiLZzEby2A2nsFcPIO5eC7AqtS7n8fjwcjICEZGRjA8PGyNh4eHMTQ0BJ/P18KjICIiol7GoIsa5pVEXPfG06xp6g/tOu9ZRcNsLIOjsQyOxtLWdDytlN1HkiSMjIxgbGwMo6OjGBsbw8jICEZHRxGJRPhiUiIiImoJBl3UMEEQEPDxEuo3bp93VdMxG8/g6GIGM4tpzCymqwZX4XAY4+PjGBsbs8ZjY2MYGhpiYEVERERtx2/MRNQWhmEgnlZwZCEXWB3JB1jziWzZrtaj0SgmJiasYXx8HOPj4wiFQq0tPBEREVEdGHQRketUTcdsLIMji2kcWUhbgVZGKd31ejAYxOTkJCYnJzExMWGNg8Fgi0tORERE1DwGXdQw3TCwfyYBAFgzHoHIntz6QrXznpFVHFlI4/BCGkcWUjiymHv+qlSfFqIoYnx8HJOTk1ixYgVWrFgBRVFw8skns2dAIiIi6hkMuqhhqqbjlt/vAAB88PLT4fPwfUT9wH7e/+elp2AukcHhhTQOz6dwZCGNxZRccr9gMIgVK1Zg5cqVVoA1Pj4Oj6fwz9DU1BQDLiIiIuopDLqIqCrDyL1M+PB8Ggfnktbyf//NsyW3HxoawsqVK60Aa9WqVRgYGGAwRURERH2JQVeHy5Z55oXILbphYD6exeGFFA7Pp3PjhXTJa1EQBIyNjWHVqlVWkLVy5Uo+e0VERERkw6CrQ5ndXN/y+x0YjQawbiKC9RNRrB2PIMhu2skhmm5gNpZ7/sp+i6Ci6cu2lSQJk5OTVmC1atUqTE5O8iXCRERERFXw23uHuuSSS/CHP/wBhw4dxGw8g9l4Bk/uPAoAmBgKYu1YBOvGI1jTxiBMEkVcvGm1NU2dTVF1zCwutVwdXkjh6GIGWokeLrxer3VboBlkTUxMLHv+ioiIiIiq4zeoDnX66afj9NNPRyqVwp49e7Br1y7s2rULMzMzVpfbj++YAQCMDwawbjyKNWMRrB4LIxLwtqSMkijgnOMnWpIX1SeZyb3/6kg+yDqykMZ8PItSr78KBAIFrVcrV67E2NgYXypMRERE5BAGXR0uFArhlFNOwSmnnAIAiMfj2LNnD3bv3o09e/ZgZmYGM4sZzCxmrCBsKOLH2rEwVo9GsGYsjOGInx0Y9ChdNzAXX3r/lfmS4WRGLbl9JBIpCK5WrlyJ4eFhXh9ERERELmLQ1WWi0ShOO+00nHbaaQCARCJhBWFTU1M4cuQIFhJZLCSy2LZnDgAQ8ElYNRLODaMhrBwJw+9l9+7dxDAMJDIqjuaDqqOLaczEMpiNlb49UBAEjIyMFHRusWLFCkSj0TaUnoiIiKi/MejqcpFIBKeeeipOPfVUAEA6nca+ffswNTWFqakpHDhwABlZxa7pGHZNx6z9RqMBrBgOYsVwCJPDIUwMBfmerQ5gGAZSWRVHY5n8kMbsYm46U6YnS5/PZ733yhwmJibg9/tbXHoiIiIiKoVBV48JBoM44YQTcMIJJwAAVFXF9PQ09u/fj3379mHfvn2Yn5+3Oud4Zu88AEAAMDoQwMRQEBODQYwPBjExFES4wvNhiqrjW3c9BwB4z5+eDK+HzwDVStdz772ajWUwlz8Xc/EsZisEV4IgYHR0FJOTk9awYsUKDA0N8fkrIiIiog7GoKvHeTwerFmzBmvWrMH5558PIHdL4oEDB3Dw4EFrHI/HrdaVZzFv7R/yezA+GMTYQAAj0QBGo36MDgQQ8ntgwEAsJQMAjJJdNPQ3TdexmJSxmJQxn8gWDIvJLErcFQggF1wNDw9jYmKiYBgbG4PX25pOUoiIiIjIOQy6+lAkEsGJJ56IE0880VoWi8Vw8OBBTE9PW8Pc3BxSWRVTR+KYOhIvSCPglTAS9eOUdcMYCvux/cAChsJ+DIX9CAc8fdExg6LqiKVlxFMyYikFsZSMWErGQlLGYjKLeFqpuL/H48HY2BjGxsYwPj6O8fFxjI2NYXR0lO++IiIiIuohDLoIADAwMICBgQGcdNJJ1jJZlnHkyBEcPnwYO3fuRDabxczMDBYWFpBRNBycS+HgXGpZWpIoIBr0Ihry5cZBHwZCXoQDXoT8Hmvwe6WOC840XUc6qyEtq0hlVaSzKpIZBYmMirEVa5BVNHz7v59DIq0gLZe+DdDO4/FgZGQEIyMjGB0dLRii0ShvCyQiIiLqAwy6qCyfz2fdmjg2Nob169cDABRFwezsLI4ePYq5uTnMzc1hfn4ec3NzWFxchKYbWEjmWnwqkUQBQb8HAa8E/7JBhEeyDwI8kghJFCAKAgQBEJAbi4IAwzCgG7DGumFANwyomg5Vy40VVc+NNR1ZRcsPOuT8dFpWIat62fJOrs+1DM4sZgrqaHBwEENDQxgcHMTg4CCGh4cxMjKC4eFhHD16FBs2bGj+ZBARERFR12LQRXXzer1WL3nFVFVFPB7H4uIiFhcXEYvFEIvFsLi4iEQigWQyiWQyiWw2C003kEgrSFS5Da/VBEFAKBSyhkgkgmg0WjBEIhEMDg4iEAhUbK2bnZ1tYcmJiIiIqBMx6CJHeTweDA8PY3h4uOJ2iqJYAVgmk1k2ZLNZqKoKRVGgKIo1res6dF3PtWjZxqIoQhAEiKJYMHg8Hni9XmvweDzw+XwIBALw+/3w+/0IBAIIBAIIBoMIhUIIBAK87Y+IiIiIHMOgi9rC6/ViaGgIQ0ND7S4KEREREZGr+HM+ERERERGRixh0ERERERERuYhBFxERERERkYsYdBEREREREbmIQRcREREREZGLGHQRERERERG5iEEXERERERGRi/ierj7z2J55vDAdx4krojhnQ+UXGNttn9MwY8wDQE37P7ZnvmC+1LbmNtXKUWq7Wvd1UqU866nXVpTdXh6TE/nVcl4B4PsP7wUAK//i7YrrwF7etcFAxW1rWddIHdeyjxvnrlRdNJuH0+WsN712fD4b5VZZO+1vVD3bVNqv+PPZSBrN1km3XF+dXM5W/C0jd/XS365+waCrz9z0+90NBV23bJ2DKC4AqC3ouun3uwvmS21rblOtHKW2q3VfJ1XKs556bUXZ7eUxOZFfLefVvl25oKu4Duzl/erbXlJx21rWNVLHtezjxrkrVRfN5uF0OetNrx2fz0a5VdZO+xtVzzaV9iv+fDaSRrN10i3XVyeXsxV/y8hdvfS3q18w6Oojj+2Zx+NTCwCAx6cW8Nie+Zo+VI/tmceT+2IFyyrtb8/Hvqy4pcrcplI5Sm1X675OqpRnPfXairIXl8e+vNnWk2rnFci1ciWyWkH+9u2K68C+3eNTC9h2KIVzNvhLblvLNdRIHdeyjxvnrlJdNJqH0+WsN712fD4btX1Oc6WsnfY3qtly2fezfz6dLl8r03FbcTmbaSF0Wiv+lnXqeekVbtU3z6O7+ExXHylupSier3W/RpZXyrtSOUptV+u+TqqUZz312oqy13u+mkm3kWXF08vq73flt601zUplKaWWfdw4d5XqotE8nC5nvem14/PZqFu2zlnTTpa10/5G1bNN1f1+197rsluur+JyDg0Nta8wRVrxt4zc5VZ98zy6i0FXnyjVSmG2ytS7X6X9y21v37Z4m3LlKLXd9x/eW9O+TqpU3nrqtdbjdrKstZSrmXSL07S3cpXarlQdLKu/veW3rXYNNVLHtezjxrmrVheN5OF0OetNrxXXuFOKW/CdKms76sDNa3jZfnvbd112y/VVqpzbDqXaVyCbVv0t68Tz0ivcqm+eR/cx6OoTjbZ+1Lu+llYep1pMai1jsyrl2UxrX6X9G9Xs+Wxkv3paLetpeaq33qvtUymvasvcOHf1tho1mqZTrQq1pNeOz2ej3Cprp/2NqmebRtNuRRpOpuO2kuVssIXQaa36W9aJ56VX9NLfrn7DoKsPNNr6UWm/UvtX275US1W5cpRLq1IrihsqtfCUOxZ7mQKBQNV0nCp7vefLqXTNNMu1ctm3q1a+attWu4bqreNGWmZrSbeaWs5VvXk4Xc5602vFNe4Ut8rajjpw8xp24nicqpNuub7KlrOBFkKntfJvWaedl15Rrb7N7xxOp0vOYNDVBxpt/aj3eZhmf7lv5l7iVj8fVUue9vv4m0mnVo08v+TU9vW0YjXLqV/Iq6VVy3XtRstho9s6XU4nW7Q67ddSt8rajjpw8xp24nicqpNuub46uZyt/lvW7uPtRdXqu9FnB3keW4NBV49rtPWj1l/izf0rtfrYVWsJKfVcTq1laOXzUUDlYzHLtO1QquaWombUe76a/ZW6VJrV6sMpjeTTyDVey3XtRsthI3nYe+FrNA27eq/ZVlzjTnGrrO2oAzevYSeOx6k66Zbrq5PL6UbZOvl4e1Et9d3Is4M8j63DoKvHNfprZLtampppMWn181E1pfG72o6nFc8HNbJ9L/3C1cg13qrnrprdx94LnxP51nvcrbjGneJWWdtRB25ew06k3crPWCdcX51cznb9LeuE89IraqrvBp4d5HlsHQZdPazR1o9GWpqcauWo57mfUvu28vmoWjy+t7bjacXzQfXm51QddIpGrvFarms3Wg7ryeOxPcvfo1dvGsXp1XPNuvF8mlvcKms76sDNa9iJ43GqTrrl+urkcrbyGdJm0qTyaq7vOp8d5HlsLQZdPazRX626+ReNVj8f5aRWt/A58St1t3HrGm9Fy2Gzz17Wum296bnxfJpb3CprO+rAzWvYieNpx/Na7by+OrmcrX6GtJltqbRO+HvE89g8Bl09qtHWj1qfzepUrX4+ykmteD6o1vx6rZXL5NY13oqWw2afvSyXRrPpOf18mlvceJbOzXSdzLMWjT5T6+Z12Y66bUQnl9ONsnXy8faiXvrb1e8YdPWoTnkuqh1a/XyUk1r9LJ0Tv1J3m3b0IudUvs0+e1ltn1ac905+pqWRfdxK1839K6XrxPE4VSftqNtGdHI53ShbJx9vL+qlv139TjAMw2h3IVpp27ZtUFUVZ555ZruLQkREREQlTE1NYf369e0uBvWBVl1rHtdzoIZkn7kfAOB/ycsa3s5cZ6fseQreDafD/5KXIfvM/VD2PAV1ehcAwLNio7Wu2Pz8PIaHhyvmbaYHoGw65cpmlqfcsVRSKq1K29nXV8vTnnalYyuVTuL2G6Ec2onQBX9e9fhqPYZKGq2/etMqdZ4bybvcPrFYDAMDAw2Vu9Y8OlkjZT506BBWrlzZ8P5OlKGdGi2vvd5akV8vaLTOul2z57wX663Z/7P78fPjtl68znoJg64Olbj1MwCq/1GqtJ25zs4edCVu/QyUPU/ByCQBAEIgXDZYisViy4Ku4rzN9IDqQVdx2czylDuWSkqlVWm7gsCoSp72tCsdW7m0jUwS+v7nqx5frcdQSaP1V29apc5zI3mX22d+ft6xoMvJOmmVRsosy3JT+ztRhnZqtLz2emtFfr2g0Trrds2e816st2b/z+7Hz4/bevE66yUMujpQ9pn7IT/7gDVdqRWm3Hb2dcXkZx9A4vYbl603UjHIzz5QMc9yeZvp2vMol06pstnLU0v+ldIqtX+puqpWz+XqsPjYSqWTuP1GGKmYtX2l46v1GGqth3r3rSet4rJWO7ZWlLedeTit2TI7cczdVm+tLm+31Q81j+d8uUbrhHVJ/YwdaXQge6tHqdaqWrartF896daad6l9yqVTbdta8q81rUrpV8uzkbovlXa1vOrNu57yNKKZstaat5PlbWceTmu2zE4cc7fVW6vL2231Q83jOV/Oif+zWZfUbxh0dZhSLQmlns2qtF2lVi6T2QpTSrk8K+VdqUWo0r6lylMt/2ppFe9frXWmln3K5VEq7YVvfWRZ/ZY7vlqPoZJar5lm03Lq3DlZ3nbm4bRmy+zEMXdbvbW6vN1WP9Q8nvPlGq0T1iX1OwZdHaaRlpviZU78etRsS1i5bRtpCWlkm0Zaser9Ba5cC1/6v/+jpn2r5dNMXTnRylG8zKlz52R525mH05ots9Mtpo2m0UqtLm+31Q81j+d8uUbrhHVJ/Y5BVwdptOXGvl2pZ7UaUWsLWz3p1LNvva1t5favpXWm1n1KbV9yO12rad9q56qRVqN69q0nrXquq0p5O1neclqRh9OaLbN/31bHW0wbSaOVWl3ebqsfah7P+XKN1gnrkohBV0dppuWmlnVOlKeZ50yc+NW9nnKUa4lyep9GNduaV219q+u7lu1bce226vPhpGbLLN31tab2d6IMrdbq8nZb/VDzeM6Xa7ROWJdEDLo6RrMtN6ZKz2rVq97nnCql00gLXLOtbWVbohzep1G1nKtGWo1q2bfetOq9rup5nq94n0AgUFdejeTRSZotc/aZ+6E+/2DD+ztRhlZrdXm7rX6oeTznyzVaJ6xLohwGXR3CrZabZjnV05Abzxj1i0ZajerZpp7t6tHI83yJWz+DoaEhx/JsdJtWcvL5xUa36bZ6a3V5u61+qHk858u5+dxWv9Ul9ScGXR3AzZabZpm/QIUPPdNU3o22wDnV2tbNGmk1qrRvMbfqtJHn+eRnHwB2bGkoPyfrpFWaLbMTx9xt9dbq8nZb/VDzeM6Xa7ROWJdESxh0dYBO/4UncetngF9/pb35l5juJ432AlnLtq3o7a2ePOIOtoo6sa2bWtk7YTMtpo1s65ZWl7fb6oeax3O+XKN1wrokWsKgq826oeVGfvYBKM/9oa3519OrYC9qtBfI4n2LuV2njTzPpzjY82KlcrX7V9Vmy+zEMXdbvbW6vN1WP9Q8nvPlGq0T1iVRIQZdbcZfdmrTjufZOk2jvUBW2qcVdepkeduZh9OaLbMTx9xt9dbq8nZb/VDzeM6Xa7ROWJdEhQTDMIx2F6KVtm3bBlVVceaZZ7a7KERE1EZTU1NYv359u4vRVVhnjWG91Y91Vj/WWWNaVW9s6aKazM7OtrsIXSmRSLS7CF3n0KFD7S5C12GdERERdTYGXVQTBg+NYbBaP1mW212ErsM6IyIi6myedheAiMjOMAzMp2SkZQ1Bn4ShoBeCILS7WEREREQNY9BFRB0hnlHwX1un8b2H9mI6vstavmY4iL84dw3esGkFogFvG0tIRERE1BgGXUTUdg/unMVHb30aGUVbtu7AfBpfuOtF3HjfLnzuilNx4bGjbSghERERUeP4TBcRtdWDO2dx3Q+3IqNoMAAUd6dqLssoGq774VY8uJPPyREREVF3YdBFRG0Tzyj46K1PwzCWB1vFDACGAXz01qcRzyitKB4RERGRI3h7IRE1ZD7ZfI95P33igNXCVQuzxevHjx7Am85a1XT+3Ww47Gt3EYiIiKhGDLqIqCGv+pcH2pKvAeDG3+7Cjb/dVXXbXvbEJy9udxGIiIioRry9kIiIiIiIyEUMuoiIiIiIiFzEoIuIiIiIiMhFfKaLiBpyz4df2tT+hmHg6v94HIcWMzV3pAEAAoCVgwHc/JdnQxCEpspARERE1AoMuoioIU70nnfV5rX4wl0v1r3f289fi5GIv+n8iYiIiFqBtxcS1cAwDOixWahHpqDHZmEY9bTNUDlv2LQCAa+EWturRAEIeCW8/vQVrpaLiIiIyEls6SKqQE8uIP27HyD565ugHd5tLZcmj0H4Ndcg+PKrIIaH2lfALhcNePG5K07FdT/cClR5QbIZmH3+ilMRDXhbUTwiIiIiR7Cli6iM7B/vxpH3n4zYdz8B7ciegnXakT2IffcTOPL+k5H9493tKWCPuPDYUXz5yk0IeCUAWNbqJeSHgFfCV67chAuOHW11EYmIiIiawqCLqITsH+/G3A1XwJDTgGHkBrv8MkNOY+6GKxh4NenCY0fxi78+H39+5iqMhaWCdauHg/jbS4/HnR+8iAEXERERdSXeXkhURE8uYP4L78wHVnrljQ0dgIj5L7wTE//2HG81bIIkCPiTE8bw2g0iDhsDUFQdf3LCGIZCXvZSSERERF2NQRd1FC12tN1FQOrub8PIplD5CSMbQ4eRTSH5m28i9Or/UbguMQctFi65mzQw1lQ5e01S1gAAgiAg7PMgOuhxpIdEIiIionZj0EUd5ch7N7a7CA0ykLjln5C45Z8KlvoAHCmzx8ofx1wvVTdJZdWC+bBPKrMlERERUXfhM11E1BHMli5TiEEXERER9QgGXUTUEVJFQVfYz4Z4IiIi6g0Muoio7TKKBlUvfIaOLV1ERETUK/hTMnWUiW/uamv+hmFg9u9fCf3ovuXdxFciCBDH1mL0/95X0NPe/n37sGbtWhdK2luKW7k8omC9t4uIiIio2zHooo7SCT36RV73AcS++4kG9rsWnsHxooXJjjimTpcs6kSDrVxERETUS3h7IVGR4MuvguALAUKNHw9BhOALIfjyK90tWA9b9jwXgy4iIiLqIQy6iIqI4SEMf+Q/AUGoHngJIiAIGP7b7/HFyE1Y1nMhO9EgIiKiHsJvNkQl+M94NUY+fivmv/BOGHIqt7DgGa/cc1uCL4jhv/0e/Jte1fpCdqDP/WY7frhlPz71xpPxxk0ry263mFbwys/fXyGlfQCAoaAX9/7ty5ouVyKr4q3/9gimY1k88cmLm06PiIiIqB5s6SIqw3/GqzHxb89h4F03QJrYULBOHJpA+I3XYeKm5xlw5f32hRn86NH9NW37/KG4y6VZohsGPv1fz2M6lm1ZnkRERER2bOkiqkAMDyH82r9C6DXvh7L7j1B2PwXBF4QQiEAMRiCGBttdxI7wu+1H8bGfPg29xg4fn5vOBV2TA378+zvPwrOHYta6qLaATScdCwAQIJTcv1aKpuPTv3oe//3ckabSISIiImoGgy6iGgiCAM/kMdDnpq1lRiYJQ1UgeLxtLFl76YaBm363G996YE/NARcAPJdv6XrJqgHohgG/rXv4sCQh5Gv+T9PhWAYf/+kz2Lp/sem0iIiIiJrBoIuoRkJwABAF2KMLPbXYt13CP7hzFl+6ewd2HEkCAE5eGbWCqWqenzaDruiyTjQksbnWrYyi4XsP78V3HtyLlKzBIwrYOB7G9sOJptIlIiIiahSf6SKqkSCKy24nNBLzbSpN+/31D7Zix5EkPKKAa/7kGNzwppfUtF8iq2L/XBpArqUrJRe+o8vTZNB117NH8LXf7kZK1rBmOICvv+MMvOLE/gyMiYiIqDOwpYuoDkJ4CEgsWPN6HwddAoBXnjSOa1+5EceMhXFwIV3Tfs8fisPI7x/wiPjG/Xvw/HQci2kVAa+IE0Z9eLcxgJce33igFA148M7z1+Ht569F0Cvh0T39e56IiIio/Rh0EdVBDA/BfjOcnlqEYRgQhOZaZ7rRbR84H+tHQ3XvZ95aKAjAe25+Eprtds1kVsOTB9N48pan8IZNK/DJ158Ej1hfg/wFG0dw5/UXIcgXLBMREVGHYNBFVAcxMly4QFVhZJMQApH2FKiNGgm4gKVONHQDWDXoxytPHMexE2F4RAH759O4a9tB7JmX8V9bpxHxe/DRS0+oK/3xqL+hchERERG5hc90EdVB8AUg+Aq/1Bu22w2puoyiI+iVcOJkBJ97y6k4Z8MwhkM+RANeXHDsKG54zWqctnoAAHDLlv14kR1gEBERUZdj0EVUJyE8VDDfz891NeILbz0Nf/j4y3Hze85BcS/zYZ8Ev0fEx16Ta90yAPxy66GWl5GIiIjISQy6iOokFgddKb4HqhFeSUQqW9hdfCj/HNYpKwcwkb9N8OkDsWX7EhEREXUTPtNFfSez5b+Quue7UHY+AT25CHFwHL5TLkL4Ne+H7/hzq+4vRoYK5o1UHIamQtn5OFK/+Sbk5x+GtjANQfTAM7QCsbMvReh1H4BnfJ1LR9SddMNAWikMusJ+Dxby0ysG/TgSz2I+Jbe8bEREREROYtBFfcOQ01j4yvuQeeQXBcv1uYPIPHArMg/cisjb/jeib/67iukIocFlL0mOffcTSN15U2F+yEI8vBPJO76G1L03Y+hv/h2Bc1/n3AF1uVRWtVchgFxL10J+WtFyK4Ne9kJIRERE3Y1BF/WNxW9cbwVc3hM3I/rW/wXvhtOgLc4gdde3kLrzJiR+9P8BhoHoWz5WNh1BlCAGB6Anc7cVZp64E+m7vw0A8J18ISJv/jt4N2yCnpjDod//HN7/vhFGfA7zX/ofGPvne+DdcLr7B9uhth1YxMd+8jTmUwr++pUbsWZkqQdEv0eEV8rd8azpBvbOpgAA6xrsJZGIiIioU/CZLuoL8vZHkP79LQAA/6ZXYfQf74D/tFdAjI7Cu+YkDP7l5xB9+z8BABI/+zzUQzsqpmd2pmGoCjIP/BgA4DvlpRj5h1/Bf/rFEAdG4Vl1PPQL3oqxz/wut72SRfyWT7t2jN1g1WAQh2NZZFUdD+2aK1gXsr1X63fbjyIp5249vOi40ZaWkYiIiMhpDLqoL5gBFyQvBq/5CgSPd9k24TdcB2liA6Bkkbz9xorpmc91qVNPw8gkAQCRt/49BGl547FnYj1Cr3oXACD71H0wVKXxA+lyoxEfNm8cAQA8tHMOO44sdQcfzgdd82kVX7jrRQDA5IAfl75kovUFJSIiInIQgy7qC8quPwIAvMedDWlsTcltBFGE77RXAACyT95dMT2zB0M9fhTw5nrZ86w5uez2nhUbcxOqDD0+W3vBe9AHX3Us/B4RBoBv3r8Hv31hBjPxLDKqjl9vm8bH7jiAQ4sZeEQB//D6k+D3FD7Tdcuj+/Gmrz2MN33tYfZsSERERF2Bz3RRXzDfpSWNr624nTgwBgDQZqagp2IQQwMltxP8IQheH/xnXAL/GZfAyKYAJVM2XXV611Ie4cF6i99TTlgRxf97y6n4+E+fRkbR8Ys/HsIv/lj4Lq6QT8I/vuFkXHDs8lsLF1Iy9uSf98oU9X5IRERE1IkYdFFfEIJRAICRTlTczkguvehYmztUNugCcs91GQtHctP+EPTkAqTR1cvTzKas2xu9G8+E4AvWXf5ec9a6IXzsshPx++0zeO5QHHNJGaIoYMVAAKdNePH+S16ClYOBdheTiIiIyBEMuqgveNeeAnX3VsgvPAxDTpcNfORnHrCmjXTlW9fEyBD0fNAFAHpioeR2sf/8X9AXDgMAQpf+zzpL3j1WDQXxxCcvrmnbVFbDUMiLN56xCm88Awh4ROtZr6mpqYoB1/tfvhHvf/nGusrWyD5ERERETuEzXdQXAhf+OQDASMwj9oNPldwmdfd3oB54YWlBlQ4vzOe6TEY6BkMvvN1NvP8/kbrrWwBy3ckHX35VfQXvUUlZLZgP+fkuLiIiIupdDLqoLwTOuszqJCN1x9cx/y/vgrzjMeiJeagHtiP2/X/E4r9/EOLIqqWdSvRwaCeEhwDBtsAAjOSCNZv41Vfh+eVnAQDiyCoMXf9tCCI/cgCQkguD07CPje5ERETUu/hNh/rG8Ae/jbn/9zYo27cg8/DPkHn4ZwXrvcefi/AbrsPCv7wTACAEIhXTE0QJQnAARmrpNkQ9uQAhMoL4Dz6F5C++CAAQh1dg9JO/hDSy0uEj6l7JbFFLl48tXURERNS7+LM79Q0xOorRT/0aA3/5eXiPOxuCPwwhOADv8edi4L1fxOin7wJstwdKw5PV08y/r8ukzR/GwhffZQVcxshqjP6fX8Oz+gRHj6WbabqBtKIXLAv7+fsPERER9S5+06G+Ini8CF/2PoQve1/J9crU0wByXceLkZGq6YnhIWjYCwDQUzEkvv9JaAe2A8j1VJh8x7/As+JYh0rfG4pvLQTY0kVERES9jS1d1DcMXYOemKu4jfzUPQAA73Hn1JSmYL0keQ7x7y0FXP6zLsPIp+4AomONF7hHpYo60Qh6RUiiUGZrIiIiou7HoIv6Quru72D6ylEcvuZE6Ol4yW2UPdug7HwSABA47w01pSsGwjDkDOI/+jT0hencvi97G4b/7ocQA2FnCt9jktmiTjR4ayERERH1OAZd1Be8J24GDB1Qskjd891l6w05jcV/vx4AIA6vRPClV9ScdvI334A+dxAA4D/7NYj82YcgiLxdrpxl3cXz1kIiIiLqcfyJmfqCd+3J8J/5p8g+eRfiP/w/gKogcP4bIQSiUHY8jviP/xnqnqcAQcTg+74EwVf4ct4jHzwbAOA77mwM/fU3rOWZx38N5bk/AACk1Scg+NK3Qp87BH1iXW6DbAp6JlGQluAPQxD693a6VHFLF7uLJyIioh7HbzvUNwY/8HXM/Z/XQd3/POI/+EfEf/CPhRt4Axh835cQOPs1y/bVDr6YGw8V9miYvOPrS9sc2I6Ff313wXofgMNFaY1/dRs8E+sbPo5upuo6Mmphz4V8MTIRERH1OgZd1DekwXGMfea3SN7xdaQf+hnUgy8CugZpfC38m16F8OuuhWfymLrSVLY/6lJpe1NxK5cA3l5IREREvY9BF/UVwR9C5M8/gsiff6Su/Vb+OFZy+Yr/PAQAyD5zP4z00m2EnrUnwTN5DKamprB+fX+2apWSLOouPuiTIPbxrZZERETUH9iRBpEDxHzX8SY9Md+egnS4VLawE40wW7mIiIioDzDoInKAGBkqmDeSC20pR6crbunirYVERETUDxh0ETlAKGrpMuQsDDnTnsJ0sJTMd3QRERFR/2HQReQAIRABPIUBhM7WrgKKpiNb3HMhW7qIiIioDzDoInKAIAgQQ4MFyxh0FSpu5RKFXEcaRERERL2OQReRQ8TIcMG8kVhoT0E6VLKoE42glz0XEhERUX9g0EXkkOLnuvTUAmAYbSlLJypu6eJLkYmIiKhfMOgickhxt/HQDRi6WnLbflTcc2HYx040iIiIqD8w6CJyiODxQgiECxeqSnsK04GK39HFTjSIiIioXzDoInLQsvd1aQy6gFzPhbJWeKtlhN3FExERUZ9g0EXkoOLnutjSlZPMLu+5MODlnx8iIiLqD/zWQ+Sg4ue6DF2HoWTbU5gOkpKX31oosOdCIiIi6hO8v4eoBpqmQZZlyLKMbDYLRVFKDrIsI733IBRFhabrWMQhbH1hB3TJB1VVoWnasrGu61UHwzBg5HtCLDcGcu8LqzaIoghJkqyxfdq+zOPxQJIkeL1ea95c5vF44PV6Kw4+n88axzPFQRf/9BAREVH/4Dcf6nmKoiCdTiOdTiOTySCbzVrT5mAuy2azVmBln1bV/u6F8IknngAAnHXWWQ3tLwgCJI8XktcHyeNFKODHg6EA/H4/fD4f/H6/NaRSKRw9ehR+vx+BQKBgXSAQgM/nYysZERERdRUGXdQ1NE1DOp1GKpVCKpVCMpm0ps2gyhzsy5wMmARRhMfjzQUQVstPblqyT0u5eQDY8UwuYDnlrIvg9fkgShJEUVoaiyIEUcy3RIkQxNxYFMXcvCAAggABAPLBRsG0LQAxDAMwDBj2acMAYFitZYauQzf03Fi3jQ1zPtf6pmkqdE2Drmm46577YBgGNp58BnRNy69ToWkqNDU/aEqu9U41lyvQNc0ql6rIUBUZAJCKAUcbPAeiKFoBWLkhGAxag30+EAjA6/U2mDMRERFRYxh0UduYQVQikUAymbSG4nl7YNUoQRDg9fnh9fnh8fnh9fqsea/PB483v86buyXO4/Xlh1yAZU1L9X1kVEWxgq5jTjodni79wj8wNAoAOO28P6lrP0PXoWoqNEXJBV2qAlVRoKkKVFWGqij5QbYGxZyWbdOKDEWWrcDQDKgb4fF4CoKy4iEUCpWc9/v9bGEjIiKihjDoIkfpum4FTtWGdDpd8DxSrbz+AHz+APz+IHyBAHz+IHz+ALx+P3y+gLXe6/Pnlwfg8Xj5hbkNBFGEV/TB6/UBCFfdvhLDMKBpKhQ5mwvI5GxuUGzTchaqnIUsZ6FkM9YyOT+GYUBVVcTjccTj8bryF0VxWSBmH5ebZssaERERMeiimqiqirm5OevLaiKRKBib08lksu5AyhcIwh8Iwh8IwR8IwmeNc8t9+eDKHwjC6wtAFNnpZj8SBAEejxcejxcI1b+/eYujImchmwFZNgPZnM9mIMsZKFlzfQZyNjfo+Q5PzNbXeni93pLBmH1Z8XwgwOuciIiolzDo6nOyLFtBU6Uhk8nUnqggwO8Pwh8M5YZACAFz2j4fCMHnD0Do4S+XgihgxdqN1jS1j/0W01BkoK59NVWFnE1DzmYhZ9NW4CZn01aQllufWQrgshkYhmH1bBmLxeoqq/ksWrlgzT42P6O8BZKIiKgzMejqcUePHsX09DTi8ThisVhTwZQkeeAPhhAIhXPjYBiB4NK0tc4f7OlAqh6S5MHmi1/f7mJQkySPB0FPFMFwtOZ9zJY1MxAzgzIlH7gtLbcHammoigLDMKzn1ubm5mrO034LZPFzaYFAYNnzavYORiRJaqRqiIiIqAYMunrY7Ows/vXLXwaq3O4nSR4EQrkAKhdQhQvmzWmPl111E9XK3rIWjg7WvJ+ua0staiUDM/M2yIxtu2yuR8kGb4EEAJ/PZwVgZpBW3AOkvXdIe2+RXi+fmSQiIqqEQVcPW1xcBAwDkseLydXrEQhFCoOp/JjBFFHnEEUJgWDuFtx65G6BtD+LlrWmrefX8s+wmdOKnLW68Tdf/r24uFh3mSVJWtZtf7lu/e3vX7O/h83j8fDvEBER9ay6g66Pf/zj+NnPflY50XyXzBMTE9i0aRMuueQSXHzxxWW3NwwD99xzD37xi1/gqaeewtzcHLxeL1avXo2LLroIV199NVatWlVvUSkvFBnAua94bbuLQUQuyt0CGUEwHKlrP13Xl3p8tHUuYu/90T7I+R4izflcr5Jawy1sVvklqeBF2NUG86Xa9pdr+3w++Hw+trwREVHHcaWly94l886dO3HbbbfhvPPOw4033oiBgcIH2BcXF/GhD30If/jDHwqWy7KM7du3Y/v27fj+97+PT3/607j88svdKC6RazRNxdaH7gMAbLrglXW/54vIbaIowpfvLbRehmFAVRUo2SxUJQvF7MpfyS517Z/vMdIK1GzvXVOVLFRFAZB7b5/5Tr5mCYJQEJTZgzFzudfrRSaTwe7duwvW2cellkmSxICOiIjq1tQ3wOuvvx4nnHDCsuWapiGRSOC5557DT37yE6TTaWzZsgXXXnstbr75Zus/LF3Xce211+LRRx8FAKxatQpvfvObsXHjRiSTSTzwwAO46667IMsyPv7xj2NgYKBiixlRpzF0A/t2PgcAOH3zKwD2VUA9RBCE3IvGvT4AtXcyYmd2OGIGYoqSzc8XvjRbVRQo+Y5JDk3tAAAMDI9B01RrvaYqVprZbBbZbNapQ7XkjtlbdvB4PAXT5nypcfF0qUGSJHg8Hr5CgIioyzUVdJ199tnYvHlzxW3e9ra34aqrrkIsFsOWLVvw61//Gq99be5Wt5/97GdWwHX++efj61//OkKhpecYrrjiCtx77734m7/5G6iqik996lO46KKL4Pf7myk2ERF1CHuHI8Ea3p+tKgpuzwddL3vNFfDYXj5tGAY0VckFbGouaLPPl5rW1KV5TVXz02p+uWotNwzdysN8/q2VRFEsCMLs0/ZxpelaBlEUyy4TRRFzc3MIhUIF6+zrzWm2BhIRFXL9Xqfjjz8e11xzDT73uc8BAH75y19aQdePfvQjAIDf78fnP//5goDLdPHFF+Md73gHvvOd7+Dw4cP4/e9/j0suucTtYhMRUZcRBAEerw8er8/xtHVNg6apuYDMHKtKwTJVVXLb5ed1a52Wm87Pm2lVGxfkr+stD/SaIQjCsmCs0mAGauXWl1tnX15quni/4uXlxtXSq2d/QRCQSqWQTCYrbkNEva0lD5hceumlVtD1zDPPAAASiQSeeuopAMCFF16I8fHxsvtfdtll+M53vgMA2Lp1K4MuIiJqKVGSIEoSvL7W3GlhGAYMQ88HbOagQtNt83rxOLe9oeemrXUF8/rSett+uq7B0PX8et3axz6t6zoMQy/cztBLvpbE7GBF07SW1Fe3EwSh5gCuUiBaLsgsnq8W5BbvU2vwXC6grjRfbhlRr2lJ0DU5OWlNz8/PW+PTTjsNhw4dwsaNGyvuPzi49I6bWCzmTiGJXCB5PHjtlddY00TUHEEUsGLtRmu6V+W+hEsQRQnwVt++nQzDWArEjHwAV7CscL0ZrBm6Yc0bhrG0Pr99Li0DMHTohrmtsbRNQRpFyw2jYJlZDtjXF0/r9uWF61CQZoU0bOkU51W1DvPlpKUg1HyesVyAVnz7ay3T9ayvN41K6diXmcdH/aUl3wLj8bg1HY3mHrZeu3Ytbr311pr237lzpzU9NDTkaNmI3GQ+r0JEzpAkDzZf/Pp2F4NsBEGAkG8JpPIKA8ISQVu5gE4vDgCLAtRSaeql1+mGDui5cXHQarZamq2Y5v5LwXJheroVMOeCY6PMvH1/+z7muvJ1ZXTVLbX1qrXFr9ZBEARkMhlEIpGSLZmVbrW1b1/LuLjFtXhdpWX1bAOg6jJz3r681DadoCVB18MPP2xNn3LKKXXv/9Of/tSa3rRpkyNlIiIiImoVs/USAHuytcm18GlWUFfuNtalFtHctoUtqtpSi6auLQ/6bGmY2xi2AFMv2tZeDmtZfrvi+WWtuDUElQB4+20blArSTJ/61Kdcz9/1oCsej+PLX/6yNf+6172urv3vv/9+3Hdf7j1Ho6OjuOiiixwtXz/QNRWLczPtLgZ1MVnOdb3N64iIiNol1yrjyQetHX7fbV5BS+OyW2z1wuBP15dupy3TQmi/Zbe4VdOefqnWUt28VdZq7SzdMopl++angaLbZ5fWdQPzNt9qt/u6xZWgK5vN4vDhw3jkkUdw0003Yd++fQCAk046CW94wxtqTufQoUP42Mc+Zs3/1V/9FQKBgOPl7XXJ+CJ++18/bHcxyAVPPPEEAOCss85yNZ+Zg3sBgNcRERERUQOaCrquvvrqmrddt24dvvKVr8Drre2Xibm5OfzlX/4lZmdnAQAXXXQR3vGOdzRUzn61evVqrF271uq8pJRMJgNVVcuup862e/duAO4HXcccc4yr6RMR9bJAIAAPO1Oqi6ZpkPicYF1YZ41p1W2erv8FWLt2LS6//HJcffXVGBgYqGmfmZkZvPvd78auXbsAAOvXr8fnP//5jnoYrhv4/X68733vq7jNbbfdhieffBLHrxrEMStqOz+9SNV03Lv1AADg4k2r4ZG6o7vaP9ybe7fdn5611tV83E7fSd16LomotxiGgf9+cj8A4C1veQtOPPHENpeou0xNTWH9+vXtLkZXYZ01ZmpqqiX5NBV0XX/99TjhhBOWLff7/YhGo1i3bh1GRkbqSnPv3r1473vfa1XAypUr8e1vf7vudKg+q0bDOGPjWLuL0Tayqllf1E8/ZhQ+T3f8UjQUzr0Etp/PXbFuPZdE1Ft0W9BFRNRU0HX22Wdj8+bNTpUFTz31FN7//vdbtxSuXr0a3/3ud7F69WrH8iAiIiKyMwwDmm5A1XSo5ljVoWj56fygaDpUzciNVR2ytV1+UHUoqmZNExGZOuYG4/vvvx/XXXcdUqkUAOC4447Df/zHfxS8WJnILaIg4LhVg9Y0dS+eS3KSYRgwAOi6kX9B79K0bsDWe1euZcMAAGNpP8O+DMBSp1mN9p4lwLyshaVF+emidQIgQCjaxtZdctG+uXXFuRUqVWojf4CGbYNlx26rj1y95VaYdZjr4jtfh/ltcnWcW6frBjTDgK7DWqbpOjRz3bJBX5rOB1Kalluumss0A6quQ8sHUW4Jh8Pw+XyupU9E3aEjgq57770X1113HRRFAZB7F9dNN92E4eHhNpeM+oVHEvGmCze2uxjkAJ7L3qPruS/FstmKoC61KpitD2bLg9UKoelLLRe21gut6Mu3XvRF3ZzXbV/wqb8IggCPxwOv12uNyw0+n6/stDk/OTmJmRm+boOo37U96NqyZQuuv/56K+C66KKL8NWvfhWhUKjNJSMiomZouoGsoiGraMjIGrKKiqyiW8tk1Rznlsn5aVnNj5VckKXqnRn4iKIIQRCssTkNwJo3B3NZqXGtit8x0+i42rJy8+Yye7lLTZvjcnVhry9z3l6XpQZJkpZNezweSJJkzUuSZC2zr7MvKzWYwZU5LUkSO+4iIse1NeianZ3F9ddfD1mWAQCvfOUr8eUvf5nN8EREHUTTDaRlFelsfpA1pGUVGTk3nZFVZORcYJVR8tOK5vgzLYIgFLQgFLcu2IdSX6rtX7wlScLc3BxWrlwJj8djfWm3D8Vf9u2DPZjqJ+wdjYioMW0Nuj796U9jbm4OAHD66afjX//1XxlwERG5TDcMpLMqUlkVyYyCZFZFKpObzw2KtT4t51qjmuH3++H3+xEIBJYNPp/PWmdO+/1+a9oMsnw+Hzwej6OBDgMIIiJqlbYFXTt37sSdd95pzV922WW4//77q+43NDSEc845x82iUR9SNR13PbEPQO6dVHy3U/fq53Op6QaSGQWJtIKEbZzMKEhmVGs6lVVR76NKgiAgGAwiFAohFAotmy41BAIB+P1+vqyTiIj6XtuCrp///OcF94t/9rOfrWm/8847D//5n//pVrGoT+mGgaencq2urz5zTZtLQ83o1XOpajoSaQWxtIx4SkEsJSOezgVW8YyMRCrXYlWPUCiEcDiMSCRSMDYHc70ZWJnP6BAREVF92hZ0bd++vV1ZExF1HEXVsZjKYjEpI5aSsZiUsZjKTcdSMpKZ2gIqURQRjUYxMDCAaDSKaDSKSCRSMG0GV2yBIiIiao26g64bbrgBN9xwQ9MZ33TTTU2nQUTULQzDQCKjYCEhYyGZxUIii4WkjMVkLtCqpZXK4/FgcHDQGgYGBjAwMIBMJoPjjjsOAwMDCIVCbJEiIiLqMG3vMp6IqFcYhoFUVsVcPIu5RAbz8SzmE1ksJHNjVav8IFUgEMDQ0BCGhoYwPDyMoaEhDA4OWstCoVDJjiSmpqawatUqtw6LiIiImsSgiwiAVxJx3RtPs6ape7XiXGq6jvlEFnPxLGZjGczFM5jNB1iVevoTRRGDg4MYGRmxhuHhYWsIBoOulJeIiIjai0EXAQDSWXXZSy/7iSAICPj4cegFTp5LVdMxF8/iaCyN2VgGs/EMjsYymE9ky/b+JwgChoaGMDo6umwYGhric1RERER9iN8y+5zHk7sEtmw/gm175rB+IoINkwPYMBnFQIjvTKP+YBgGFpMyZhbTmFnMYCaWzgVX8Qz0MsGVz+fD+Pg4xsbGrPHY2BhGRkbg9XpbewBERETU0Rh09bkLLrgA8Xgcu3fvRjqbxfP7F/D8/gUAwGg0gA2TURwzGcXa8Si8Ht52R91PVjXMLGZwZCGFI4tpzCzkgixF1UtuHwgEMDExgYmJCYyPj1vjgYGBvm0ZJiIiovow6Opz4+PjePvb3w5N07B//37s2LEDO3bswIEDBzAbz91O9fiOGUiigDVjYWyYHMAxkwMYHwz01BdO3TCwfyYBAFgzHoHYQ8fWb+zncjjqtwKswwtpHFlIYz6RLbmfJEkYHx/H5OQkVqxYgcnJSUxMTDC4IiIioqYx6CIAuS+c69evx/r16/GqV70KqVQKu3btsoKwxcVFTB1JYOpIAr/bdhDhgMe6DXHDRBThQHffTqVqOm75/Q4AwAcvPx0+D5+76SaGYSCRVjC9kMLB2SQeeeFIxe2j0ShWrFhRMIyOjvJ5KyIiInIFgy4qKRQK4dRTT8Wpp54KwzDw1FNPIZVKYceOHdi9ezeSGQXPTM3hmak5AMDEUBDH5IOw1aNheNgDILnEHmBNz6UwPZ9rxUqVeM+VIAgYHR3FypUrrWHFihWIRCJtKDkRERH1KwZdVJXZG9umTZtwwQUXQFVV7N27Fy+++CJ27tyJQ4cO4Uj+1q1HXjgMryRi7XgE6yei2DAZxdhAb92KSK2VzCiYnk/hkBlgzadKvkhYFEWMj49j1apVWLVqlRVg+f3+NpSaiIiIaAmDLqqbx+PBxo0bsXHjRgBAIpHAzp07rVsRE4kEdk3HsGs6BgC5WxEnolg/GcX6iSiiQfaKSKVlZLUgwJqeTyGeVpZtZwZYq1evtoKsFStWsNdAIiIi6kgMuqhpkUgEmzZtwqZNm2AYBg4fPmwFYXv27EEyo+KZvfN4Zu88gFyviOsnIlg3EcXa8QiCHfB+LEkUcfGm1dY0uU9WNRyeT+eCrPkkpudSWEjKy7YTBAFjY2NYvXq1FWStXLmSARYRERF1jfZ/26WeIgiC1THBRRddBEVRsHfvXuzatQs7d+7EwYMHrV4Rn9h5FEDuebB14xGsG49izVi4LS8plkQB5xw/0fJ8+4Wq6TiykLZar6bnU5iNZVDqFVgDAwNYt24d1qxZY7Vi8RZBIiIi6mYMushVXq8Xxx57LI499lhccsklSKfT2LVrF3bv3o1du3ZhZmbGeh7ssRdnAOSCsLVjEawdj2DNWBghP1s0uomq6ZhZNAOsNA7Pp3A0li75kuGBgQGrBctsxZqZmcH69etbX3AiIiIilzDoopYKBoN4yUtegpe85CUAYL2YedeuXdizZw9mZ2etIOzxHbkgbDjix5qxMFaPRrB6LIyRiJ8dc3QIWdWs83V4IYXD8+myAVY4HLYCK7MVKxqNtr7QRERERC3GoIvaKhqN4vTTT8fpp58OIBeE7dmzxxqOHDmC+UQW84kstu3JdU8f8ElYORzCqtEwVo2EsXIk1JZbEvuJYRhIZFTMLKRwZDFd9UXDgWAIa9csdXKxatUqDA4OMlgmIiKivsRvqtRRotEoTjvtNJx22mkAgFQqhX379mHv3r2YmprCgQMHkJFV7D4cx+7DcWu/obAPk8MhrBgKYXI4iMmhEIL+2i9vRdXxrbueAwC8509PhtfTv51pZBUNs7EMZhbTmDHHi2lkZK3k9tFoFMGhMYSHxxEdmcDAyDjOPHYVJgYCLS45ERERUWdi0EUdLRQK4cQTT8SJJ54IAFBVFYcPH8b+/fuxb98+7Nu3D3Nzc1hIylhIynhh/4K1byToxcRgEOODQYwPBjA2EMRI1F/yxc0GDMRSsjXd68yWq7l4BnPxLObiGczGMjgazyBRoot2YKkXQfP9V9Z7sIIhPLhzrmDbcB0BLxEREVGv4zcj6ioej8fqdGHz5s0AgHQ6jYMHD+LAgQM4ePAgDh48iPn5eSTSChJpxXpfGAAIAAbDPowOBDAaDWA44sdQxI9IoPc669B0A4l0LhhdSGTzgWkWC4ks5hJZKKpedt+BgQFMTExgcnLSGsbHx0t2075YFKQJAII+yenDISIiIupaDLqo6wWDQauHRFMmk8GRI0cwPT2N6elpHD58GEeOHEEmk7FaxXYeihWkI4kCBkI+/OzB3RgIeREN+jAQ8iEa9CIU8CAc8CLk80AU2/tckmEYkFUdyYyCVFaFqhl4dPsRJDMK4mkFsZSMWEpGIq1UbLMTRRFDQ0MYGxvD2NgYJiYmMD4+jomJCQQCtd8amMyqBfMhnwSRz24RERERWRh0UU8KBAJYt24d1q1bZy0zDAOJRAIzMzOYmZnB0aNHMTc3h9nZWSwsLEDTNKvTjkpCfg8CPgkBb37skxDwSvB6RHil/NgjwiMJEAUBoihAEgQIggBBAAwDuWDIyN3IqOkGVE3PD7lpWdUhqxqyigZZ0ZFRNGRkFemsirSsWr0DSpFxSADue+pAybJKkoShoSGMjIxYw/DwMMbGxjA8PAyPp/k/AamiZ71CfrZyEREREdkx6KK+IQgCotEootEoNm7cWLBO13UsLi5ifn4ei4uLBUM8HkcikUAymYRhGEhlVaSyKoDKwZnbfD4fLrnkEuuYotEoBgYGMDg4iKGhIQwODiIcDkMU3e0UJFkUdIXZkyQRERFRAX47IkLuVrvh4WEMDw+X3UbXdaRSKSQSCaTT6YIhk8lAlmUoimKNFUVBKpWCz+eDpmnQdR2GYeRbvAQrGBIycXgEAx5RhEcS4R0YQXBwFH6/H4FAAH6/H36/H6FQCKFQCOFwGMFgsOTzVe2QKnF7IREREREtYdBFVCNRFBGJRBCJRGreZ2pqCuvXr6+4jbLnKWhHl24PlEZWwrvxjEaL2VKKpkPWCp8cC/P2QiIiIqIC/fsyIqIOIYSHCub15EJbytGIZLbw1kJRAIJeBl1EREREdgy6iNpMLAq6jGwahtLe58VqlZKX31oosOdCIiIiogIMuojaTAhGgaLOLrqltau4pSvETjSIiIiIlmHQRdRmgiBAjAwVLNMTC20pS72SRS1dYXaiQURERLQMgy6iDiCGC3tNNLq1pYudaBAREREtw6CLqAOU6kzDMIzSG3cIWdWh6kU9F/L2QiIiIqJlGHQRdYDi2wuh6zDSsbaUpVbJovdzSQIQ8PJPChEREVEx/izd4x7bMw8AOGdD+Zf+urFvJd9/eC8A4O3nr6uYR6V19jQqKZeGE8dmT6NUeo/tmYeiBlHuLV3mPgDwwnQcx4+diTO0vdDjswByz3U9cUSvmEc9ZSy17oXpOE5cEV22/vsP78W+uRRefcrksv3Msqi6jqGQFwspBQAQzPdcWFwvL0zHAaBkPp2sWn279floFac/A06mW0t+Tn8eOp1b54tax6n6b/T/zVLb8Jrobjx/3YVBV4+76fe7ATT2gWxm31rSffv56yrmUcu6akFXuTScODZ7GqXSM5ddcNxExf0BWMHPTW8+AfJzDwHIPdd10/2zFfOop4yl1pULum76/W6kZA07j6aW7Wcvyxfeehqe3LsIAIj4PcvyNPMAui/oqlbfbn0+WsXpz4CT6daSn9Ofh07n1vmi1nGq/hv9f7PUNrwmuhvPX3dh0NXDHtszj8enFqzpen8Rtu+7NhhwpEzff3gvEvnOF2644/my5atUdnsa3394b9nAq1wazdRLqbS///DeZelVy8O+3vT41AKemDFw+sAY9NhRPHZYweNTsbJ51FPGSvX7+NRCwXp7/RaXsbgs26cTGA55MZ9SEPJJZeulVD6drJ7z1y3HZOf0Z6DW674Z1T5zjZS3W7h1vqh1ts9pjtR/rX/Xy+XR7OeIOgc/092HD2D0MHtLin26kX2HhoYcL9NPnjhYcnmp/GtdV8t2zdRLrelVy6Ncvjf9bjc8q48HAHzrOb2pMtdah/WmX7Dt73Zjw1gIABD2exo+7k5Tz3F0yzHZufUZcLNemv0Md/M5c/NvFrXGLVvnrOlm6r/Z/xud/r+Q2ofnr/sw6OpRxS0pZitDo/tuO7T8NrN62VtQAMDe8Z29fJXKXpxGIqtZz3dVO4bH9sw3VS/l0raX5/GphbItPOX2t3t8b66164/GajyxP1E2j2plrnSc5VrZHtszv6x+Sykoy94Fq7Xrhel42Xqpp+ztVu0aceIaaic3PgO1XPfNqPaZa/TzEAg404LvJrfOVzdds93usT3zeHLfUsdIjdZ/PX/XS+XR7OeIOkc3/03rZwy6elQzrQwl9/1d87+i1NqCUqnstR5XuWVOtL7U0xJUalnV/X+3G996VnGtDJV+AW3k17Kbfrcbx4yF8R9/mKpt+w7/Ra6R89fpx2Tn1mfAzXpp5Hqvtt7JFnw3tfJ8kTucqv9m/29s9nNEnaOb/6b1MwZdPahcS0ojvwhb++5t7lewWlpQSv1abl93wx3Pl0yjuLWr0vE3Wi/V0i4uT7k8atn/8b0LeDzfOUXZbSqUudLxl6tfc321c1Ryv70L+OUfD1Y9Lns+nfqLarXPTjOfrU7gRPnLpeFWy2ZNn5kGPw9OtOC7yc3z1S3XbLdzqv4b+bte7Q6HZstE7dHNf9P6HYOuHlTrc05O79tourVuZ38GrNJ+Tv2C2Mx25fZ18pfEWu7Zr2edG2VpdTmaVa3u2lG3TnKi/K0+182Wq+IxO9CC7ya3z1c3XLPdzqn6bzSdanc4NFMmao9u/pvW7xh09Zhqv2Y18otwLftWUksrl6nSdvZnwErt9/2H99b0a14pzbQC1pNHM/uXSq/aPfvFGmnJqkW96XbiL6q1XP9ufD5axYnPdyOfgWbqpZ78Gvk8NNuC76ZWnK9Ov2a7nVP138zf9Wp3ODRaJmqPbv6bRgy6ek4tv1I18otwPds4sU8jmm1J6sb73eu9Z7+TdFpZnShPpx2TnROfbyd6znNzv0Y+D516zlp1vjr1+HuBU/XfqtZip/Ij9/Az3d0YdPWQWn8VbuQX4Ur7VlJPK1ezElmt6ZaoRlsB26Xee/Y7SSf9oupU3XXSMdk58flupo6c6HGv3nzc+pvWCq08X514/L3Aqfp34m9TL9yJQPxM9wIGXT2kmfvDnbq3vJltO0EzrYDtUu89+52kU8rcimft2smJz2y3/NreyOeh085Zq89Xpx1/L2jF81pu4jXRefiZ7n4MunpEvb+GNdNCUuuvKK1s5XJKM62A7VLvPfudpBN+kXP6/HbCMdk58fl2oo6c7iW0Uj71fh466Zy143x10vH3Aqfqv53/9/Ca6Cz8TPcGBl09oplnrdx6Tqtbf2npxuekuqGM5bS77G7k3+5jsnPi8+3U8bSqB7VOfva0mnadr045/l7gVP23+5y0O39aws90bxAMwzDaXYhW2rZtG1RVxZlnntnuohBRCVNTU1i/fn27i9FVWGfUKrzWqFV4rVGrtOpa87ieA3W07DP3AwD8L3lZxeUzD92OgYGBZds5kVel9fZlxeuL1yl7noJ3w+lVy1huW/vy9MM/BwAMvecLTR3XzMwMxsfHq25XLPXEXdAOvADvhtPgWXUCDE2Gd2JDQTnV6V0AgOD5l5dMM3H7jcg++wD8p7wUkdddW3OZKx2j75SXYj6ZhaobkLXc7zWjYS+CPm/daZYT3PJjJJ4OLSuzvSy1nmt72ZstV7N5uFWOxO03YkQ3gPV/7Wi6pbSiLltZjkOHDmHlypUtyavT1XoNjykK0MCXk26tx2bLbe6fWHEKRkdHHSsXUSm1/E2j9mHQ1ecSt34GwPL/UJYt//VXkBDFpv7DLJdXpfX2ZcXri9fV+kW83Lb25fLzDwGoLeiqdFypVKqm7Zbt94t/gTr1NLwbTsfwR38A+cVn4BlfD0EQrHIamSQAQNv3XOmg69bPwEjFoDxz/7IApp6y2Lcf/ofbse1gomDdRHSkoTTL0W//VySAskFXPefayXI1m4db5TDTjb7B/aCrFXXZynLIstyyvDpdPddw+IyLXUm/EzVbbnP/yCd+5liZiMqp5W8atQ+f6epj2Wfuh/zsA5CffcD6Na7U8uwz90N9/sFl2zmRV6X19mWJ229cVqbidUYqVrWM5n7F2xYvh64BuoaFb32kqeOqdztzW+W5B62yKHu2QQwNwcgmC8pplrFUmonbb8xtA8BIxZC4/caGylK8vfLcAxgOea11XkmAzyPWnWY5ZrmLy1xcllrOdXHZmylXs3m4VY5q9eWkVtRlp5WjU47ZbW5fw91aj82W274/dmxxoYRE1E0YdPUx8xe4atPl1jmRV6X1jZapUhnLbVtun/R//0fZtCqVsdHtSq2P3/oZeFYfDyOxUHbfanVQy7mupTzxWz+DDWMhaz7s8zSUZi15VbtOasnLqXI1m4db5WjF8bUjr04pR6ccs9vcvoa7tR6bLXfx304i6m8MuvqU+QucqbgFyb681HZO5FVpvdl6ZTJbbUqVyb6uUhmL8ymXV4EKrV3Vjqve7cqVUcm3dsk7Hi9bTnua9lYuk9kSUk9ZSpVHefYB+Hc+bLV2hXxS3WmWU1zuci10dvXUpRu/steSh1vlqFZfTmpFXXZaOTrlmN3m9jXcrfXYbLlL/e3shuMmIvcw6OpT5VoR6m35aCavWtc3op40q+VVrrWr1jycKEv81s8g+et/q1TMqq125c5vPS2DZlnM1q6wX3Ls/DVyndS73Olf2RttjXOiHK04vnbk1Snl6JRjdpvb13C31mOz5e7W4yYi9zDo6kOlWgyA5S1I5TTzK2dxGuXWF7fU1Ktaa1pdeZVo7ap2XCb/vq01bVetjMqzD0B57g8Viyk/+wAWvvWRssdjPa9WQ1kqlcfe2hXe80hdaZZTqnXOLHOpFrpqedV6fppRSx5ulaNafTmpFXXZaeXolGN2m9vXcLfWY7Pl7tbjJiJ3MejqQ279yl7vdrW2rDXKyecIilu7am0dku76Wk3bVUuzVtWeQau1LOWWmczWLuXnn60rzUa2reU6aea5vkbVkodb5WjF8bUjr04pR6ccs9vcvoa7tR6bLXe3HjcRuYtBV5+p1GJQj2Z+5bSn4URZKqVfqTWtLrbWrlqOy8xXff7BqtvVkmY95WxEPS2DQK61S7znG1BqqIdqyrXamMq10JXLq9bz04xa8qj4rGAT5ailvpxq7WpFXXZaOTrlmN3m9jXcrfXYbLm79biJyH0MuvqMWz2nuZ1Xo5xsTTNbkWr9pbOeX0Q7pa5KTdeyvdvb1KKeunTjmUG3tmlVuu3Oq1PK0SnH7Da3r+FurcdWtGR14nETkfsYdPURx1pT8pr5lbNVHG1N0zXMfu7Kmp97q3W7ar8mt0q9LYO1PHdX7Vfdaq029ainLpv5tbmV9VOs1vpyorWr1uN0+5f7VpajU47ZbW5fw91aj82Wu1uPm4hag0FXH3Hj17VmfuXsRvKjtzueZifVlRvP2XXKszhO5d0tZW5Fa55TeXVKOTrlmN3m9jXcrfXYyt4JO+m4iag1GHT1Cbdanpr5lZNynGrpcYIbz9mV+1XXyVYuUz3pNfJrsxvXtlPPvhVrprWr3uN065f7VpajU47ZbW5fw91aj82Wu1uPm4hah0FXn2jlL9H8BY+Kdeo7a+otQyvfCeVE3q1szWtlS7ob5eiUY3ab29dwt9Zjs+Xu1uMmotYRDMMw2l2IVtq2bRtUVcWZZ57Z7qIQEVEbTU1NYf369e0uBhGVwM9n/VhnjWlVvbGli2oyMzPT7iJ0JSfqTZczUGJHocTnrcFosGv4Rh2YT+PxqXk8PjWPp/YvYO9c0rW8eK3Vj3VGrXLo0KF2F6Ersd6IyNPuAlB3SKVS7S5CV3Ki3gTJA+3FRwFbm7R00mYIkZGm067VQlpBIrsU6EX8Xtfy4rVWP9YZtYosy+0uQldivRERW7qIOpwgeSAEowXL9MR8y/I3DAOHFzOYS8pIZFUYhoGQT2pZ/kRERETdji1dRF1ADA9BS8WteT2x4Hqe8YyC/9o6jVu27Mf+hbS1fDTsw1Wb1+AtZ69GNOBeixcRERFRr2DQRdQFxMgwtJl91ryRXHA1vwd3zuKjtz6NjLL82bHZpIyv3rsL37x/Cp+74lRceOyoq2UhIiIi6na8vZCoCwjhoYJ5Q5FhyOnSGzfpwZ2zuO6HW5FRNBgoeJRsKX8AGUXDdT/cigd3zrpSDiIiIqJewaCLqAuIgTDgKWyYduMWw3hGwUdvfRqGUTrYsjMAGAbw0VufRjyjOF4WIiIiol7B2wuJuoQYHoK+eNSa15MLkEZWltx2PtlYT1k/feKA1cJVC7PF68ePHsCbzlrVUJ7FFjNaw+V323DY1+4iEBERURdi0EXUJYqDrkrPdb3qXx5oQYny5QBw42934cbf7nIw1T0OpuWcJz55cbuLQERERF2ItxcSdQkhMlwwr6cWYeh6m0pDRERERLVi0EXUJcTQYOEC3YCRWmxPYYiIiIioZgy6iLqE4PFCCEYKlukudx1PRERERM3jM11EXUQMD0FLJ6z5cs913fPhl9adtmEYuPo/HsehxUzNHWkAgABg5WAAN//l2RAEoe58i+3bvx9r16xpOh0iIiKiTsGgi6iLiJEhaEf3W/Pluo1vtJe9qzavxRfuerHu/d5+/lqMRPwN5VksFpDYSyARERH1FN5eSNRF7C9JNgwD2sIRKAe2Q4/NwjDqaZ8q7Q2bViDglVBre5UoAAGvhNefvqLpvImIiIh6FVu6iLqIEIhAVzKQt96L7BN3Ql84bK2TJo9B+DXXIPjyqyDagrN6RANefO6KU3HdD7cCVV6QbAZmn7/iVEQD3obyIyIiIuoHbOki6iLy1nuweOM1SN97M/SFIwXrtCN7EPvuJ3Dk/Scj+8e7G87jwmNH8eUrN1ktXsWtXuaygFfCV67chAuOHW04LyIiIqJ+wKCLqEtk/3g35m64AlCyyLVBFbVDGQZgGDDkNOZuuKLpwOvOD16Iv730eEwMFD6rNTHgx99eejzu/OBFDLiIiIiIasDbC4m6gJ5cwPwX3mkFVhUZOgAR8194Jyb+7bmmbjW88ry1OHf9ELYfSSKr6vB7RBwzGsLJqwYaSpOIiIioHzHoImohLXa0of1Sd38bRjaFyk9Z2Rg6jGwKyd98E6FX/4+G8jT50ikMqGlo4VyrlupAhx1ERERE/YRBF1ELHXnvxhbmZiBxyz8hccs/NZWKF8CxALb/f7sAAKrGoIuIiIioHnymi4jqomh6u4tARERE1FUYdBFRXRS2dBERERHVhUEXEdVF1Q1HXsRMRERE1C/4TBdRC018c1fd+xiGgdm/fyX0o/uq91xoJwgQx9Zi9P/eB0EofttW7WRVw+NTCwXLVN2AV2o8TSIiIqJ+wqCLqIWkgbGG9ou87gOIffcTDex3LTyD4w3lafIbBrSjUsEyRdPhldhQTkRERFQLfmsi6gLBl18FwRcChBo/soIIwRdC8OVXNp23KAjwiIWtWnyui4iIiKh2DLqIuoAYHsLwR/4TEITqgZcgAoKA4b/9XsMvRi62POhiD4ZEREREteLthURdwn/GqzHy8Vsx/4V35l+UDBS8LDn/3JbgC2L4b78H/6ZX4XO/2Y4fbtmPT73xZLxx08qK6c/Es/jew3tx/4uzOLSYgVcSsWY4iItPGseJKyIF29bb0hXPKPjBI/tx3wsz2DuXgigIWDUYwMtOGMNfnLsG41F/XekRERERdRMGXURdxH/GqzHxb88h9v1/RPaxO6AvHLbWSRMbEH7N+xF8xVUQQ4P47Qsz+NGj+2tK9/GpeXz4x9sQz6jWsqyq4/npOJ6fjmMo5MW7L1yPdaMhAPW1dO08ksBf/3ArDseyBct3zCSxYyaJnzx+AP98+Sl46fGNPe9GRERE1OkYdBF1GSE4gMBZl8F/5qUwMgkYchr+0y+GNLbG6qXwd9uP4mM/fRp6DQ1Sh2MZfOhH25DIqhgN+3DtKzfivGOGAQBb9szjq/fuxFxSwTcf2IOPXXYCwn4P1BpbupJZFdf/6CkcjmUR8Ut4358cg5cdP4agV8JT+xfx5Xt3Yv98Gh/76TP4wf88F+vzQR0RERFRL+EzXUTdRpUBAIIgQAxGIQ1OQBqehCAI0A0DX//tLnzkx0/VfAvgf/xhComsCr9HxI1XbcLlZ67CqqEgVg0FcfkZq/Dlv9gEQQDiGRUP7ZrLFaHGlq6fPH4ABxcyEAB85k2n4h3nr8P60RAmBvx49SkT+Ld3nIGwT0Ja0fC9h/c2VB1EREREnY5BF1GXMTRl+ULRgwd3zuIvvrEF/37/HugGcPLKaE3p3ff8DADgT04Ywwkrlu9zyqoBrBkKAgD2zuaeJVNqaUIDcG8+7ZesGsBFx40uW79qKIgz1w0BAJ4+EKspTSIiIqJuw9sLibpNcdAlSRBEEX/9g60Acj0NvuelG/Da0ybxZzc+XDW5X/71BdhzNIWQXyq/Ub7zQjHfi2Gtz3R9811nYWo2VdM7nT182TIRERH1KAZdRF3GUNWCeUHy5sYAXnnSOK595UYcMxbGwYV0TekFvBJOqtAq9vCuOeyby6V10mSuF8Nab130SiKOm4iUXb/jSAJbds8DAM7fOFJTmkRERETdhkEXUbcpbuny5IKu2z5wviMdUWi6gcW0gn3zadyxbRo/f/IgAOD4yQjOPSYXGNX6TFcxwzCwkFZwaDGDe56dwa2P74es6ThuIoyrL1jfdNmJiIiIOhGDLqIuY+Q70jCZLV1O9fz3+NQ83v+9PxYse/NZq3DeMSOQrNsL63tPl+nAQgZv/OpDBcsuOWUC/+u1JyIa4J8jIiIi6k3sSIOo22iFtxeaLV1OObSYWbbsjm3T+NVTh6DlO9AwUN+7uiql/dsXZvDZ32xHMquW2IOIiIio+/GnZaIuY6iFtxeaLV1Ouei4UdzzkZci7PNg50wS33pgD+59fgb3vziL+aSC97xsAwBA1Qx4K/S9UcrxExH85oMXYSjkxb65NH6wZR9ue+Ig7th2GHuOpvDtd5/t6LEQERERdQK2dBF1mzLPdDllLOLHcMgHn0fEySuj+PwVp+G1p00CAJ4+GMML03EAgKrX39I1FPJiPOqHVxKxcTyM//26k/Del24AADx7KI7/2nrIseMgIiIi6hQMuojaZPHbf4dDbx1A6rffr3kf7eg+zP7DpVj8t7+2lglS4w3Wyp6nsPC1D+DItafh0NvHMf2uNTj6D5cidc93YdiCqr+5+Fhr2nyfVqPPdRV7z8vWI+zLNZn9dvtRR9IkIiIi6iS8vZCoDTKP3o7Und+oax9DTmP+y++FkU1B8IeXVnh8DZUh8fN/QfyWTwO6tpSHkoXy/ENYfP4hpB+6DSN/9yMIvgAmBwIYCHgQy6iYTeY68mjkma5S/B4Jx4yH8fSBGA7OL3/mi4iIiKjbsaWLqMUyj/0a8198F2DUHrTomQTmPvsXUJ5/aNm6Rp7pSt39HcR/8ClA13Bg5CX46qb/h0OffAyj//deBDb/GQBAfuo+xL73SWsfs2XLK4kF8+XEMwo+8uNt+ItvbMHtT01X3Dar5OrC7+WfJCIiIuo9/IZD1CKGriP+43/G/OeuBIq6fa9E2fccjn7iFZCfuq/0BnU+06UtzljBlHjyS3Hdqn/Ar/UT8Zt9BnzHnYPhj/wnAuf/OQAgdc93oCfmsHX/ItJKrkVs7XAwV64qLV0RvwePTc1j++EE7nzmcNntDscy2DWTBACcXOElzURERETdikEXUQtk/3g3jv7dRUj85P8Bhg7vxjOr7qMn5hH77idw9GMvg3ZgOwR/GNLqE5ZtV+8zXenf/RBGahHwBjB67ddw9vGrAAC/3HrICn7Cr/9APm0vEtufxGfv3A4A8EoCztkwBCDXe2ElgiDgspfkOuD4w45ZPLJ7btk2qq7j/97xAjTDgADgTWeuqutYiIiIiLoBgy6iFpj7v2+CuvcZQPIicsUnMPTBb1fdJ3nH15G8/UZAleE5ZhNG//lueI8t0aV6nbcXZh76KQAgeMGfwzOxAR981bEIeEVkVR3v/e4T+PFj+3F47DT4vrEfz/79Vrx3SwTPHcr1WPj601diKJR7hkzJd7Rxy6P78aavPYw3fe1hq5MN0/v+5BgMh3Ll+9CPnsK3HtiD3UeTmE/JeHjXHN5385O4/8VZAMDbz1+LU1YN1HUsRERERN2AHWkQtYIgIHDuGxC98h/gWX0C1CNTNe0mDq9A5E0fRejV7861aJV4DkyooyMNQ5Wh7HkaAOA79U8AAMdORPAvbz0dn/jJViymsrjh19uX7ScJAq6+YB1OXztoLTOf6VpIydgzmwIAZBStYL/RiA83XnUGPvTjp3A4lsWN9+3CjfftWpb+2zevxQdffVzNx0FERETUTRh0EbXA+Bcfg2fV8XXtE3zF2xF5098WBlWl3o1Vx+2F6sEXrfd8eVYeBz2TRPJXX8HGP/wEP5jeBd0ADoQ24NdDr8Z/j12KscEQzjtmBFeetwYhn4TnpxNLadXYe+FJK6P48TXn4UePHsC9z89g71wKqmZgLOrDWeuG8NZz1uDU1WzhIiIiot7FoIuoBeoNuADAM7F++cLili6PB4IglNx/1VAQT3zy4oJl+vxSL4JGNomjf3cRtOmllicRwNrEDrwvsQN/PfwCht/zA4iBXPf0s4nCzj/Mlq73v3wj3v/yjRWPJRrw4r0v24D3vmxDxe2IiIiIehGf6SLqJkVBV73dxevppZaqhRvfD+3wHoQv/zDGv/oUVvzgKMa+8AgCL70CACBvuw+L37je2t4rFQZ3Tr2ni4iIiKjXMegi6iZGUY+BdQZdhpyypvX5aQxe82UMXPUpeCY2QPD44F17Moav+xaCF78LAJB54MdQdv0RwNL7uaz9DUDTK/dgSEREREQMuoi6ilH0TJfgqe8OYcEXsqY9609F6OKrS24XveofAVECAGS2/DK3vbT8Nka2dhERERFVx6CLqJsU315YR8+FACAGI9a07yUvK7udNDAGz+oTAQDK/hcAAB5RQHHYpVR5VxcRERERMegi6i7FHWnUeXuhZOucQ/AGKm4rhKK5CTmdmxeEZa1dtfZgSERERNTPGHQRdZPiLuPrDbomN0II5Fq7tMN7Kme1cAQAII6sspYVP9el8pkuIiIioqoYdBF1k2W3F9b5TJcown/GJQCA7NZ7oGcSJbdTp3dBO7wbAOA74TxrOXswJCIiIqofgy6iblLUe2G9z3QBQOjV/yOXVDqG2Hf/fnkWuo7Yzbnlgj+MwPl/Zq0rbuniM11ERERE1THoIuoihqEVLihze2Hyzptw5INn48gHz4a847GCdf7TX4ngy94GAEjf8x3MffZKyC88Aj0+C3nHY5j/f1cg+9gdAIDoO/4JYmjQ2tcjsqWLiIiIqF713ZtERO1V9AxVuZcj67FZaAdfBAAY2fSy9YN/dSMMTUHmwduQfex2ZB+7fdk24Td+EKE/fW/BMrZ0EREREdWPQRdRNynuvbDOZ7pMgseH4Q9+B5k/+Quk7rkZyouPQk/MQxycgO/4cxC67H3wn/LSZfst672wuGMPIiIiIlqGQRdRG3gm1mPlj2N17WNoKsKv/QDCr/2AtaxcS1f0rX+P6FuXP69VLHDWZQicdVnNZWBLFxEREVH9+EwXUbfQlOXL6uwyvlnsvZCIiIiofgy6iLqEoRYFXQIAqbWN1cve08WWLiIiIqKqGHQRdYvioEvyQBCE0tu6pLj3QlU3oBsMvIiIiIgqYdBF1CWMotsLG3lHV7OKW7oAtnYRERERVcOgi6hbFAddLX6eC1jeeyHA57qIiIiIqmHQRdQllj3T1eLnuQBAFIQSL0hmSxcRERFRJQy6iLqFphbMtuP2QoDv6iIiIiKqF4Muoi5hqHLhAk/rby8E+K4uIiIionox6CLqFh3wTBfAd3URERER1YtBF1GX6IRnugDAK/JdXURERET1YNBF1C2WPdPFli4iIiKibsCgi6hLFL+nC23rSIPPdBERERHVg0EXUZco7khDaNfthey9kIiIiKguDLqIuoBhGIBaeHsh2taRBlu6iIiIiOrRnp/KifqQrutVB03TcgEWcoGWOeiqgux8AgaWAhzf4RkI3ljJvARBKDuIomiNqw2lFL+ni890EREREVXGoIv6kq7rkGUZsiwjm81CURTIslwwNgdzXlVVayg3r2lawaAoihVM6U7fhvf7bc6mV0QQBEiStGwQRBGyJkCURIiiBFHy4MVoAB6PxxruvfdeiKKI17/+9dYyr9cLr9dbdtrn88Hr9SKbzUJV1VxeglC9oEREREQdjkEXdRVN05DJZJYN2WwW2Wy27HQ2m7WCLDOI6hSCIEIQhdxYECCIIgQgH3AIgCBAEJbmqwUiSy1lOgwDgGHkWshsLWe5QYeh58bl0jGDymrmDxfO33fffQCAoaGhqvuWI4qiFZCZQZk5FM8XD36/v+S83++Hx8M/e0RERNRa/PZBLadpGtLpNFKpFNLpNDKZDFKpFDKZDNLpdMFgLjODK6eDJUEQIHm88Hi9ubHHA0nyQvJ4IHlyY09+LEoeSJKUH9umPRJE0ZyXIIq51iAYBn5/x48BABdf/k74/P7cunxgJQpiPqBqb2uOYeQCMt0WhOm6DsO67VGDrmu5+XyLna5r+encWDPnNQ2apuKue34LGAaOPeVMaJqaW66q+encWFM1a15Vldx6VbGCRl3XrYDZSZIkFQRj9oCs3Hy5wev1tv38ERERUedj0EVNURQFqVQKyWQS6XQayWQSqVSqYDADLHPaiS/RHq8XXp8fHq8fXp8fXq8PHq8PHp+v9LTHmxubwVV+2gyC3KDaAsRgKAKPtz0dX1Qj5JrRIEEEJGfSHBgaAQCceu7L6t5X1/MBmmoGY0rBtKoq0BTFCtRUVYGqyPlp2VqnKuby3LSWf8+ZGfSn0+mmj1MQhLIBWSAQqGt5uWfoiIiIqPsx6CKLYRjIZDJIJBJWIGWOp6ensWXLFit4Mtc10/Lk9fnh8wdyQZM1HYDX74cvP/b6AvD6fNY2ZoAl8AtqzxJFCaJPgtfndzRdQ9etAM0MyJbmbcsUBYq1TF7aTpYLlgNLn5lMJtN0+cq1rlUL3nw+H2KxGJLJJG+fJCIi6lD837mPaJqGZ555BnNzc0ilUlZwZQ+yGunsQRBE+AIB+PxB+PNjnz8Arz8An39p3j54ff6eD5wEUcCKtRutaWovQRStwL1ZhmHkWtaUwkBMkWWoqgxVlpetWwraskuBnSxD1zUAsJ43jMfjTZVNkqSC2yQDgUDZgK7SbZXms3O8fZKIiKh5DLr6yLZt2/DTn/606nYerzcfQAXhCwThD4RywVLAtsyfG/v8AXi8Pn4xK0GSPNh88evbXQxygSAI8HhzzwIGEG4qLU1Tba1s+cDNHpgVBGzmNrn1ZoCn5G+jzKWnWS3SzRJFsezzb6XGlToxMQfeRklERP2o7qDr4x//OH72s59VTtTjQTAYxMTEBDZt2oRLLrkEF198cdntdV3Hr371K/ziF7/AM888g0QigdHRUZx88sm4/PLLcemll/JLvQPML2Hh6CBWrT/OCqL8gVBBQCVJjMWJWkXKd8ziDwSbSsd++6Sy7JZJc5lSeNtkvmVOUQpvuzQDOF3XHbt90lTcG6W9Va1Sj5T29aWm+YoBIiLqZK58u1ZVFfF4HPF4HDt37sRtt92G8847DzfeeCMGBgYKtl1cXMS1116LRx99tGD59PQ0pqencd9992Hz5s344he/iNHRUTeK23eGx1bglLMvancxiMhB9tsnmwvf8q8LMIMwtfB5N83+XJxqPvOWu9VSs29fsJ0C5HulNN9/50RLnJ39FQP2YKz4tQP2efN5Vft6+7vjigePx8PAjoiIGtJU0HX99dfjhBNOWLZc0zQkEgk899xz+MlPfoJ0Oo0tW7bg2muvxc0332z9p6XrOv7qr/4Kjz/+OABg9erVeMtb3oK1a9fi0KFD+PnPf46dO3fikUcesfb1+XzNFJmIiKoQBMGx59+AXBCn65otaMsFZ+ZzcdZyq4dKtbBHSlVdto3Zc6WRfw7VrVcMFKv0gu9y6yots79UvNRytuAREfWGpoKus88+G5s3b664zdve9jZcddVViMVi2LJlC37961/jta99LQDgtttuswKuzZs34xvf+AYCgYC177vf/W5cf/31uOeee/Dkk0/iV7/6Fd70pjc1U2SiltE0FVsfyr0keNMFr+Rtm9S3BEGwbqNE0+1whcxXDJivBSh+xYD5/jc1P85tkx80pWDa2kZVl7bTVCuwA5Za6px45UAtBEEoCMyKAzT7IEnSsoCt1L7mcq/Xa00Xb1u8nM/iERE1x/VvgccffzyuueYafO5znwMA/PKXv7SCrltuuQVArretz372swUBF5D7RfEf//Efcc899wAA7rjjDgZd1DUM3cC+nc8BAE7f/ArH3oFFREvcesWAna7rtmBNKXzhtxnIWS8AV6DnX/xt7mO9EFzToJsB3bJ5DbqZbv6dckCuldAM9NpJFMXcC+Dzt3EWB2X2sX2oZV2tg1mGSsvZKkhEnaolP71feumlVtD1zDPPAMjdgrhq1SokEgmMj49jxYoVJfednJzEyMgI5ubmcPDgwVYUl4iIyCKKIkQx97L1VjBvx9Q1DdlsGvfcdjMA4GWvuQIQkA/ktKXgT9Py2xcHb7k0dH1pW13ToOlF25Sc1wrKpOu69UoRt2/hbEbuXIkFwVgt4+L9ys3XMwiCYE3Pzs4iFouVXF9pXOt0rQOAZfNE1BotCbomJyet6fn5eQC51q0vf/nLVfdNJpPWe2vGxsbcKSAREVGHsN+OKQhLt/UNDI/B4/W2pAyGYcAw9KXAzQzG9MJ5TdNg6Hp+nAvoDM0c67mxbbvc/vpSWroO3Uwjv60Z4BkF2+oF63RdK7jt02SuV1W1xFFRsXIBWanpesb29EtNV1pmLldVtezL3lsRMBr5zn+cUE95K9VJuW3MeUVRrL4PircvdY4qncdap0vNl1tX7YeC4nRq+WGh1A8V9fxgIYoiZmZmoCgKjjvuuJrPUyNaEnTZX/YZjUbr2verX/2qdVvFZZdd5mi5iIiIaLncFxIJoigBrYnz6mYGhrmAbCmgK1xmzmtL8wXL8/NGbmzfxjAM6IYGQ7flYxgF+5vz9ulcWkbResM2FK03lqZRNG8OKNi/eH558FlfHToXXBB1s09/+tOupt+SoOvhhx+2pk855ZSK22qahqNHj2Lbtm24+eab8cgjjwAAzjvvPLzlLW9xtZxETpI8Hrz2ymusaSKiegmigBVrN1rTtKQgMKSlYAxmYJYPqnITJbYBgKUgLrfetiy/wLCPkU83P85vZO23rDylyli4YGmyxFTpA628utsY9gMyCtdYU0bJDarXsXlebedrabrwvBVOG1g617CumcJzv3SdGCWur6Vtbdejbd5KN59XwQ8KtnWFy6v8CIH8jxAGCn64KP+jhQ7Dtq3bXP8mGI/HC24jfN3rXldx+8svvxzbt2+35gVBwNvf/nZ8+MMfZnfx1FXMbreJiBolSR5svvj17S4GdQFBEABBAENzos7kStCVzWZx+PBhPPLII7jpppuwb98+AMBJJ52EN7zhDRX3PXToUMG8YRh4+OGH8dvf/rZqwEZLDhw4gPvvv7/gvnbzebqDUzsw86N97SoaVfBY/iXh55x7bptL0pxD+3YBAO780TfbXBIiagkBCEcG4fMHqm9LRNSBzjnuL11Nv6mg6+qrr65523Xr1uErX/kKvBUeAlZVFe9617uwceNGZDIZPPjgg7jjjjuwY8cOfPjDH8aLL76ID37wg80UuW889NBDVk+RxXRdQzaTanGJqBbbtz8PADjttJe0uSTNWb9uHQDwOiPqI9k0P+9EROW4fnvh2rVrcfnll+Pqq6/GwMBA5cJ4PPibv/kba/7Nb34z3vKWt+Caa65BNpvF17/+dbz0pS/FOeec43axu56W7+73pLVDWD9RX+cl3ULVdNzzxwMAgFedsRoeqftf3vmHe0MAgEvPXtvmkjSnHeXvxeuBqBscnE1i2545jI+P48ILL2x3cTrS7OwsRkdH212MrsI6qx/rrDGtqremgq7rr78eJ5xwwrLlfr8f0WgU69atw8jISDNZ4IILLsB1111nvefre9/7HoOuOqwejWDTMb3Z1b6sataX7NM2jMLn6f6HqYfCuecWe/WcuakXrweibuARRWzbM4eBgQH+/1zG1NQU1q9f3+5idBXWWf1YZ41pVb01FXSdffbZ2Lx5s1NlKeuKK67AF77wBei6jieeeML1/IiIiHqJrhtQNB2KqltjVdMLlqlF6wu2y8/LBfMaFDW3jIiIKuuKfqwHBwcxOjqKmZkZzM3Ntbs41CFEQcBxqwataepvvB6oWxmGAc0eFNkDH1WDohlWgFMcFBUHSssDKQOKqkNvQXfIGzZscD0PIqJu1baga+fOnfjmN7+Jffv24c///M/x5je/ueL22WwWABAOh1tRPOoCHknEmy7c2O5iUIfg9UCtouZbfGQ139KjaPl52zJ1aZmSn1YK1hcGVnqL3j0kCAK8Xi98Ph+8Xm/BYC4rXmdfvri4iFWrVi1b7vf7EQqFWnMQRERdqG1Bl6ZpuO222wAAkUikYtC1e/duxGIxAMAxxxzTkvIREVHvMAwjFwQpGrKKhqw5rWqQFR1ZRYOsavlxbl5Rl5bLip5f726rkSRJViBjDsXBT6kAqVygVLxMkqTc+5waxGdGiIga07ag6/jjj8fKlStx6NAh3H///di3bx/Wri3d49m///u/W9OvfvWrW1VEIiJqM8MwoOpGLvgxAyZbYJQtsSy3nZ4PqJbWOc3eymMfF0/b571eb8E6c94eFBERUe9pW9AlCALe9a534YYbboCqqvjoRz+Kb37zm4hEIgXbffe738VPf/pTAMD4+DiuvPLKdhSXiIjqpOmGrWVpqUXJHgxlrRak5fPmNk7eeidJEvx+P/x+PwRBwMDAgDVfbfD5fAgEAlbAxACJiIhq1daONN7xjnfg7rvvxmOPPYYnn3wSr3/96/G2t70NGzZswNzcHO68805s2bIFQO4XxS9+8Yt8possqqbjrif2AQD+9Ky1fC9Tn+P14BxdN6xb6ey33S1vTbLdpmcLrMzb9lTNuWhJEISCACgQCBSMzeni+eJtvF6vlSZvlSMiolZpa9Dl9Xrx9a9/HR/60IfwwAMP4NChQ/jSl760bLvR0VF86Utfwrnnntv6QlLH0g0DT0/lerN89Zlr2lwaajdeDzmGYVgtRllFRcZ+C56s2eYLg6VcIJV/lklz9lY88xa6SkFSqWDJvo3X64UoMpAmIqLu1PYu4wcGBvDNb34Td999N376059i27ZtWFxcRDgcxsaNG/GqV70KV155JVu4iKivqJqOtKwhI6tIZ3PBUzqrIiNrSMu5+Ux+faYooHKKx+Oxbqcr14pUqlXJ3N5cx9vwiIio39UddN1www244YYbHC2EIAi45JJLcMkllziaLhFRp9B0A6msgmRGRSqjIJlVkcqqSGVUpPKBVSqbG6dltemOH8yAqdRgb0EqbmmyB0seT9t/lyMiIuoJ/B+ViKgJumEgmVGRSMuIpxUkzCGTGyezChLpXCBVL1EUEQwGEQwGEQqFCqYDgYA1jsfjWLduHYLBYMHteERERNQZGHRR1/JKIq5742nWNPU3t64HWdUQS8pYTMmIWYNiTScyCmp9bZMoigiHwwiHw4hEIta0OYRCoYJpv99f03NMU1NTWLduXZNHSkRERG5h0EVdSxAEBHy8hCmn0evBMAwkMgrmE1ksJGQsJrOYT2axmMgFWqls9RYqURQRjUYxMDBQMDaHSCSCaDSKYDDIziCIiIj6EL+x9riMrMIwDAiC0O6iELVVRlYxF89iLp7JjRNZzCcyWEjIVXvrCwQCGBoaKhgGBwetIRKJMJgiIiKishh09SjzAfg/PDuNp3bP4tiVgzhu5QDWTUT5/iLqWYZhIJVVcTSWwWwsg9l4BkdjGczFMkhWaLESRRGDg4MYHR3F8PAwRkZGMDIyguHhYQwNDSEYDLbwKIiIiKjXMOjqUS996UuRTqexa9cuxNMK/rjrKP646yi8koj1E1Ecu2oAG1cMIhrs3oftdcPA/pkEAGDNeAQiW/P6SlbRMLOYxsxiBkdjufFsLI20XL7L9Gg0irGxMWsYHR3F6OgohoaG2FMfERERuYbfMnrU5OQk3vGOd0CWZezevRsvvPACXnjhBcRiMew4tIgdhxYB7MPkUBAbVw7i2BUDWDkS6qrbEFVNxy2/3wEA+ODlp8Pn4buAepFhGJhPZDGzmMaRhTSO5AOtWEouub0gCBgeHsbExATGx8cxPj6OiYkJjI2Nwe/3t7j0RERERAy6ep7P58OJJ56IE088EYZhYHp62grADhzYj8MLaRxeSOOh56YR8nuwccUANq4YwIbJKDupoJZTNR1HYxkcXkjlAqz8UO6Zq4GBAUxOTlqDGVz5fL4Wl5yIiIioPH6r7iOCIGDlypVYuXIlXvGKVyCRSODFF1/ECy+8gB07diCVzeLpqTk8PTUHQQBWj4axccUAjlkxgInBYFe1glHnk1UNMwtpTC+kcXg+F2QdjaWhl+h+3ePxYHJyEitWrLDGK1as4LNWRERE1BUYdPWxSCSCM888E2eeeSZUVcXevXuxfft2bN++HTMzM9h/NIn9R5P4/dOHEA54cMxkLgBbPxFFyM9Lh2qXVTQcWUjj8EIK0/MpHJ5PYy6eQanXW/n9fqxevdr6gWDFihUYGxuDJPH2USIiIupO/OZMAHItCRs3bsTGjRtx2WWXYX5+Htu3b8eLL76IXbt2IZlRrFYwAJgcDuKYiQGsn4xi9Wi4LT0iSqKIizettqapM5gB1vR8PsBaSGEuni25bTQaxcqVK7Fq1SprvLCwgA0bNrS20EREREQuYtBFJQ0PD2Pz5s3YvHkzVFXFli1bcp1w7NiBw4cP4/B8Gofn03j4hcPwiAJWj0WwfiKC9RNRTA6FIIru34ooiQLOOX7C9XyovKyi4fB8CofzQValAGtgYACrVq0qGKLR6LLtFhcX3S42ERERUUsx6KKqPB4PVq9ejQsvvBAAEI/HsXPnTuzYsQM7d+5EIpHA1JE4po7EARyCzyNizVgEa8cjWDceaVkQRu5KZVUcWcgFWIfnU5heSGMhUTrAGhwcXBZgRSKRFpeYiIiIqDMw6KK6RaNRnHHGGTjjjDNgGAZmZmawa9cu7Ny5E3v27EEmk8Gu6Rh2TccAAF5JxMqRENaMRbBmLIyVI2H4vXw+p1MZhoHFlGz1HGj2JBhPKyW3D4ajiIxMYGB0AtGRCZx7ykasmxxpcamJiIiIOheDLmqKIAiYmJjAxMQEzj//fOi6junpaezZswe7d+/G1NQU0uk09s4ksDf/ImMAGBsIYNVoGKtGwlg5EsJoNMDWsDbIKhqOxjL5lwynrXdhyWrpLtpHRkasZ6/M57CePyojkV16IbEvEGpV8YmIiIi6AoMucpQoitYX8gsvvBC6rmNmZgZ79+7F1NQUpqamsLCwgKOxDI7GMnhq9ywAwCMJmBgMYcVwEJPDIUwMBjE6EKjYQYei6vjWXc8BAN7zpyfD62FnGuXIqobZWBZz8QyOxnIvFz4aK/+CYUmSCrpmN3sRDAQCy7b1zqsAloKucu/UIiIiIupXDLrIVaIoWi+uPffccwHkngnbv38/9u3bh3379uHgwYOQZRkH55I4OJdc2lcARgcCGB8MYmwggNGB3Hgw7IMoCDBgWEGDUbLz8f6iGwbiKQXziQzm4lnMJ3JB1mw8g1iq9K2BQO52Ufv7ryYnJzE+Pl5zF+3eosBY0XguiIiIiOwYdFHLRaNRnHzyyTj55JMBALquY3Z2FgcPHrSG6elpZDIZzCzmBjuPKGAo4sdQ2Gct2z+TwOhAENGQF2KPvsTZMAxkFA2LSRmxlIzFpIzFZBYLSRkLySwWkzK0Um8WzguHwxgfH7duB52cnMTExARCoeZuB/RKhfWtsqWLiIiIqACDLmo7URQxPj6O8fFxbNq0CUC+M4fFRUxPT2N6ehozMzM4cuQIjh49ClVVrdsTTT/5w65cWgIQDfkwEPIhGvTmBx8iQS9Cfg/CgdzY5xEhdEhwphsGMrKKdFZDKqtC1Qw89uIRpDIq4hkF8ZSMeFpBIq1UvXVPkiSMjIxgdHQUY2NjGB0dteq22eCqHE/Rs3hKhcCPiIiIqB8x6KKOJAgChoaGMDQ0hJNOOslarut67pmwo0cxOzuL2dlZHD16FPPz81hcXISmafkWoNLPKpkkUUDAJyHo8yDgkxDweuDzivB5JPg8ubFHEuCRRHgkEZIoQBQFiEKubEsBmwHDMMtmQLMGHapmQNV0KKoORdMhqzqyimYNGUVDJqsioyw9DyVFxiEBuHfrgbJlj0QiVt0MDQ1hZGTEGgYHByG2+EXRy28vZEsXERERkR2DLuoqoihaAUYxXdcRj8etACwWiyEWi2FxcRGJRAKJRALJZBKynLsNL5lRkcyobTiK0gKBAC655BJEo1GEw2GEDBkRI4to0I+BoA+DK9dh9JRz4fV6213UAsVBl8pnuoiIiIgKMOiiniGKIgYHBzE4OFhxO1mWkUwmkU6nC4ZsNotsNgtZliHLMhRFgaqq1qDrOnRdh2EY0PVca4691UsURXg8HkiSZA2yLGNkZARerxc+nw+BQAB+vx+BQACBQAChUAihUAjBYHBZxxXK7q3QZg9a89LoSMcFXMDyZ7rY0kVERERUiEEX9R2fzwefz4fh4WHX85qamsL69esb2tdQC2+RFKTO/LgWd+vP3guJiIiICvHFRkSdSiu69dHjK71dmxW3dBlgaxcRERGRHYMuog5lqIXv1hKkzru1EMi92LqYyh4MiYiIiCwMuog6lVb0QmNPhwZdooiiXuP5ri4iIiIiGwZdRB3K0IpbujrzmS6gVLfxbOkiIiIiMjHoIupAhq4BxbfodejthQB7MCQiIiKqhEEXUScqep4LAIQO7UgDyN1iaMeWLiIiIqIlDLqIOlBxd/EAgI6+vbCwpYvPdBEREREt6dxvceS4x/bMAwDO2TBc03Kn0ncjnxem4zhxRdRKq568y03Xkm+1bYvTVtQgZuqod/PYDMPACWNn4gx9Lx4/okKa3IDzhOW9BNZzLKXqrR6V9m/mXV1uXjdOpO3UddtOlY6h0c9Do/l2an06Ua52HVsrzlu70urU66XTsd5ar911/tieeTx7QMcpxnzVv/VO5ul0mr2MQVcfuen3uwEs/3CUW+5U+m7kU/zlv568y03Xkm+1bUulbaql3s1jA4ATV0Rx05tPwLce3A5h9zTOO2FVzcdVrmzNBF2V9l/2TJdee0uXm9eNE2k7dd22U6VjaPTz0Gi+nVqfTpSrXcfWivPWrrQ69XrpdKy31mt3nS/9H1066HKjfO0+5m7DoKtPPLZnHo9PLVjT9haiUsudSr/W9Y3k8/jUgvUrS615f//hvSWnq5WplvKXy8e+vlK925eZx/TDbTE8sT9RMt9y+VWrA7Pe6jkH1fZvtPdCN68bJ9J26rptp0rHUM815ES+nVqfTpSrXcfmdL5OptdsWp16vXQ61lvrtbvOq/0f7Ub52n3M3YjPdPUJe6tLLdNOpe9mPuZ8o3nXU6Zatm22HPXWWz3HUmnfWlTb31P0oq5an+ly87pxIm2nrtt2cuoaciLfTq1PJ8rVrmNrxXlrV1qder10OtZb67W7zqv9H+1G+dp9zN2IQVcfKNWC8tie+bLLnUq/1vWN5mOmVU/eiaxWcrpSmWopf6V8ivcrlV6plrFKZayUX7U6KHcM5dSyfyMtXeXqNRAIVFzfSJkbuSadum7bqdIx1HMNOZFv8TXeKfXpxHlu17XidL5OptdsWr3w+WsH1lvrtbvOq/0f7Ub52n3M3YpBVx8o9yt/LS0izaRf6/pm8qm2XT351NLaUW5ZPWVrpj4qtYqVS6+e46pnO/vyRnovLFcPQ0NDFdfXwolr0qnrtp0qHUM915CT+TqVh1OcKFe7js3pfJ1Mr9m0OvV66XSst9Zrd51X+z/ajfK1+5i7FYOuHlfqFxBgeQuRfXkzv0YWp1NtfbP5lFIt72r71ZJvpV+RquVRrUWr2v7lWsVKla9S2Wo5B7XuX9zSpRmAVvxy5xrSfXxqAdsOpZq6bpy4Jp26btup0jHUcw05lW+l1t922T6nNX2e23WtOJ2vk+k1m1YvfP7agfXWeu2u82r/R5f7W+9kK7YTafYLBl09zs0WpWrblmvVcTqfRvKuNZ9ayt/qX3eabVWrJ61a13uKWroAQK3Qg2HFev1dc9eNE9ekU9dtOzlxDG7//Wg0D6fcsnWu7Don6sjNY3M6XyfTazatXvj8tQPrrfXaXefN/n/odJ68zipj0NXD6m3pMTX7a6Q9nWrrncinkbxrKVMtx1dLi4HT6mkVa6Y1o5Z6N/f3iAKKw65yz3VVrde9jV83TlyT1c5pN/yaV60e6rmG3GqRbjQPpzy2Zx5P7ouVXd9sS3CtaTTC6XydTK/ZtNpVp92O9dZ67a7zWv7eVvpb72QrdjNp9hMGXT3M6Xv7nUy/1fnUo54WrF74VceJZ74EQVjW2qWUea7LzeumV6/JejlZPjdbpJvdrxlOnOd2XStO5+tkeq1oyer0z187sN5ar9113qr/7+rdntdZeQy6elSjrVymZn+N7LR86lFPC1atLQadrJ5n2SrtPxr2FSxXS7R0uXndOJV2Lee0k3/Nc/oz42aLdL15OKXWsjbbElwtjUY4nW+96Zm9i7pRtnbVabdjvbVeu+u8Vd/BGsmT11l5DLp6lNu/gLTql/R2/WLSb7/U1PMsW7n91w4HC5aVauly87rp1GfrWq0bWlbc2N+tvJxoVe3klsd60zN7F3UiLaf371est9Zrd523429Ku4+5FzDo6kFu/wLSql/S29HKZeqFFqx61PMsW7n9nz0Ux3DIay0rfqbLzeumXS2infZrnlv10IoW6U7o7asUJ1pVnTo2p/NtJL1th1KulK1dddrtWG+t1+46b8fdDO0+5l7BoKsHuf0LSKt+teEvJa3VbG+MN/1uNzaMhaz54nd1uXndsEU0p5t6y3MzHafzcKJVtR3PX1Tbp6H0fufc85T2fdpVp92O9dZ67a7zdrSctfuYe4VgGMbyBy962LZt26CqKs4888x2F6WrTE1NYf369e0uBnU4WdXx8O45GAYwGfXjpJXRdhepL/DzSdTZ+BmtH+uMWqVV15rH9RyoLbLP3A8A8L/kZTUtL7fenA/H48gm9pbdr5KFb30EADD0ni9UzK/cdLWyldum2rE2sr19n2Kl9pufn8fw8HDJsit7nrK282443VqnZzPQEnOArkHZ+QTEUBSC5K1YtlrLnrj9RgBA5HXXVj0+e52YZa2035FYBjuPpmD+jKPoek3nsdjMzAzGx8crHkelcpsauVbt6dT72XEiDyc08xlohUr5tboshw4dwsjcjrrzbHU5W6nasR06dAgrV66sax8iao1G//+k1mDQ1aMSt34GwPL/BMstL7fenIdhICEIDf2nmv7v/wBQOuiy51duulrZym1T7Vgb2d6+T7FS+8ViMQwPD5cse7mgy0gtQNu9FQCQuv1GCNLSx7TW81at7JWCp1J1UkvQpRm5li6Tohk1ncdiqVTpZ0aqKT4vjX4BbPSz40QeTmjmM9AKlfJrdVlkWW4oz1aXs5WqHZssy3XvQ0St0ej/n9QaDLp6UPaZ+yE/+4A1bW9dKbW83H4ArHn7NvX8x7rwrY8AumZN2wMve36J228sOW3mV6lspbYplUaleqpl++J9Sq2rlk+5epWffWBpf00BAGhH90N54eGqeVQ7r6bE7TfCSMWs6VIBVLk6sadRLvDyFr2nKzr1SNXz6NSXtFLnpZH0G/3sOJGHE5r5DDhdlmrlK86v1WUBAP++rXXn2Y5ytkojx9bL9UFE5CR2pNGD7L/41zJdbn21bWphtnIVT9dbzkplK7dNtTLXu32j62qpV/t2hpoLutJ/uLWmPBope711Ukv6HrHwz8nAf39p2X61lrVeTlyrxfs0cn00k4cTnPoMuKVSfq0uCwBId32t7jzbUc5WaeTYerk+iIicxKCrxxT/4m+2oJRbXmm/Ui06xftVYm/lAgDomvV8V3F+ZgtM8bT87APLWlyKy1Zqm+I0istcKf9yx1iplavcfvZf0kuVveT+krdkK1epPKqdV5O9lQvIHa/5fFe5tOzbV9rPZG/pmjj4GPTn/1BQrlLnsdZrqZJy56Xe9Bv97DRTVqfqoFTa9X4GnCxLLeWz59fqsph5qs8/WFee7ShnqzRybL1cH0RETmPQ1WPK/eLfTAtJLXmUUtyyZV/mdH7Vtqn0q3qt6TVSDvsv6bVI3PoZiAOjJVu5SuVRazlrWdbIL9t2Xmnpz8nIvf/acLnq1WjLZC3bOtXiWy0PJzjxGXCzpaJSfq0uS6N5tqOcrcL6ICJyF4OuHlLpF/9KLQHVWnDK7VfJslYuk65h9nNX1pVfqRaXerep9Kt6te1r3adUPvZf0mshP/sAUnd/u2QrV3EetbbwFLdymeytVvVcA+Vauzz5lq7iVi77ftXKWq9GWh/rSceJFt9a8mi2daC4RbWUWj4DbrVUVMqv1LODbpbl/2/vvsOcqNo2gN+Ttr3vslQBkQVFQUQBBQQFQcCC2BGwfTaKoMALiqBiAaSpvKhYsFNEwRc7Ik2kCtKRvvRle9/0+f4IGdI3vWzu30WuncycmTlzchLy5JSpLT/OzhnsMgsmlgcRUeAx6KpDvB3D4uvYF0cctXKZabf95PH5/MHVr+qu0nuyjzfncXVeV2l8Hb9luc3TvDpKLxMEKGSCw1YuT4/lz339lcbXffzVIueIuy2q7rzegWip8Pc4SV/5OjbTk22RgOVBRBR4DLrqCE9bq8xcjTGqbT9nv2Y6beUKMVe/qjtL721LoCfnseVOy15tr5s5785auSzPVfrJGI/z6qy1q945x61crnj7y7g3rY++HMfT43pyDl/HibnboupO3fR3S0Vt1+6qfgai1cSb1yKQr1+osTyIiIKDQVcdEYpfFp2d01UrV6h50zoRjFaQQHA3796+Xo6OnWIxY6Gvx/LnPoH6Vd7XcXD+yIc/W9x8zUsgjhXIGR7dTRPsMgsmlgcRUXAw6KoDvP2l3leOfs0M11YuM3dakSx52xLo6XkCQbt/g3v58PL1sm3t0uz7EwYPW7nMfJ1p0Nvj+/re8WaGN1+O5e2xLbnbkhrImSU9EcgZHt05ZyBfv1Dz5trcGT9ouw8RETHoqhNC+Yui7bnDuZWL/M+f9+gJxAyate0T6NkTPT1HIFqtvBWMsgnn43gz3jHSWne8uTZPZmSNtPIgIgokBl0RLlStXGaWv2aGeysX+Z+5tcufLRqxsbEu0/lrDJa/3jvezPDmzbF8PbanAj2zZDDz4k1+vBmbGUmtO96WhyczskZSeRARBRqDrggXDr8kmvPAVq7o5O24N2fHSk1NrTWNL8f3x3FcHded9d4cyx/H9lQwxrsF63jBGpsZDp/J7mB5EBEFlyCKohjqTATTnj17oNfr0b59+1BnhYgc0Ov1UCgUoc4GEZHfnDhxAk2bNg11NiIKy8xzLDPvBKvc2NJFbikpKQl1FiJSUVGRR+kNGjV0pQXQleRBV3AKuvPHA5SzwCip0mL7iRJsP1GC3afLcLSg0uNj5OXlBSBnddv58+dDnYWIdO7cuVBnIeKwrhEReYc/J5NbysvLkZaWFupsRJzKykpkZGS4v0NVCQzHdkpPhdgEILu5/zMWKAJQqTGP6zOgUqNHi6xEjw5hMHBcoKfUanWosxCRtFptqLMQcVjXiIi8w6CLKIyIBp3Vc0GhDFFOvKOUyyCKIqq0Bmj1RsQoZDAajZDJ2KhORERE0YtBF1E4sQm6II+ct2iFWodlO87iq82nUFR1sQXh/bXH8GDHJri9XX0kxUZWEElERETkD5HzjY4oCoh6m5YueWQEKRuPFmHc0r1Q6wywnZnnbKkas1Yexrw1xzDj3itxQwsPulsSERER1QHs80MUTmyCLihUocmHBzYeLcKzi3Y5DLgAQLzwUOsMeHbRLmw86tnkIkRERESRjkEXURiJtDFdFWodxi3dC1GEw4DLkghAFIFxS/eiQq2rJTURERFR3cHuhUThxMmYrpKq8Jxl7bsdZ5y2cDlibvH6ZtsZDLymocM0ZWpD2F5vuPK1zNISwr9FlYiIKJIx6CIKI87GdPWcvSEU2QkIEcC8tccwb+0xF6lyg5SbuiTX6z13TLrZf9kgIiIiO+xeSBRO7MZ0hXf3QiIiIiKqHYMuojASaWO6iIiIiKh2DLqIwoQoioDBYL0yQqaMJyIiIiLnOKaLKFzYTqKBi2O6/ni+a7BzUytRFDF0wXacK1O7PZEGAAgAGqTE4ovHOkAQBLvtp06fRpPGjf2Wz2jAMiMiIgpvDLqIwoTtJBoApDFd4Tq73KBOTTBr5WGP93uocxOkJ8Y43FYeKw/b6w1XLDMiIqLwxu6FRGFC1GthrC6HoSwfxupyiAIgyOShzpZLt7erj1ilHPbtVY7JBCBWKcdtbesHNF9ERERE4YQtXUQhZqwqRc26haj86T0YC05K62Vp9WG88zjiug+CLCE1dBl0ISlWiRn3XolnF+0CarlBsjkwm3nvlUiK5Vg1IiIiih5s6SIKIc3OVch/+nKUf/4CjAWnrLYZS86j/PMXkP/05dDsXBWiHNbuhhYZePfBdohVOm6VEy48YpVyzH2wHa5vkRHU/BERERGFGoMuohDR7FyF4mn3QtTWAKII+3YiERBFiNoaFE+7N+wDrx9GdsZd7Rsiw2ZsUaO0OIzt0xK/ju7CgIuIiIiiErsXEoWAsaoUJbOGmIIt0eg6sWgEIEPJrCGo98GBsO1qCAi4MScT3VpmoFprgE5vxI05mUiNVzqcpZCIiIgoWjDoIvKBobzQdYLKYhjKE+xWV6/6FKKmGq5HQVkQjRA11aj67WPE93rE43z6Qp6c6Va6SrUeACAIAhJiFEhJVXBGPSIiIiIw6CLySf7/XepyuwpAvt/OJqJy8RRULp7ityO6o8E35W6lq9DorZ4nxvDjhYiIiAjgmC4i8pMKtXXQlRTLoIuIiIgIYNBFRH6g1Ruh0VuPTWPQRURERGTCoIuIfFZp07VQLgBxTqaQJyIiIoo2/CmayAf1Pj7mcvvpU6fQuEkTq3WiKKLoxZtgLDx1Yap4NwkCZJlNkPHmmrCbDdC2a2FirCLs8khEREQUKgy6iHxQ68x+iVUO0yT2H4byz1/w+HyJ/YdDkZLl8X6BZjeei5NoEBEREUnYvZAoBOK6D4KgigcEN9+CggyCKh5x3R8MbMa8ZNu9MClWGaKcEBEREYUfBl1EISBLSEXamC8BQag98BJkgCAgbexXYXljZI3eYDeJRmIsx3MRERERmTHoIgqRmKt7IX3CUkAZA0C48LAgCIAgQFDFIf2FbxHTrmfA8zTjt0O45rXVWLHrnNv7VKoN0vKmo0V4bsludJ2+3q/5qtTo0e+dv3DNa6v9elwiIiKiYODAC6IQirm6F9LGfwP1xu+g2f4LjKXnpW3yes2Q0PdpxPUYBFl8SsDzsvZgAZZsO+3xfhVqHQCgoEKD73e6H6y5yyiKeO2Hf5FXrvH7sYmIiIiCgUEXUQiJoggYtIjt0Bcx19wKUV0JRXYzKBq0gJCYHrQZANcdKsT47/bC6MFkimaVGgOMRhFfbzkFrU03Q1/pDEa89uO/+P1Avl+PS0RERBRMDLqIQkisLgP0pkkoBEGAEJcEZbO2EFSxQTm/URQxf91xfLIh16uACzC1dP1+IB8niqr9mrfz5WpM+G4fdp0u8+txiYiIiIKNQRdRCBkriq2eC7HxQQu4Nh4twturjuBIfhUA4PIGSThwrsKjY2j0BhzJr8LKfeehlAu4smEy/jnlW5Ck0Rvx8Z/H8dnGk6jWGqCQCbg0KwGHzlf6dFwiIiKiUOFEGkQhZKwosnouS8oI2rlHLNyFI/lVUMgEPHVjc0wb2MbjYxRUaPDVlpMwikD/tg1QP8X3gPGv3Eq8t/Y4qrUGNE6LxfuDr0aPVrXcD42IiIgojLGliyhERKPRrqUrmEGXAOCm1lkYftOlaJ6ZgLOlNR4fY96aYyio0KJlvUTc2DID6w4V+iVvSbEKDOl8CR7q3ARxSjm25Zb45bhEREREocCgiyhExOpSwGg98YQsKT1o5182rDOaZsR7vf+Gw4X4bV8+YpUyPNixMQRBgErue+P51Q3j8Wu3KxCn4r2+iIiIqG5g90KiELEbzxWXCEEZE7Tz+xJwlVRrMeXHfwEAd7VviLQEFQAgRuH7R0p6vIIBFxEREdUpDLqIQiSU47l89fqPB1FYqcVVjZLRsfnF1jmVH4IuIiIiorqG35CIQkA0GmGstB6nJEuOjKDrfzvPYs3BAqTGK3HftY2l9QqZAIU8OPcVIyIiIookDLqIQkCsKoHtjbFkicEbz+WtMyU1mPnbYQDAUzc2Q2LsxWGhSbEcIkpERETkCL8lEflB2af/QfUvHyBl2PuI7/GQy7SiXofKH+dBvWk5DEWnAQDy1GzEdbsPCbeN8Cn4MlaXoern96He+iMMeccgGvRQNGiB2I63IaHfM74dWxQx6X/7UaU14PZ29XF5g2SUVOuk7Qy6iIiIiBzjtyQiH6m3/YTqXz90K62oVaP4zYHQ7t9gtd5QcBKVy2aies3XSJ+4HMpLrvA4H7rju1A8/X4Yi89ardef3IfKk/tQ/funSH9hKZSXtvf42ACQV6bGzgs3Pv5hVx5+2JXnMv01r60GANzetj5evdPz6yEiIiKqKxh0EflA/fcvKJnzMCAaa08MoPS9Z0wBl0yO2C73QnVFFwhyJYwVRahcNgPGknMomX4/Mmdthiw2we18GIrPofjNu2EsywcEGRL6D0fczUMhT8mE7vguVCx8Fbpj/6Do5X7InLYOikY53l4yEREREXmIQReRF0SjEZXfTkXldzPcDri0R3dAvfE7AEB8z0cR0/4W0wYBiOt2H2KuvBGFE3vCUHAC1T+/j8SBY93OT+W300wBF4CUYe8hvvsgaVtM25uhan09Cif1gf74TpR9/DwyXv7R7WObNUiJxYbxNwIACiu0+DevQtqmlAvodGk6PtlwAp/+dQIApLQKGYeOEhERUXTjtyEiD2l2rkLhf7qg8tvpgGh0u7te1Q9zAQCy9EZQtesprZclpEKQK6Bs3g5xNz4IAKhe/YXb+RENetT8ZQrmYtr1tAq4zARVHJKHvA4A0O5bD+2Rv90+vnQMQUC8SoF4lQJ6o4gYpVx6ZCTGIF6lgNJi9kJzWk4jT0RERNGO34aIPFT85kDoT+4D5Eok3vsCUkd/Wus+oihCs+sPAICq5bUQLFp/LO/PFXtdPwCAIT8Xutw9buVHf+4oxJpy0/6dBzhNp7qiC6CKAwBo/vndrWM7U6HRWz33ZBKNxdtOY+B7mzHwvc3Ye6bcp3wQERERRQJ2LyTylCAg9rrbkfTgZCga5UCff6LWXQwFJyBWlQIAZBmNrA+XdHFGQWXzdtKy7thOKJtdVeuxRYv7fcmzmjjPtkwOWUIqjNoa6E/sq/W4rlSqvQ+6Squ1yC2qBgCodQaf8kFEREQUCRh0EXkoa87fUDRs6dE+hvyT0rIsJeviBpkAWWLaxaep9QG5EjDoYMjPdevYQlyitGysqXSaThRFiNWmliVD8Rk3c26vRmuA3uYeY5wunoiIiMg5flMi8pCnARcAGCuKpGVZ7MUgSRafCkEml54LMhmEuESIlSUwXmgZqzU/DVoAyhhAp4F29xrEdbrDYTrdkb8haqpM+ampsNveMDUOOybdXOv5Km26FqrkAmIUpmt4uvuleLr7pS73dyeNP/YhIiIiChcc00UUBKJOffGJQiUtypIz7NIKF8ZdiTqNW8cWVHGI7dAXAFC99itoD2+zP79eh4qvX764Qq9169iOVNh0LUxkKxcRERGRSwy6iILAsjXLksxiPJcvku5/CUJMAqDToPi1O1G54l3o83NhrCyGZu86FE25Ddr9GyBLb2jawSLw85Rt0JUUw6CLiIiIyBV+WyIKAiHG4kbH5lYmmQAhIc0uraitMe2jinX7+IpGOUgb8yVK5jwMsaYCFV+9hIqvXrLIgAyJD0yC4ewR1Kxf5NGNl63yJop23QvZ0kVERETkGlu6iIJASEiRlkWNaeY+WWKa1dTxgOmmy+KFyTAsp5J3R8zVvZA1azPiez8OeVZTQKGCLK0+Yq8fiIzXViJp4DgYKwpNx06t79V1qHVGTqJBRERE5CF+WyIKAkWDy6RlY/mFwMdBUGUszQMMOgCAPNP59O/OyDObIOX/5jjdrsvda8qPF5OBAPZdC2MUMmkSDSIiIiJyjC1dREEgS0yDEJcEANCfzzWtcxB06Y7tkpaVzdt6dA5RWwOj2vmU8bpTB2AsOWc6dsvrPDq2WYVaZ/U8keO5iIiIiGrFoIsoCIwVxVBeejUAQHd0O0RBgBCfYpdO/ffPAABZWn0omtZ+Y2SzopdvRd7gbJTNf9ZpmprVXwAwjS+LaVf71PCOVGqsb2bMroVEREREtWPQRRQExooiqK7sblouPgvd/g1247l0x3ehZv0iAEBCv2EQBMHt4ysvuxYAoNn+K/QFJ+22aw9tRdXKjwEA8b0egexCq5snHE6iwZYuIiIiolox6CIKAmNFEZRNr5KCo6qf5qF80RTozx+HofQ8qld/iaLX7gD0WsjrNUN878ftjlH163zkj+6A/NEdoD3yt9W2+FufhBATD1FdiZI374Z6+y8wlOZDn3cMlf97G8Wv3QHoNJDXvxSJ977g1TXU6AycRIOIiIjIC/zGRBRgolYNUW2asTC+3zOo/OZNGPKOomr5TFQtn2mVVpZSD+kvLXfYEmUsL4Lh7GHTMTU1VtsUWZcgZcSHKH33cejPHETJ9Pvt9lc0uRxp45dAFp/s1XU4mkRDpeDvNkRERES1YdBFFGDGiiJpWRabiKSH34ThfC7UG76B/swhiDoN5FlNENuhLxLufA7ylCyvzhPX6Q4opv+JqhXvQLNnHYyleRBUcVBc0gZxXe5GfK9HIfhwU+RK25sis5WLiIiIyC381kTkI0W9pmjwTbnT7caKYqvn8pR6iOvQF4n9nvHoPEn3vYik+150mUbZuDVSh73v0XHdVWEzniuJ47mIiIiI3MK+QUQBZtnSBQCypPQQ5cR7oijatXQlsqWLiIiIyC0MuogCyWiwG3/l6P5c4a5aa4DBeg4NzlxIRERE5CYGXUQBJOq1Vs8FhVK6SXIksZ1EI5aTaBARERG5jd+aiAJI1OusngtJ6R7dfytc2N6fi5NoEBEREbmPQRdRINm0dEVi10LAvqWLXQuJiIiI3MegiyhAjOoqiEaj1TpZcuQFXaIosqWLiIiIyAcMuogCxHbWQkGpgiw2MUS58V6V1gCj7SQaDLqIiIiI3MagiyhARLup4iOvlQtwMImGUgalnB8dRERERO7iNyeiALG9KXIkdi0EYHd/Lt4UmYiIiMgzDLqIAsBYUwFRZzNdfGLk3RQZACo0vCkyERERkS8YdBEFgN14LlUsZLEJIcqN94yiiCpOokFERETkEwZdRAFgLLcZzxWhXQurNA4m0WD3QiIiIiKPMOgi8jNRFGGstBnPFaFdC22nio/jJBpEREREHuO3JyI/E2vKAb11sBKpLV12N0Vm10IiIiIij/EbFJGbjEaj9DAYDFbPjUYjRFGEKIrQF56CvqIGIkSUyMuhlBmhLC6DKJZKaQBAEAS3/8rlcgiCAJlMBplMZrVs+9y8nz9UqHVWzzlzIREREZHn+A2KIoYoitDr9dBqtVYPnU4nrdfr9dDpdHYP83q9Xg+DweDwr+0628DKHCx55bfN/iuIWuzcuRMymQwdO3aEXC63eygUCrvnCoUCSqVSWjanOVOmhSBXQCaXQyZXILYiGdUFcVAqldJDpVJZPZfL5UG7ViIiIqJIwKCLAspoNEKr1UKj0UCtVkOj0dg9bNdbBlS2z30KfAJBECATBAiCDBAEmBqZBFNrkyBAgLnFSrDaDojmfxeYn4gX/ojmtYAoQhSNEI2mv0ZRlLY7cvToUQBA27Zt/X65+91II5fLrQKymJgYadnZIyYmRvpbUlIiLZvXKxQKv7bgEREREQUTgy5ySRRFaLVaVFZWIi8vDzU1NdJDo9GgpqYGarXa6UOj0QQkUJLLFZArlaa/CgXkCotlF39lcjlkMvmFlhs55OZli3Uyi3Wi0YDV//saAHDL3Y9AqYqBIJNBJsgg+LkrnyfEC4GXUbzQrdF48e9fW3ZAhIibBwyBaDS31hlgNJj+itJzAwzmvwY9jAYDjAZTK9/Fv6ZtBr0eRqMBBr0OBr3puWn9hdZD/cVuiOZWQrVa7bfrlclkVoFZbGysFJRZPmzXm5+b/6pUKgZvREREFHQMuqKEKIrQaDSorq5GTU0NqqurrZYtgynLh1qthsFg8Pn8MpkMCqUKCqUKSpXq4rIyxrQsrVNCoTD/VUKhVEKuUEKhVEGuUJjWKZQQZMGZA0avuxhMqGLioFAqg3Le2phb0uQO5sKRK0xv66SUtKDlRxTFC0GZKSDT63UXAjQd9Dqd9Fxatlqnhd78V2fxV6+F4cKEJEajUaqTvjAHb5ZBmWVgZrvsaB0DNyIiIvIUg6466uDBg/jzzz9RVVUlfVk1Go1eH0+QyaBSxUAZEwulKgZKVQxUqlgoVCrpuelhCqQs1ylUKsjlrGp1mWmyD1NrImL8d1zRaLQLyHRWyxrotdoL283btNBrLy7rtBroddoLgaHvwZtl4GYbnJkDtLi4OLvtls+VYRK8ExERUXDwm3AdtWHDBpw4ccJuvVyugDImFiqbhzImFipVLJQxMVCqYqGy+SuP0jE1gkxA/SaXSssUXIJMJgXvvhBFEQa9XgrSdDot9FqN1V+d9mKQZg7UrJe1pjF1fgjc5HK5W61qjpbNfznOjYiIKHIw6KqjzK1ardtfj/qNm0vBlbnrGblHLleg0823hTob5CNBEExdVpVKIN67Y4iiCINBbwrCtBdb2XRS0Kax2nZx2Tp4A0zj3qqqqlBVVeX1NcnlcikAEwQBycnJbo1vs30weCMiIgo8j7+BT5gwAcuXL3d9UIUCcXFxqFevHtq1a4dbbrkFN998s8eZe/XVV7Fw4UIApu5y5Lnk1HSkpGeGOhtEEU8QBGlMoS+Bm15n2aJmG7yZ1lkFdbqLLW22gZt5bCYAFBUVeZUny+6SKpXKatIRy/WWy8628ZYBREREjgWk2UOv16OiogIVFRU4evQoli1bho4dO2LevHlITk526xibN2/GokWLApE9IqKQEARB6i4Zl5Dk1TGsA7eLQZnlODa9ZXdJqy6VF7brtdIkMf6apMRMoVDUensAy/u7mdfXtswWOSIiimQ+BV2jRo1CTk6O3XqDwYDKykocOHAA3377LWpqarB161YMHz4cX3zxRa3/cVZVVWHixInhd08mIqIQsw7cvD/OxeBNZz8RiU4HnU5jmmXSYjZJKa3l5CYXlsULXZrNNxo3t8D5iyAIVjfhtn3Y3qTbfMNvy2XLdUqlEoWFhdL4ONubg8uCNEMqERFFB5+Crg4dOqBTp04u09x///0YNGgQysvLsXXrVvzyyy/o16+fy31mzpyJ06dP+5I1Ir8wGPTYtWkNAKDd9TdxFkaqMyyDN38w389Nmu5fCsgu3BrgQnBmfRsBvbTN8p5v0rJOB6PRdMsK8z0DtVqtX/JbG7lc7jAYc+dh3tf2uVwut9pmud5yne2DASARUeQL+DfIli1b4qmnnsKMGTMAACtWrHAZdFl2K0xKSkJFRUWgs0jklGgUceroAQBA2049AA5XIXLIfMsAVUysX49rNBpNgZhBD626BmtWmMb5du55BwCYArQLAZ90826D3uqv6WbfehfbTDcEt+xdYb7Jt0aj8ev1eEMmkzkNyGp7mFvtLAM4Z4Gd7bKjdfn5+dLYPdvttvvJQngDeSKicBOUn+379OkjBV379u1zms6yW+Gdd96Jc+fOYevWrcHIIhERhSGZTAaZKgZKxEChUEnrM7Ib+f1m5UajUQrQjBeCLsugzbzOaDQHaoYL6/TSstFosEl7cb2U3m6d3nRuowFGgxGiaLTLV7Ba+PzNNhhztuxsu+XDnTTOtnma3puHIAhslSQip4ISdGVnZ0vLJSUlTtOZuxVmZWXhxRdfxMiRI4ORPSIiogtfnlVQKFW1Jw4g0WiEUTRaBGYXAjKLZVEK0gwXAzajxTab9aL010Fa87Lhwnnt0lvvZ87fxW0GGEURcDAO23ghjV6vD0FJBp858LIMwmQyGURRhFKpdBioyeVyq7SO9ne1ztF6d/Z1lFdnaVyldWebq79E0SIoQZdlF8GkJMczdll2K3z55ZeRmpoajKwREVGEiJablQsyGeSQRdwYUlEUrQI5++DMKN1g3OU2KY1old5yu9NtlutFy/PYpxeNopQHURSt8mT1/MKyaLnNyURfpvv5mVo5bflrhtC6RBAE6WEbjImiKM1aapvG2bLlOnee17bNVfraHo6uz3ads7TO1rt6DgCFhYWorq62OpftX0frPPnrzTpHz2vb153n7m6rbbsv98z0RFA+0Tdv3iwtX3HFFXbbLbsV9u3bF7fcckswskVUK7lCgX4PPiUtE1Ho8Gbl4c3UaqOIirGv5sDLMhAzGu0DNKOUxlFao8UxbLZbrpOOKVpvt1xvFxyK1uttjmc0GgHL44giYLGfeT2k4zo7nuU60S4v7pYjAIeBKlGwvPbaawE/R8C/RVZUVODdd9+Vnvfv398ujblbYVpaGiZPnhzoLEWVyvJSlBSeD3U2KIi0GjUA8HUnIgo3F1oa5NEy9ksUIUK0CNzMwZyzgM20DnbrHO0runhuhAhIQakpK+b9LyzDvGw+p8UyYH082JzLfG2i8UKvWsvtuHBsc0BpmbeL26zOcWHZlPZiucFqm00a2B7zwroLeaPwE5CgS6PR4Pz589iyZQvmz5+PU6dOAQBat26N22+/3SqtZbfCSZMmIT09PRBZilhlZWVYuXIl1Gq1R/vl5+cDAPZv/ysQ2Yo4O3bsAABcc801Ic5J4BWcM73f1v+0JMQ5ISIiIiLAx6Br6NChbqe95JJLMHfuXCgtZpuy7FbYs2dPh61g0W7Xrl3YvXt3qLMR8Y4fPw4gOoKu5s2bhzoLRBRAcrkciYmJITm3wWCAXB4F/Qf9jOXmOZaZ51hm3glW19aAdy9s0qQJBgwYgKFDhyI5Odlqm7lbYUpKCl555ZVAZyUiGS/0iW6cmYirmkVWK6DeYMTv/5hucn1L+8ZQyEPXnWLj6ngAQN9rLwlZHoIlEq4xnOoGUaQoLFNj2+F8ZGZmYsSIESHJw4kTJ9C0adOQnDuSsdw8xzLzHMvMOydOnAjKeXwKukaNGoWcnBy79TExMUhKSsIll1zitLugZbfCCRMmoF69er5kpc7LSIrBVc0yQp0Nj2j1BumLdZum6VApQvfrS0qCaQroSCvDuiqc6gZRpMg9X45th/NDnQ0iIvKCT0FXhw4d0KlTJ4/3s+xW2K1bNwwcONCXbBAREREREYWtkMyBvXfvXpw+bfqV+88//0SrVq1q3cecplGjRli9enVA80f+IRMEXNYwRVomMmPdIDIxiiK0OgO0eiM0OgO0OgM0OgM0F55bPkoqNKHOLhEReYk3HqKAUchlGHjDpaHOBoUh1g2KJKIoQm8QoTcYoTcYoZP+itDpDRZ/jdDrTdu1eiN0evNfU1Cllf4aodMZoNEboDd4PrVzqCbRICIi74Uk6GrZsiXmzZtXa7q3334bhw8fBgApfWxsbEDzRkREvjHfx8Z0k1XAKN1E1dSyYxTN2y6uE40X0oqitJ/pJq7W6YyiCKPxwrJRtPprMNovGyyem9fpjUbTX4P5uQiD0QiDwbTNYA6wjKa/BmPg73mjUCgQExODmJgYxMbGSsvm5+ZHXFycw7HUREQU3kISdKWnp6NXr161pvv888+lZXfSExHVVaJoCg70esvWFlOwoDNcDBjM6/QGEYYLgYM5sDBeWG8bhNgGJ5bBjOFCEGQwOgmEHARFdfm+nHK5HAqFAkqlEkqlEiqVym5ZpVI5fMTExNj9NT8UCnY8ISKqy/gpT0QUAHrDhTE6FmNztHoDtDqLbmYXtusM1su6C13TdAYDdHpRCrLqCplMBkEQIJPJpIc7z+VyudU62+fmdeaH7XNzwCSTyaBQKFBaWop69epBoVBYPeRyOZRKpRRc2W4jIiLyFIMuChi9wYiVO04BAHpf04T3YiJJJNQNURShMxih1hpQo9VDrTVArdWjRmuARmuAWndhnXmiA8tlnSGgXdJkMplVUCCKIuLi4qwCBHOAYf5rXnYUgNgGKbYBjLOgx9HDMlhythwueE8bIiIKFgZdFDBGUcTeE8UAgF7tG4c4NxROQlE3DEYjqjV6VKv1qNboUaO98FdjCqRqNHpUa83PTQGVPwInyy5k5odl1zLbrmfmh213Ncsgy7a1hcEDERFReGPQRUQRy2AUUaXWXXjoL/7V6FB94W/VhSBLozN4dQ65XI74+HjExsZKf+Pi4qwelhMdmB/moCqcWnaIiIgoNDwOuqZNm4Zp06YFIi92vvzyy6Cch4jCiyiKqNLoUVGtQ6Vai8oanemhtvxrapHyhEwmQ3x8POLj45GQkICEhARp2bze9qFUKiHwXmJERETkA7Z0EVFQiaIIjfZiq9Ouo4Wo0uhRXqNFRbUOFTWmIMvdnn0ymQyJiYlISkpCYmIiEhMTkZCQIC1bPo+NjWXLExEREQUdgy4KGKVchmfvuEpapuih1upRVqVFWbUWpZUalFVrUValRXm1aZ1Of3EmvjV7zjo8hiAISExMRHJyMpKSkuz+mh9xcXEMpIiIiCisMeiigBEEAbEqVrG6SBRFVKn1KKnUoLRKY/pbqUFplSnIUrsxfiohIQGpqalITk5GamoqUlJSkJycjJqaGrRq1QqJiYmcnpuIiIjqBH4jJiKntHoDiis0KKpQo7hCg5IKNYorTUGWZWuVIwkJCUhLS5MeqampSEtLQ0pKClJTU6FUKh3ud+LECaSkpATicoiIiIhCgkEXEaFao0NhuRpF5WoUVWhQVK5GcYUaFTU6p/sIgoDU1FSkp6cjIyMD6enpSEtLk/7GxMQE8QqIiIiIwheDLgoYoyjidEElAKBxViJknAEu5DQ6AwrL1Sgoq0FhuRqFF/5Wa5zPApiYmIjMzExkZmYiIyND+puWlgaFgh8hRERERLXhNyYKGL3BiMXrjwAARg9oC5WC43OCRRRFlFRqUFBWg/zSGhSUmQKtsmqt033iEpORkpaJFpc0QFZWlvSIi4sLYs6JiIiI6h4GXRHiREEldh4rRE6jVMTH8GWji/QGIwrL1ThfWo380hrpoTM4HnOVlJSE7OxsZGdnQ69KhhiXgoTkdMiVSjRMiUXL7MQgXwERERFR3cZv72GuYcOGEAQBpZUarNxxCr//cwrN6iWhVeM0tGyUgjjODhhVdAYjCkprcL60Gnkl1ThfWoPCshqH97RSKBSoV68e6tevLz2ys7MRHx8vpdlyrBhqiwkxkmJZn4iIiIj8jd+wwlxOTg6ee+457N27F3v27MG5c+dw/HwFjp+vwModQNN6SWjdJA2XNWQAVtfoDUYUlNUgr6RaehSWqyE6CLDi4uLQoEED6VG/fn1kZma6nHJdZzBaBVwAgy4iIiKiQOA3rAiQlpaGbt26oVu3bigsLMTevXuxd+9enD9/XgrAZAJwSVYSchqnomXDFCTEOp6OO5jkMhlubtdIWibnjKKIonI1zhWbgqtzJVUoKFXD6CDCSkhIQMOGDa0eKSkpEDycqKRCbT15hlwA4lUcd0dERETkbwy6IkxmZiZ69OiBHj16oKCgAPv27ZMCsNz8CuTmV+D3HafQKDMBLRum4rKGKUhLDM3U3XKZgGtb1gvJucOZKIoordKagqviKuQVm7oJOhqDFR8fj0aNGqFhw4bS3+TkZI8DLEcqbYKuxFiFX45LRERERNYYdEWwrKwsKQArKirCvn37sH//fpw5cwanC6twurAKa3afQWZyLC5rkIIWDZLRICOBU7cHWaVah7ziapwrqUZecRXOlVRDrTXYpYuJiZGCK/MjNTU1YIFQhc008YmcoIWIiIgoIPgtq47IyMjAjTfeiBtvvBGlpaX4999/ceDAAeTm5prux1SuxuaD5xGnkqN5/WQ0r5+MZvWSwqIbYl1So9XjfEm11E0wr6Ta4Q2G5XI56tevj8aNG0sBVmZmJmRB7IZp272Q47mIiIiIAoPfsuqg1NRUdO7cGZ07d0Z1dTUOHz6MgwcP4vDhw6hRq7H/ZAn2nywBAGSnxqFZdjKa1ktEo4xEKBUce+Uuc4B1vqQGeaXVOF9SjdIq+/tgCYKAzMxMKcBq3LgxsrOzQ3pjYa3eCI3NJBps6SIiIiIKDH7LquPi4+PRrl07tGvXDgaDASdPnsThw4dx+PBh5OXl4XxpDc6X1mDLwfOQywQ0TE9Ak6xENM5MQMOMBJ9uaKzTG/HJygMAgMd7Xx6xAZ0oiiiv1iG/rAb5F+6Fdb60BuVObjScnp4ujb9SKBRo3749YmJCM67OmUoNJ9EgIiIiChYGXVFELpejefPmaN68OXr37o2KigocOXIEx44dw7Fjx1BeXo5ThZU4VVgJABAEoF5qHBplJKJBWjzqp8UjPSnG7TFGIkQpMBHhYJ7zMKTW6lFYrkZBmRpF5TXIL6tBQZkaGp39GCzAFGA1bNgQDRo0kAKtuLg4afuJEyfCLuAC7LsWchINIiIiosBh0BXFkpKS0L59e7Rv3x6iKKK4uBjHjh1Dbm4uTp48idLSUpwvqcH5khppH5VChnqp8aiXGoeslFjUS4lDRnKsTy1iwWY0iiir1qK0UoPiSg2KK9QoKlejqEKNKptgxEwmCMhMikN2cgIatshBoxatUL9+fasAK5LYjedi10IiIiKigOE3LQJgGneUkZGBjIwMXHfddQCAsrIynDhxAmfOnMHRo0dRXFwMrU6H04WVOH2hNcwsMU6J9MQYpCfFIjVBheQEFeJD8EVeFEVo9UZUqnWoqtGhokaHsmotsho0gVqrx4e/7EN5tRZGFw1vKSkpqFevnumRmoSMqnPISIyDQm7qHhnT7iYIyvBrvfKEbffCJE6oQkRERBQwDLrIqZSUFLRt2xZt27bFiRMn0LhxY+Tn55vGgp0/j7y8POTl5aGqqgqVNTpU1uhwsqDS7jgyAfj41wNIiFUgIVaJWJUcMUo5YpVyqJRyKOUyKOQyyOUCFDIBgADLnm4GowijUYTBKEJvMEKrN0CrM/1V6wxQaw2o1uhRo9WjWq13eL+repfkAIA00YVCoUBaWhoyMzORmZmJrKws6W9sbKy0n/7cEejPlEvPhbjEiA+4NHqD/SQasZHTUklEREQUaRh0kdvkcjkaNGiABg0aWK2vqalBYWGh9CgpKUFZWRlKS0tRUVEBoyiiUq1DpVoHoMbxwf0sJiYGSUlJSE5ORmpqKlJSUpCamoq0tDSkp6cjKSnJrenZjRVFVs9lSRmBynLQVKqtx6cpZALilAy6iIiIiAKFQRf5LC4uDk2aNEGTJk3sthkMBlRVVaGiogKVlZWorKyEWq1GTU0N1Go11Go19Ho99Ho9dDodDAYDRNG6759cLpceCoUCMTExiImJgUqlQmxsLOLj46VHQkICkpKSoFKpfL4u0WiEsbLEap0sOfKDrgq19X3DEmM4iQYRERFRIDHoooCSy+VITk5GcnJyqLPiMbGqBLaDv2SJ6SHKjf9UaqxbunhTZCIiIqLAiswbJxEFgbGi2Oq5EJ8MQRH5E044aukiIiIiosBh0EXkhLHcZjxXHehaqNEboDVYt96xpYuIiIgosBh0ETkgGg0wVpdarasLXQtt78+lkAmIVfJjgIiIiCiQ+G2LyAFjpc14LgGQJdW9oCsxlpNoEBEREQUagy4iB0Sb8VyyhFQI8sjvhmd3U2SO5yIiIiIKOAZdRA7Y3p9LqANdCwH7li6O5yIiIiIKPAZdRDZEgx7GqlKrdXVhEg21zgCdzSQaiQy6iIiIiAKO37jqqL9zTTf1vbZZmtWyp/ubNYmLdXmO2rb9nVuCg3kVaFU/yWqdN3n05no8OZ6xsgSwiE3+Uafi6L4qtG4gd3k9tr7efBJGERjS1Lt82K43sy0/ADiYVwEATssXMLVy5ZWp0Sg1FsVVWqQnqhCnlNeap9qu093rqG2bN7zJXyTyR7n5u+zD7dyBOEcoyyyYAvk+cqcMo6WcvcGyCQ1fv0MROcKgq46av/44ANOHhOWyp/ub/ff+Ni7PUdu2+euP2/2n7m0evbkeT45n27Xwk8MKHCrItct7bV9SzMcdcr17UZez67J9LWzLD3AcdNker1Kjx6/7zkMhEyCKIgRBQNfLMmvNk6dfxjypF77yJn+RyB/l5u+yD7dzB+IcoSyzYArk+8idMoyWcvYGyyY0fP0OReQIg6466O/cEmw/UQrA1NpiXv47t8TtliTzPmZ7zlXj2mYxDtPYHtd2GwDp+fYTpXbrPMmjq/N6w9HxLO/P9Y86FTtOVzjNu3mdbT6+3nwSlRqDtPxQ50s8zoftesu0lnmw5CiP5uPtOV2OXafK7I7lrAwtz+3sOt29jtq2ecOb/EUif5Sbv8s+3M4diHOEssyCKZDvI3fKMFrK2Rssm9Dw9TsUkTMc01UHWbaAOFt2d39p3brjTtPYprfd5mi7t3n05no8OZ4oGiFWl0vrPjkst0vv6nq9zacn5eEoD662m5e/33nW5Xlr2+bLddS2zRve5C8S+aPc/F324XbuQJwjlGUWTIF8H7lThtFSzt5g2YSGr9+hiJxh0FXH2LaMmFtbAOtWEHf3l/Y9eXFf2zSWx3W0zfZ4tuvczaOr83rDcV7LIEs2dbkztXJVusy7o3xYtnIBpuv7evNJj/Lxd26J89fCQR5cbd9+ohRfbz6JnTatXI7y7ixPrtLWdh21bfOGN/mLRP4oN3+XfbidOxDnCGWZBVMg30fulGG0lLM3WDah4et3KCJXGHTVMZ60FHm63bzNVctOIFowXK335XzOWvQUjVoCsG/lcvdYnubTm9YsT3lzfn8eJxivna/HDEf+KDd/l324nTsQ5whlmQVTIN9H7pRhtJSzN1g2oeHrdygiVxh01SHOWkYsedKS5Ghfy/7N7m7zlDu/iLpK6w5XLXo7CkTsFBvZtXK5Ys6HbSuXmbPWLm9bszzlKE+W57IsQ1f1wJOWMXN6V3XGn6+dL8cMR/6o8/5+33giGOcOxDlCWWbBFMj3kTtlGC3l7A2WTWj4+h2KqDYMuuoQX8Zsubu/py0d3vJknIE353V5vHXH8Ml+nVfH9FeLUrB5O67OnX2C+dp5e8xw5I9yC9b7NVTnDsQ5QllmwRTI95G3PSb8ce66gGUTGv74XCVyhUFXHeHOLzRmnrQk2XLVYuJqm6fc+UXUUVp31Hq8k2XYftJ+/FNttp8orbV8LFu7PHnNAq22cWSO0prVto8nrWy1iZZfIv1R5/39vvFEMM4diHOEssyCKZDvI196TPh67rogWupguPH1OxSROxh01RG+jvMIx19uPBkn5kn+Q3mt4TwTkifjyPx5HYF47cKtbD3ljzrv7/eNJ4Jx7kCcI5RlFkyBfB/5qwzrQjl7g2UTGoH4rCCyxaCrDvCmxcSTlqRQ8WScmLu/PIX6Ws2tXaHOhyOejCPzpGXM3WPVJlp+iXT3On0Zn+nOMbwVjHMH4hyhLLNgCuT7yB89Jrw9d10QLXUw3Pj6HYrIXQy66gBfxy6E8y82/m4FCYdr9ffMhKHiz+sIVatFOPJHnQ9ly28wzh3q1tFIrVtAYN9HoaxLdUG01MFwE4zPISKAQVfE86WlwZ8zDgaKJ+PEavvlKVxalyo1hrDIh6/8OcNiIF67SPwl0tPrNF9jbGysz8fwh2CcOxDnCGWZBVMg30eB+HyN1HL2RrTUwXDj63covgbkCQZdES6Y42kiAWd9ilzBnokuHHk7hiY1NdXnY/hDMM4dLjNeRlrdAgL7Pgrl+MC6IFrqYLjhdygKJkEURTHUmQimPXv2QK/Xo3379qHOChE5cOLECTRt2jTU2Ygoer0eCoUi1NmIOKxrnmOZeYfl5jmWmedYZt4JVrnxf+kwp9n3JwAgpk03n/ezXKfZ9ydqNn8PRf1Lkdh/uN2+5Ytfgyw5AxnjFgEASkpKkJaW5lHezNsBQJe7G8pmbWu9DttjVv40D/q8Y1DUv1Tav/KneQCAxP7DnV5nbeez3E/U1kB7ZAegUEGIiYcgV0CedQlkMXFuXacl27xpju1CzKXtPH793OEoX7bnF40i1C06o0ZrgNZghCAC2ed32O0XLLXVS38ds7Ztnp7T1zwGWl5eHho3bgzAf58Z4XjNnuQpkPkP1bHD8TUhIiL3MOgKc5VLpwLw/D9ZR/tZrqtcOhXafzdBiE2wC7oql06F/uBmq3Xl5eV2QVdteTNvB9wPumyPWbl0KkR1FYTYhItB14U0if2HO73O2s5nuZ+xvAhidTkAQAQgKJRQNMpx+zodHdecN13ubujcuG5vuHqNLctGOeF/OJRfBQDISFAhwcs65Q+11Ut/HbO2bZ6e09c8BprBcHHso78+M8Lxmr15LwbrvReMY4fja0IUKGyx8RzLLLwx6Apjmn1/Qrt/g7Tsya/ytvtZrqv8aZ60LFaXo/KneVLgZZkOAIpmPCi1dnmSN9vjAIB2/waX12F7TF3u7ovBUHU5tPs3oPSTMdK60k/GuLxOZ+ezPY88o5HVdiEpHYIguHWdlip/mucwb7Vdtzcc5cvZ+ROPbkZaSluUVOuQcWabV3UqUHn2to67OmZt2zw9p695DCZ/fWYACLtr9uTaAvmaherYkVQPiYjIHifSCGOWLUWWy97s5+pYrrZpt/3kVd6c5dfVddge01Hamt8XOFx2dJ3u5K1y6VTIkjOttsuSMpymdcVyu2Xe3NnXU7W9xpbnr1g6Fc0y4wEAyp9mBixPtfG0Xnp7zNq2eXpOX/MYTP76zAjHa/b2vRiM914wjh2OrwkREbmPQVeYsm0pMreWeLOfZcsWAKk1xPJ55U/zHLZOAabWrqysLLfz5uw4rq7D0TFt8wkAMBocLju6Tnfypt2/Abrju6wCL1lyhlvXacmylckun7Xs6ylH+bJsAbQ9v27/BsQc3YymBf9Ad+CvgOTJmzzbvl7m/FhOf+7pMc3X42ybp+8rb9+HoeDPz4xwu2ZPri2Qr1mojh1J9ZCIiBxj0BWmPGk98td+ztJpt/2EuLg4t8/hya/QnuSxNu60rjlKU7H0TSgatQQACEoVZLGJHufT29fGG7W1ADpSsXQqElfOCVieauNuWVYunWo1/bm3x3S2zdO6F6i6Ggj+/MzwJk0g+fpeDOR7LxjHjqR6SEREjjHoCkPOWoo8/VXezGGLkYM0zlqnAKB45iC38uaqlcs2bW359pSz66wtb7r9f0GXuwey5Eypa6Enr4FdK5cT/vh12mlZ2bSs2dLt3wDdAc/rlD94Ui+1+zcAR7Z6fUxnLZ7mbZ68r7x9H4ZCzKldfv3M8PQ4geTJ6xDI1yxUx3ZVn8OtHhIRkXMMusKQp7+8u7PNV+axXbWd3908BHt8Qm15M7d2mbsWelLO3o6d8UYgyirQ5e/p8St8bHnxdWxYoM4RSPKV7znd5q/rCNU1++u9GMj3XqiOHW71kIiInGPQFWZq++XZ01/l/algUu9a8+ZuHjxpGfOH2vJmbu0SkrM8eg3cbeVytK+nAlVWgfzF3Js863xsnfHk9TDztPU1nFoZNPv+hP7fjU63++szIxTX7MnrEMjXLJTHdlWfw6keEhGRawy6wkwgx235yvbeXb7ypGUsGCqWvgmZKtaj18BfrSqB3C+Uxw7EtQYjr5EwzsksmJ8ZoZzxsrY0gXzNQn3sQO5PRETBwaArjHg7viJYrUX+5knLWDDo9v/ldPyELYczBrrJm1+nA/0aB+IXc1/yHIoWXU9bX8OhlSHYnxmhnPHSGVfjnhylDdR7L1Tv63Coh0REVDsGXWHEl/vqkH94Upa1zRjor/N4k94bgbznkL/2D8b4s0gY5+TN+f31mRHKGS+DnTacjh2M4xARUeAw6AoT3o6viNRWrnDlUctVLTMGuuLJr9PBHPcWqHsO+SM/wSgHT1tfw3FWP2f89ZkRyhkvnQnUuMpAjnvzZ31maxcRUfhThDoDZBIJ4yvIvyqXTkVMm25upQsWd/PkznH8wTI/4Vrf/VVm3pw3GPs4O04grzkYLZqBeu+F6n0dqnpI5A2j0QhRFEOdjaASBAEyGds6opkgRlmt37NnD/R6Pdq3bx/qrBARUQidOHECTZs2DXU2IgrLzDsst4uMRiNmzpyFigrPx0RHsqSkZIwdOybogVd5eTl++uknrF27FkeOHEFRUREEQUBGRgaaNWuG7t2749Zbb0VWVpbbx6yoqMDPP/+MtWvX4uDBgygqKoLBYEBaWhpycnLQrVs33HbbbcjMzKz1WEOGDMHWrab7c/7xxx9o3Lix19fqrWC9P9nSRW6prq5GfHx8qLPhF6IoQnt0B6BVS+sUjXIgT3H/A8dd5eXlSE5O9vtxXalU61Bao5Oey2UCGqTEBTUPvjh37hwaNGgQ6mxEFJYZEUUKURRRUVGO/g89EzUtP0ajET99/X7QW/d++OEHvPnmmyguLrbbVl1djVOnTuHPP//E7NmzMXz4cDz++OMQBMHp8YxGIxYsWIAPPvgAFRUVdtvz8/ORn5+PDRs2YM6cOXj44YcxYsQIqFQqv15XpGLQRW4pKCioM7/SidVlEEvzrdbJ4pICcq6SkpKgB13nyjQ4W3YxoMxKVEVU0KXVakOdhYjDMiOiSCOTySCTyUOdjTrrxx9/xLhx4yCKIlQqFXr27In27dtLLVr5+fnYvHkz1q9fj+rqasyYMQMlJSUYN26cw+NVVVVhxIgR2Ljx4n0h27Zti+7du6NRo0aIiYnB6dOnsWHDBmzduhVqtRrz58/Hhg0b8PHHHyM9PT0o1x3OGHRR1DFWFFk9F2ITIKhiQ5Qb/xJFEWfL1Ciu0kKlkCFBJUdiLN/mRERE0aK0tBSTJk2CKIqoX78+FixYgBYtWtile+SRR7Bz50489dRTKC0txccff4xu3bqhc+fOVukMBoNVwNWkSRO89tpruP766+2O+eSTT2L37t0YM2YMTp48iX379uHhhx/GN998g7i4yPkBOBCio12XyIKx3DrokiVnhCgn/lOh1mHhllO4c95mPLdkN1778V9M+n4/3vjpIFbuO48Kta72gxAREVHE++6771BdXQ0AmDJlisOAy+zqq6/G66+/Lj3/9NNP7dJ88sknUsCVk5ODJUuWOAy4zNq2bYvFixfjiiuuAAAcOnQIU6eG52RYwcSgi6KKaDTCWFlitU6WFNlB18ajRbj17Y2YtfIwzpTUWG0rqtLi/bXHcevbG7HxaJGTIxAREVFdsX//fmm5U6dOtabv1asXsrOzAQA7d+602lZcXIz3338fABAbG4u3334bGRm1f2/KyMjA22+/LY3nWrp0KQ4fPuzuJdRJDLooqojVpYDRaLVOlhS5/Yw3Hi3Cs4t2Qa0zQATgaIiuCECtM+DZRbsYeBEREdVxlpNcnDlzptb0giDg/vvvx4ABAzBw4EAYDBfvQ2rZanbPPfe4bDWz1bRpUzz66KMATJNwfPnll27vWxcx6KKoYtu1UIhPgqCIzFl1KtQ6jFu6F6LoONiyJAIQRWDc0r3sakhERFSHNWnSRFqeNm2aW5MtDR8+HNOnT8f48eMhl1+c4OTnn3+Wlu+77z6P83LnnXdKy7/99huMNj98RxOOsKeoYqywnjY1HLoWllR5N/PcdzvOSC1c7jC3eH2z7QwGXtPQq3MGWlpCZAbARERE4eK2227DV199BQBYv349+vbtiwceeAB9+vTBJZdc4vZxysvLceDAAQBAcnIycnJyPM5LixYt0KxZM+Tm5qK0tBQHDx7E5Zdf7vFx6gIGXRQ1RKMBxirb8Vyh71rYc/aGoJ1LBDBv7THMW3ssaOf0xI5JN4c6C0RERBGtffv2uO+++/DNN98AAE6fPo2ZM2di5syZaNy4MTp16oTrrrsON9xwgzSWy5EjR45I9xZr1aqVy3t4uZKTk4Pc3FzpmAy6iOo4sbIUMFq0CwmALDH0QRcRERGRP02ePBkxMTH46quvrG7KfPr0aZw+fRrfffcdAKB169bo378/Bg0ahMTERKtjFBQUSMspKSle58XyHl2FhYVeHyfScUwXRQ3b+3PJ4lMgKJQhyg0RERFRYCiVSrz00ktYvHgx+vfv7/QeWf/++y9mzZqFW2+9FStXrrTaVlNzcUbk+Ph4r/OSmpoqLavVaq+PE+nY0kVRw3Y8lxAGXQuJiIiIAuXqq6/G1VdfDbVajS1btmDz5s3YsmULDhw4YDWpRUFBAZ599lnMnDkTt912GwDTFPFm5hkMvVFZWSktJyUleX2cSMegi6KCaNA7GM8V+kk0AOCP57t6vI8oihi6YDvOlandnkgDAAQADVJi8cVjHbzum01ERESRJTY2Ft27d0f37t0BAKWlpfjrr7+wYsUKrF27FoDpu8XEiRNxww03ID09HZmZmdL+paWlXp/bcl/LrobRhkEXRQVjVan1vOoCIEtMC1V2rHg7Y9+gTk0wa6XnNxp8qHMTpCfGeHVOIiIiinypqano378/+vfvj7///hvDhw9HaWkp1Go1vv32Wzz55JO47LLLIAgCRFHEwYMHIYqiVz/YmmdABIBmzZr58SoiC8d0UZ0niiL0Z4/CUJYPY3U5RFGELCENgjyyf3O4vV19xCrlcPfjTyYAsUo5bmtbP6D5IiIiotDYvXs3PvnkE0yfPh2nTp1ya59rr70WL7/8svT84MGDAEyBWatWrQCYbri8f/9+j/NTVFSEY8eOSceL1pkLAbZ0UR1mrCpFzbqFqPplPgznj0vrZanZiOsxGIomrSFLSA1dBn2UFKvEjHuvxLOLdgG13CDZHJjNvPdKJMVy8hAiIqK6aP369Zg7dy4A4LLLLrO6UbIrN954o7RsOQarb9+++PfffwEAS5YswZQpUxzuX1RUhNLSUrRo0cJq/aJFi6TZE3v06BHVQxvY0kV1kmbnKuQ/fTnKP38Bhvxcq23G0nxUfT8b+U9fDs3OVaHJoJ/c0CID7z7YTmrxsv0oM6+LVcox98F2uL5FeIxjIyIiIv9r3bq1tPzzzz+7vV9VVZW0bBmo3XfffdLMhcuXL5dawWytWLEC/fr1w+DBg7FlyxYAwPnz5/Hll18CAGQyGZ544gn3L6QOYtBFdY5m5yoUT7sXorYGEEXTw4oIQISorUHxtHvrROD16+gbMLZPSzRKs54StlFaHMb2aYlfR3dhwEVERFTHde3aVbqn1oYNG7BqlXvfcZYuXSot33zzzdJyeno6hg0bBgDQarUYPXo08vPz7fbfvn07AGDbtm14+OGHMXXqVIwcOVKaRGPAgAG47LLLvLqmuoJBF9UpxqpSlMwaciHYMrpOLBoBUUTJrCGmiTYiWFKsEg92bIL/De+M1WO64ceR12P1mG743/DOeLBjEyTFsicxERFRXRcbG4tRo0ZJz59//nl89tlnMBgMDtOLoojFixfjvffeAwC0bdsWN9xwg1Waxx9/XOp+eOzYMTz44IPYtGmTVZp3330Xn3zyCa688kqIoojPPvsMu3btAgBcccUVmDx5st+uMVLxmxiFPUO5+3cvr171KURNNVyPcLIgGiFqqlH128eI7/WIV/lzqbIYhvIE/x/3AnlyptVzQRCQGq9EajzHbREREUWjhx56CMePH8eXX34JjUaDqVOn4pNPPkHPnj3RsmVLpKeno6qqCsePH8eaNWtw9OhRAEBWVhbeffddu+PJZDLMnTsXo0aNwtq1a3H69Gk88sgjaNu2Lbp3747GjRtDLpfj3LlziImxnx35qquugkxWezvPxIkTre4N5kxGRgbefPNNN0oivDDoorCX/3+XBvgMIioXT0HlYseDQ32hAmDfCO8/Db4pD+DRiYiIAsPyxrx1XSiudeLEiWjRogXmzJmDsrIy5OfnY9GiRU7Td+nSBVOmTEGDBg0cbo+NjcUHH3yAL774AvPmzUNZWRl2796N3bt3O0wvl8uRnZ2Ns2fPYsmSJdiyZQvGjRuHnj17Op1MY/PmzW5dW6NGjdxKF24YdBERERFRUAiCgKSkZPz09fuhzkpQJSUlB3XmPkEQ8OCDD6Jfv35Ys2YN1q9fj0OHDqGkpARlZWVISEhAdnY2rrnmGvTt2xedOnVy65gPP/wwBg4ciJ9++glr167FgQMHUFxcDFEUkZqaissuuwydO3fGbbfdhuzsbHzwwQeYP38+cnNzMXz4cNx000344IMPglAC4YdBFxEREREFhUwmw9ixY6RpxKOFIAhudbHzt5SUFAwYMAADBgzw2zGTkpLwwAMP4IEHHqg17ciRI9G7d2/MmzcPv//+Ozp37my13Ty7YTRg0EVEREREQROK4INCp1WrVnj33Xdx4sQJZGRE70zKDLoo7NX7+Jhb6URRRNGLN8FYeMrBNPEuCAJkmU2Q8eYavzf9nz51Co3dvDEhERERUV3VtGnTUGchpBh0UdiznaHPlcT+w1D++QsenyOx/3AoUrI83q/2A1d5lH8iIiIiqnvYvkt1Slz3QRBU8YDgZtUWZBBU8Yjr/mBgM0ZEREREUYtBF9UpsoRUpI35EhCE2gMvQQYIAtLGfgVZQmpQ8kdERERE0YfdC6nOibm6F9InLEXJrCEQtdWmlZZjvC6M2xJUcUgb+xVi2vUMQS7tzfjtEBZtPY1X7rgcd7RzfJ8Ms5IqLRZvO40NR4pwqrgaap0RaQlKXNkwBQOvaYgul3k/UFUURfy0Ow/f7zyHQ+croDOIqJccg66XZWBI50tQP6X2GxcSERER0UUMuqhOirm6F+p9cAA16xah6pcPYDh/XNomr9cMCX2fRlyPQZDFp4QwlxetPViAJdtOu5V2x8lSjF26B6XVOqv1BRVarDlYgDUHC3B72/qYfPvlkMs8mxjEKIqYuHwffttnfUvnU8U1WLT1NH7YlYdZ912F65qleXRcIiIiomjGoIvqLFlCKhL6PYP4vk9DrCyGsaYSsrhECInpQb1BYW3WHSrE+O/2wujGhIvny9UYvXg3KjV6pMQpMKzHpehyWQZUChmO5lfhwz+P45+TZfhhdx4yElV4tudlHuVl3upjUsA1uHMT3H1NIyTFKrD9RCnm/H4YeeUajF26B9881RHZyWzxIiIizxmNRt6ni6IOgy6q8wRBgJCUAVlSeN0bwiiKmL/uOD7ZkOtWwAUACzacQKVGjxiFDB8OuQYtsxOlbZmJMejYPA1jl+7FmoMF+GrzKTzYsQmykmLcOnZ+uQZfbTkJAHjkhkusArZbrqiHNg2TMPjjv1Fao8OH63Mx6bbW7l8sERERTAHXrJkzUF5RGeqsBFVyUiLGjB0XlMBrwoQJWL58udPtcrkcMTExSEtLQ4sWLXDLLbegb9++SEpKcph+9uzZmD9/PgCgSZMmWLFiBeLj42vNR1lZGe68806cO3cOcrkcX3/9Ndq3b+/dRdUBDLqIQmDj0SK8veoIjuRXAQAub5CEA+cqat1v1QFTK1TvNvWsAi4zQRAw/KZLseZgAfRGEZuPFeP2WsaHmS3Zdho6g4h4lRyPd21mt71hahwGdWqC99Yewy978zC2T0vEKeVuHZuIiAgwjRsur6jEcwPaQeZhF/hIZTSKmPP9rrBp3TMYDKiurkZ1dTXOnDmD9evXY9asWXj55ZfRr18/u/QjRozAmjVrcOjQIZw6dQozZ87E5MmTaz3PpEmTcO7cOQDAsGHDojrgAhh0EYXEiIW7AAAKmYDHuzZDv6uycee8zS73MY/hEgBc1cj5WLQm6XHSckGFxu08bThSBAC4tmkaEmIcfzT0aJWJ99Yeg1pnxJZjxejRKgD3NiMiojpPJhM8HndMnhs1ahRycnKs1ul0OlRWVuL06dNYt24dDhw4gNLSUjz33HMoLy/HAw88YJVepVJh+vTpuO+++6DT6bBw4UL07t0bnTt3dnrepUuX4rfffgMAtG/fHs8884z/Ly7CMOgiCgEBwE2tszD8pkvRPDMBZ0trat0nNV6JP8Z0g85ghNHFr2Wnii8eKzlO6VZ+dAYjjhdebHVzpkVWApRyATqDiAPnKhh0ERERhbEOHTqgU6dOTrc/99xzWLx4MV577TXo9Xq89tpraNq0Ka6//nqrdFdccQWefvppzJ07F6Io4sUXX8QPP/yAhIQEu2MeP34cb775JgAgISEBM2bMgFzOnjEc0UcUAsuGdcbMe69C80z7D6vaKOUyxCicf3h9u/2MtHx1E/dmZ8wv10B/YWBZo1TnE2QIgiBNoHHGjUCRiIiIwtsDDzyACRMmAAD0ej3efPNNh10hn376abRp0wYAcObMGcyYMcMujVarxZgxY1Bdbbplz0svvYQmTZoEMPeRg0EXUQg0zah9AKo3dp0uw9K/TUHXNZek4rJ69uO+HCmtuTj9fFItrWOJF7oelqv1XuaSiIiIwsngwYNx9dVXAwAOHTqEP/74wy6NQqHAtGnToFSavicsXrwYmzZtskozZ84c7Nu3DwDQp08fDBw4MLAZjyAMuojqiNzCKoz5ZjcMoogYhQz/uTWn9p0u0OiN0nKswvXHQqzStF1rsQ8RERFFLkEQ8Oijj0rPf/jhB4fpcnJyMHLkSACmSVEmTpwotWpt2rQJn376KQAgOzsbU6ZMCXCuIwuDLqI64FhBFZ766h8UV5larF7s1wo5DmY3dEYeRvctIyIiouDr2rUrFApTb5Zt27Y5Tfd///d/aNu2LQBTN8PZs2ejsrISEyZMgCiKEAQB06dPR2pqajCyHTEYdBFFuH9OluLxz7ejoEILABjbu6Xb08SbxSkvfhRoamnBUutM22NqaREjIiKiyJGYmIimTZsCAIqKinDixAmH6eRyOaZNm4aYGNN9QBcuXIixY8ciLy8PAPDoo4/aTcRBDLoowmj2rEXxjEE4/2RLnHswA3mPNUXRlNtRvW4hRKP/urtV/vhfnLsvGRXfvOm3YwbCz3vy8MxXO1FWo4dMAF7q3wqDOnk+YDUx9uJEppUa12O1KjWm1rTUePdmRiQiIqLI0KhRI2m5oKDAaboWLVrgueeeA2C679eaNWsAAK1bt5bWkzVOGU8Ro/zLl1D1w7tW68TKEmj3roN27zrUbPgG6eMWQVDFOTmCe7SHt6Fyyes+HSMYPv7zON5bexyAaZzV1LvaoLuXU7jXT4lFrEIGtd6Ic2Vqp+lEUcT5ctO9v+onO5/lkIiIiCKP5RTwZWVlLtM+/PDD+P3337F9+3YApok2Zs2aBZVKFdA8Riq2dFFEqF79hRRwKVteh/RJK1DvwyPInLYOsV3vBQBod61G2SdjfTqP9sjfKH7zboiaap/zHEhTfzkoBVwZCSp8OOQarwMuAJAJAppnmT5oD+ZVOE13JL8KOoNpGtnW9Z3fz4uIiIgij3lMFwBUVVXVml4muxhK6PV6HD58OCD5qgsYdFFEqFw+CwCgaHIFMl7+CTFX9YA8tR6Ul7ZH2rOfSIFXzdqvYCg+69U5qlZ+jKKX+0KsKvVXtgPi7VVHpGnhL0mPw2ePdsCVjZJ9Pm7XyzIAAFuPl6BGa3CYZt2hQgCASi7Dtc1SfT4nERERhY/Kykpp2dGNjy19/PHHdhNuvPLKKygsLAxI3iIdgy4Ke8bKYhjOm1p14m58AILKvltbQu//My2IInRHtnt0fO2Rv1H08q0o//h5QKeB8tL2Puc5UFb/W4AvNp0EADRNj8cnD3dAozTfulOa9buqPuSCgHK1HvPXH7fbfq5MjYVbTgEA7ry6AZJiOaaLiIioLrFs3UpKct6jZf/+/Xj3XVMPpIyMDHTq1AkAUFpaikmTJgU2kxGKQReFP+FiNRUNOsdp5BbDE2Vyjw5fOucRaA9sBAQB8X2eQMaUX73JZcBp9Ua89eshAKaWpikDLkecSoZqrd7pQ2ewnlxk8bbTGPjeZgx8bzP2nim32tY0Ix73XWcaQPvFppN446d/caygCiVVWvxxIB+Pf74dpTU6pMQp8EiXpsG5aCIiIgqaY8eOScvNmjVzmEatVmPs2LHQ6UzfycaNG4c33ngDsbGmH8VXr16NZcuWBTyvkYYTaVDYkyWkQt6gBQznjkL917dIvP1ZCMoYqzQ1a74yLShUULa4xsMzCFC1uRFJD70C1WXX+ifTAfD7/nzkV5gmsdAajHh4Qe0tek/e2AxPd79Uel5arUVukWm8mlpn34Xw2Z4tcKqkBhsOF+G7HWfx3Q7rrppxSjnefqAdGqRwEg0iIqK6JC8vT+oamJWVhXr16jlMN336dBw9ehQAcOONN+Kuu+4CADz33HOYOnUqAODNN9/E9ddfjwYNPLuFTV3Gli6KCEmDXgUEGfSnDqD49Tuh2bsehrIC6HL3oHT+s6heZboDeuLAcZCn1ffo2OkTlyHj5R/DOuACgD1nXM8i5A8xCjneub8tXr3jcnRomoqkWAUUMgENUmIx8JqGWPJUR7RrnBLwfBAREVFw/fHHH9Jyt27dHKZZt24dFi5cCABISUnB669fnO156NCh6NChAwCgoqICEydOhCiKAcxxZGFLF0WEuE53QBi3EOVfToT2wEYUT7nNarssozGSHpyE+Bsf9PjYioYt/ZVNrzVMjcOOSTe7TDOhbytM6NvKp/M83f1Sq5YvRwRBwO3tGnh8g2UiIiKKTEajEYsXL5ae33HHHXZpiouLMXHiROn5pEmTkJ2dLT2XyWSYOnUq7rzzTtTU1OCvv/7CwoUL8dBDDwU28xGCLV0UMYzV5RBiEx1vKy+A7t8tMJRzxhwiIiIiT3z22Wc4dMg0brxdu3a4/vrr7dK89NJL0g2T+/Tpg9tvv90uTdOmTa1ujjxz5kycPHkyQLmOLAy6KCKULRiHsv8+Cf3xXYjv/Tiy5vyN+gsLUe+Dg0h+9C0IihhUr1qA4pf7wlCaH+rsEhEREYU9g8GAL774ArNmmW7No1QqrVqzzJYsWSJ1P8zMzMQrr7zi9JhDhw7FtdeahmxUV1djwoQJMBqNTtNHC3YvpLCn2b0G1b/OB2Aa25U44OIvKPL0Bkjo+zRUl3dB0aTe0J85iIpFryL1mXmhyi4RERFRWNi+fTsqKiqs1mk0GpSXl+PIkSNYv3691BIlk8nw8ssvo127dlbpc3NzMW3aNOn5q6++ivT0dKfnFAQBU6dOxR133IGamhps374dn376KR5//HE/XlnkYdBFYa/6j88BmMZtJdzxrMM0ymZXIf6Wx1D141zUrF+ElMdmQIiJD2Y2iYiIyE1GY/RMsBDKa33nnXfcSle/fn1MmjQJvXr1slqv1+sxbtw4VFebZj4eMGCAXRpHLrnkEowZM0aaaOOdd95B9+7dcdlll3l4BXUHgy4Ke/pzRwAAqpbXQXBxDy5Vm66o+nEuYNBDf/44lJe0CVYWiYiIyA2CICA5KRFzvt8V6qwEVXJSIgRBCHU2AAByuRwJCQnIzs5G69atcdNNN6Fnz57SfbYszZs3D7t37wZgCsxeeuklt88zePBgrFy5Elu3boVGo8H48eOxZMkSKBTRGX5E51VTZNFrAQCiXuP2LqLO/bREREQUHDKZDGPGjou6qcQFQYBMFpypFKZNm2bVHdAXo0aNwqhRo7zaVxAEfPnll37JR13AoIvCnrxhS+hP/wvtv5sg6jR2N0Y20x7YdGEHBRT1XU+LTkRERKERrOCDKJyw1lPYi7vhbgCAWFmCikVTHKbRnf4X1Ss/BgDEtO8NWUJqsLJHREREROQSW7oo7MVefxdUqz6Fdu86VP04F4b8E4jv9wyUjVvDqK6EZvsvqPjmTYjqSgjxKUge/JrV/vmjTXdHV13WAakjPgzFJRARERFRFGPQRWFPEASkjfkSJbMfhnbPGqi3roB66wq7dLKUekgb8xUUDVtarTecPWz6m5pttw8RERERUaAx6KKIIEtIRfrE5VBvXYGadYugO7oDxopiCDHxUDRogZhr+yGhzxOQJaaFOqtERERERFYYdFHEEGQyxHUegLjOAzzar8E35R6fy5t9iIiIiIgc4UQaREREREREAcSgi4iIiIiIKIAYdBEREREREQUQgy4iIiIiIqIAYtBFREREREQUQAy6iIiIiIiIAohBFxERERERUQAx6CIiIiIiIgogBl1EREREREQBxKCLiIiIiIgogBh0EREREVFYOXHiRKizEHFYZuGNQRcREREREVEAMegiIiIiIiIKIAZdREREREREAcSgi4iIiIiIKIAEURTFUGcimHbs2AFRFKFSqUKdlYii1+uhUChCnY2Iw3LzHMvMcywz77DcPMcy8w7LzXMsM8+xzLyj1+sRHx+PVq1aBfQ8UffKCIIQ6ixEJL6JvcNy8xzLzHMsM++w3DzHMvMOy81zLDPPscy8E6xyi7qWLiIiIiIiomDimC4iIiIiIqIAYtBFREREREQUQAy6iIiIiIiIAohBFxERERERUQAx6CIiIiIiIgogBl1EREREREQBxKCLiIiIiIgogBh0ERERERERBRCDLiIiIiIiogBi0EVERERERBRADLqIiIiIiIgCiEEXERERERFRAClCnYFAE0URP/74I7777jscOHAA1dXVyMrKwnXXXYeHHnoIbdu2DXUWw87WrVsxZMgQt9J27doVn3zySYBzFL5effVVLFy4ECNGjMDIkSNrTb9u3TosWrQIu3fvRnl5OTIyMtC2bVs8+OCDuOGGG4KQ49Bzt8zOnDmDm2++2a1jNm/eHL/++qu/shhyZWVlWLx4MdasWYPjx4+jqqoKSUlJaNWqFW699VYMHDgQKpXK6f7R+rnnS7lF6+deYWEhvvzyS6xbtw4nTpwAANSvXx9du3bF/fffj8suu8zl/tFY13wps2itZ46Ul5ejf//+yM/Px1133YVp06Y5TRuN9cwZd8stGuuaP645kHWtTgddarUao0aNwtq1a63WnzlzBmfOnMEPP/yA0aNH48knnwxNBsPUwYMHQ52FiLBp0yYsXrzYrbRGoxGTJ0/G0qVLrdbn5eUhLy8PK1euxODBg/HSSy9BEIRAZDcseFJm0VoPN23ahOeffx7FxcVW64uLi7Fp0yZs2rQJX331Fd5//300adLEbv9o/dzztdyisb799ddfeP7551FaWmq1/tixYzh27BgWLVqEESNG4Omnn3a4fzTWNV/LLBrrmTNvvvkm8vPza00XjfXMFXfLLRrrmq/XHOi6VqeDrokTJ0oF16JFC9x3333IzMzEvn37sHjxYlRXV2PWrFnIzs7GnXfeGdrMhhFzpU1ISMBbb73lMm1GRkYwshR29u7dixEjRsBoNLqV/p133pECrgYNGmDQoEFo1KgRjh07hoULF6K4uBhfffUVMjIyMGzYsEBmPWQ8LTPLD8/Zs2cjJibGadqEhASf8xcO/v33XzzzzDOoqakBYPolrmfPnkhNTcXZs2fx/fff4/Dhwzh8+DAef/xxfPvtt0hOTrY6RjR+7vmj3KLtc89cZhqNBgDQo0cPdO3aFYmJiThw4ACWLFkCtVqNOXPmIDExEYMHD7Y7RrTVNX+UWbTVM2fWrl2L5cuXu5U22uqZK56UWzTWNV+vOeB1TayjNmzYIObk5Ig5OTni0KFDRbVabbX9yJEjYseOHcWcnByxY8eOYkVFRYhyGn7uueceMScnR7z//vtDnZWwtHbtWvHaa6+V6ldOTo747rvvOk1/5MgR8fLLLxdzcnLE/v37i6WlpVbbz58/L95yyy1iTk6O2KZNG/H06dOBvoSg87TMRFEUR40aJebk5IhdunQJUi5D76GHHpLKZ+HChXbbdTqdOGbMGCnN1KlTrbZH6+eer+UmitH3uTdo0CCXZXb8+HHpPXvNNdfY1ZVorGu+lpkoRl89c6SsrEzs2rWr1f8H48ePd5g2GuuZM56UmyhGZ13z5ZqDUdfq7EQaCxYsAAAoFAq8/vrrdr+St2jRApMmTQIAlJaW2nX7ilZGoxFHjhwBALRs2TLEuQkvWq0W7777Lp5++mmUl5e7vd/nn38Og8EAAHjllVeQkpJitb1evXqYPn06AECn0+HTTz/1X6ZDzNsyAy7+YpWTkxOIrIWdo0ePYtu2bQCAXr164cEHH7RLo1Ao8MYbb6BevXoAgGXLlkl1C4jOzz1/lFu0fe4dOXIEf//9NwBTq6CjMmvWrBn+7//+DwBQWVmJ9evXW22PtrrmjzKLtnrmjLl7nG1rsyPRVs9c8aTcorGu+XrNwahrdTLoKi0txcaNGwEA3bp1c9h/HwD69esnNS/WpUH4vjh58iSqq6sBRM+XXXds3LgRffv2xbx582A0GhEfH49HH3201v2MRiN+++03AKbyvPbaax2ma9++Pdq0aQMA+O233yCKov8yHyLelhkAaDQaaYB6tNTDTZs2Scuuui3ExMTgpptuAmCaOCI3NxdA9H7u+VpuQPR97pWWlqJz587IzMzErbfe6jRd69atpeUzZ85Y7R9tdc3XMgOir545Ytk97j//+Y/LtNFYz5zxpNyA6KxrvlxzsOpanQy6/v77b2ncSOfOnZ2mk8lkuO666wAAu3btQllZWVDyF84sx9FEyxvVHStWrMDp06cBAFdeeSWWLl0qfYFz5dChQ9KAa1d10XJ7fn5+nRgA622ZAcDhw4ellohoqYcymQwtW7ZEYmIimjVr5jKtZWupuQUxWj/3fC03IPo+96699lp8/vnn+Ouvv3Dvvfc6TZeXlyctZ2VlScvRWNd8LTMg+uqZrfLycqml4P7778f111/vMn001jNHPC03IDrrmi/XHKy6Vicn0jh8+LC0XFvBm6d2FUURhw4dkgozWh06dEhaNjfPnjp1CidOnIBcLkfTpk3RsGHDUGUvpNLT0zFixAg88MADkMvlKCoqqnUfy/J0ty4CpgHblr+YRipvygxwXA/Pnz+Po0ePQhRFNG7cGE2bNg1InkNl0KBBGDRokFtpzV0oACA1NRVA9H7u+VpuAD/3HCktLZW628TFxeHGG2+UtkVrXauNqzIDWM/M3eMaNGiA//znP3YzQNpiPTPxtNyA6KxrvlxzsOpanQy6LJv0GzVq5DJt/fr1rfarS29Ub5h/KcjMzMT69evx4Ycf4ujRo1Zp2rRpg9GjR9v9h1KXPfTQQ3jllVcQGxvr0X6e1MUGDRo43C9SeVtmwMV6KJPJcOrUKUyZMgW7d++2StOsWTMMHz4cd9xxh1/yGynOnz+PP//8EwCQlpYmBZ/83HPNWbkB/Nwz02g0OHPmDP744w988cUX0rTUEyZMQHp6upSOde0id8sMiO56Ztk97tVXX0ViYmKtwQPrmXflBkRnXfPlmoNV1+pk90LL+7SkpaW5TGv5a6c7FbmuM1fawsJCjB8/3q7CAsC+ffvwxBNP4L///W+wsxcyV111lVfBgyd10bLrU12oi96WGXCxHhqNRjz33HN2ARcA5ObmYty4cZg4caLb09DXBdOnT4dOpwMA9O/fHzKZ6WOcn3uuOSs3gJ97gOmWDm3btkXfvn0xc+ZM5OfnIysrC3PmzMEDDzxglZZ1zcSTMgOit55Zdo8bMGAAunfv7tZ+0V7PvC03IDrrmi/XHKy6VieDLrVaLS27urcPAKhUKof7RaPq6mqcOnVKet60aVO89dZb2LBhA3bv3o0VK1Zg8ODB0s17586di2XLloUquxHBsk5Z1jVHLOtqtNdFy24CWVlZmDx5MtatW4c9e/bg119/xbBhw6Ty/PbbbzFv3rxQZTWoFi9ejJ9++gkAEB8fb3WDRn7uOeeq3Pi5Z2I5FsmstLQUv/zyi9WkIwDrmpknZRbN9czcPS4rKwsvvvii2/tFez3zttyisa75es3Bqmt1MujS6/XScm1fdC23W+4XjU6ePClVtvbt22PZsmW48847kZWVhZiYGLRq1QqTJk3C1KlTpX2mT5+OqqqqUGU57LEueq60tFT6IGvatCmWL1+Ohx56CPXr14dKpULz5s0xatQofPjhh1AoTD2k58+fj7Nnz4Yy2wG3atUqTJkyRXr+yiuvIDs7W3rOuuZYbeXGzz2TtLQ0TJo0CXPmzMH48ePRunVr6HQ6rFy5Evfccw/27dsnpWVdM/GkzKK1nll2j3N0yxRXorme+VJu0VjXfL3mYNW1Ohl0WXZpMncncUar1UrLtRV0Xde6dWvs3LkT69evx0cffYTExESH6e666y7cfPPNAExfkH///fdgZjOieFsXlUplwPIU7lJTU7Fjxw5s3LgRixcvtpsBzOz666+X7pWj0+ki/pc6V1auXInRo0dLMzoOHTrUbnp0fu7Zc6fc+Lln0qFDBwwePBj9+vXDY489huXLl+P+++8HAFRUVGDs2LFSObKumXhSZtFYzyy7x/Xr1w+9evXyaP9orWe+lls01jVfrzlYda1OBl3x8fHSskajcZnWsvBqa1KMBoIgIDs7G0lJSS7TWX5xMd8wkuxZ1kXLuuaIZV31dixUXZKRkWE3EN2WZT3cvn17oLMUEkuXLsVzzz0n/Udw1113Oexqws89a+6WG8DPPUdkMhlefvlltGrVCgBw7NgxbNiwAQDrmjOuygyIvnpm7h5nbhH0VLTWM1/LDYi+ugb4ds3Bqmt1MuiyvFt3bYPcLLfX9gWPLrr00kulZfNsTWTPk7poeb8H1kX31PV6+N///hcvvfSS1IXh7rvvxptvvin1S7fEz72LPCk3T9T1+mZLLpfjnnvukZ7v2LEDAOuaK87KzBN1oZ5Zdo+bNGmSV699NNYzf5SbJ+pCXfOUo2sOVl2rk0GX5Q0yz5075zKt5UDYunbPgkAyj6Uh1zypi5bbWRfdU1e7YRqNRkyePBlz586V1g0dOhRvvPGG1ax7lvi55125eSIaP/eaN28uLZtn+GJdc81RmXmiLtSzX3/9VVp+/vnn0apVK7tHz549pTTLly+X1pvfv9FYz/xRbp6oC3XNU46uOVh1rU6WdosWLaTlw4cPo2PHjk7Tmm+IJgiCdDO1aLV+/Xr8+++/KCoqwtNPP+1y2szz589Ly87G3JD1DY8tb77niCc356vLtm/fjn/++QdFRUW47777rL7A2KqL9dBoNOI///kPfvjhB2ndqFGjMGzYMJf7RfvnnrflFo2fe8uXL8eaNWtw6tQpfP7551a/8tqy7GqTkJAAIDrrmq9lFo31zFfRWM/8IRrrmq/XHKy6VieDrvbt20OpVEKn02HLli146KGHHKYzGAzYtm0bANMgPFcfotFg5cqVWLp0KQDg2muvxS233OI07aZNm6Tlq6++OtBZi1jNmzdHZmYmCgsLsWXLFpdpzdtTU1OjOujaunUr3n77bQBAdna2y6Br48aN0nK7du0CnbWgeOWVV6TAQSaTYfLkydKEIa5E++eet+UWjZ97hw4dwm+//QbAdE19+vRxmtbyHnnmbjnRWNd8LbNoq2dDhgypdQKIoqIiTJ48GQDQqVMnDB06FMDFlsJorGf+KLdoq2uA79ccrLpWJ7sXJicno3PnzgCA1atXO51K+qeffpKa/vv27Ru0/IUrc5kBpvvaOFNcXIwlS5YAAOLi4tC7d++A5y1SyWQy6QN079692Llzp8N027dvl6YX7tOnj1+6QkUqy3r47bffSrN/2dJoNFiwYAEA0y9Od9xxR1DyF0jLli2T3lsymQxTp051K3AAovtzz5dyi8bPvW7duknLCxcudJqutLRU+iKjVCqlm7NGY13ztcyirZ61adMGvXr1cvno0qWLlL5hw4bSenOrQzTWM3+UW7TVNcD3aw5WXauz3+weeeQRAKapH59//nlUVlZabT9y5AjeeOMNAKbm/3vvvTfYWQw7vXr1Qv369QEAGzZskL7QWiovL8eIESOkgYSPPvqoR/ePiEZDhgyR+hCPHz8eBQUFVtvPnz+PCRMmADANwn744YeDnsdw0r59e7Rp0waAqRl/6tSpEEXRKo1Wq8W4ceOkG5DecccdVt0DIlFhYaH0mQSY6sqAAQM8OkY0fu75Wm7R+LnXuXNntG7dGgCwefNmfPLJJ3ZpKisrMXLkSOkLxn333Wd1f7Noq2u+llk0IyIUaQAAHn5JREFU1jN/iLZ65g/RWNf8cc3BqGuCaPttpg559tlnpe4AjRs3xoMPPogGDRpg//79WLRokXRTtFdffRUPPPBAKLMaNtatW4dnnnlGal3o0qUL+vTpg6SkJBw+fBhLly6VgoaOHTvik08+ifh7Ynhry5YtUrP+iBEjMHLkSKdpp0+fLn0IZGRkYNCgQWjevDmOHz+OhQsXoqioCADw5JNPYsyYMYHPfIi4W2b79u3DQw89hJqaGgBA27ZtceeddyI9PR0nT57Ed999h5MnTwIwjZv7+uuvkZqaGpRrCJRZs2bhww8/BGD69XLixIlu7XfFFVdYDeaNts89f5RbNH7u2b7Hunbtit69ezu85iuuuAILFy5EXFyc1TGira75WmbRWM9cOX36tDQpxF133YVp06Y5TBdt9aw27pRbNNY1f1xzoOtanQ66ampqMGLECKv7ZFgSBAEjRozAiBEjgpyz8PbLL7/gxRdfRHV1tdM0N998M2bOnCkNEo5GngRdBoMBEydOlKaCdeTee+/FlClT6nTXQk/KbOvWrRg9erQUkDpyzTXX4J133kG9evX8ntdg69q1q10rqDumTp2KgQMHSs+j7XPPX+UWjZ977rzHunXrhtmzZzscuxBtdQ3wvcyisZ45427QFY31zBV3yy0a65qv1xzoulYnJ9Iwi4uLw8cff4wffvgB33//PQ4cOICKigqkpqaiQ4cOGDp0KDp06BDqbIadvn37okOHDvj666+xfv16nDhxAjqdDpmZmWjXrh3uuusuqZ86uUcul2PatGno27cvFi9ejD179qC0tBSJiYlo164dHnzwQfTo0SPU2QwrHTt2xK+//orFixdj9erVOHr0KGpqapCeno4rrrgCt912G/r37+/zvZfCQXFxsVeBgyPR9Lnnz3KLxs+9jh074pdffsHXX3+N1atX4/jx49BoNMjIyED79u1rveZoqmtmvpZZNNYzX0VjPfOHaKxrvl5zoOtanW7pIiIiIiIiCrW624+JiIiIiIgoDDDoIiIiIiIiCiAGXURERERERAHEoIuIiIiIiCiAGHQREREREREFEIMuIiIiIiKiAGLQRUREREREFEAMuoiIiIiIiAKIQRcREREREVEAMegiIiIiIiIKIAZdRER1gMFgCHUWiIiIyAkGXUTksVatWkmP1q1b4/z5827vO2zYMKv9g23ZsmUOz3369Glp/bJly4KeL1/89NNPGDdunN36uXPnhqycqW6oK3WoqqoKvXr1Qs+ePaHVakOdHa889NBDuO6663Du3LlQZ4WIvMCgi4h8IooifvnlF7fSVlRUYP369QHOUXSZM2cOnn/+eRQUFIQ6K0Rh680338SpU6cwbtw4qFSqUGfHKy+88AIqKiowYcIEiKIY6uwQkYcYdBGRz3799Ve30q1cuRI6nS7AufGOUqnEJZdcgksuuQSJiYmhzo7bXLUypqSkSNdEFK3+/vtvfPvtt2jfvj1uvfXWUGfHa1deeSX69++PzZs34/vvvw91dojIQwy6iMhrOTk5AICdO3ciLy+v1vQ///wzACArKyug+fJGdnY2fv/9d/z+++/o3bt3qLPjF0OHDpWuiSgaiaKI1157DQAwatSoEOfGdyNGjIBMJsOMGTNQVVUV6uwQkQcYdBGR17p3746EhASIolhra1dxcTE2b94MhUKBW265JUg5JKJo9uuvv+Lff/9FTk4Orr/++lBnx2fNmzfHjTfeiKKiInz99dehzg4ReYBBFxF5LSYmBjfffDOA2rsY/vLLL9Dr9ejcuTPS09NrPfaRI0cwadIk9OrVC23btsW1116Le+65Bx999BFqampc7nvmzBm8/vrr6Nu3L9q1a4euXbvihRdewKlTp5zu485EGuvXr8d//vMf9O7dG9dccw2uvPJK3HDDDXjsscewdOlSh10nzRN3mMtp9+7dGD16NLp27Yorr7wS3bt3xwsvvICjR4/WWiaWzBMcLF++HACwdetWKf+nT5+2SmM7CcKWLVus1h84cADPP/88unbtirZt26J3796YPXs2qqurAZhmRvzyyy9x11134eqrr0aHDh3w8MMPY9OmTS7z+M8//2Ds2LHo0aMHrrrqKnTs2BGDBw/GokWLvO5mas73+vXrUVBQgMmTJ+PGG29E27Zt0adPH0yePBknT550eQydTodFixZhyJAh6NSpk/Q6jBkzBjt37nS4j2WZabVazJ49GzfccIN03h9++MHlOR944AG0atUKDz/8sMPtP/74o3T8xYsXO0xjnoRm5MiRdtvWrl2L4cOHS/WqU6dOGDJkiMuyvvnmm9GqVSssXboUGzZswB133IErr7wSXbp0wZgxY1xeD2CqFyNHjpTy/dlnn9mlCUVZW/r4448BAHfffbfTNH/++SeeffZZdOvWDVdeeSWuu+46DBw4ELNnz651rKQvdby4uBjz58/HwIED0bFjR1x11VXo06cPXn/9dZeTZdxzzz0AgM8//zxsu2sTkT1FqDNARJGtf//++OGHH6QuhvXr13eYzty18LbbbpOCAmc+/fRTzJgxw2oadI1Ggz179mDPnj1YtGgRPvzwQ1x22WV2+65fvx6jRo2SAgYAUKvVWLZsGX777TfcfvvtHl9jTU0NnnvuOaxZs8ZuW1FREf766y/89ddf+PHHH7FgwQLI5XKHx1m8eDGmTJlidV15eXlYtmwZfvzxR3z44YdB/zX+559/xn/+8x+rL28nTpzA/PnzsXXrVixYsADPPPMMNm/ebLXf5s2bsXXrVsydOxe9evWy2mY0GvHWW2/h008/tVqv1Wqxbds2bNu2Dd988w0++OADZGdne5XvM2fOYOLEicjPz5fW5ebmIjc3F//73//w9ttv46abbrLb79y5c3jyySdx6NAhq/V5eXn48ccf8eOPP+KJJ57AmDFjIAiCw3O/8cYbVoFRbm4uGjdu7DK/3bt3xz///IN//vkHGo0GMTExVtstA9ht27bhgQcesNqu1WqlNJbXVVNTgzFjxuCPP/6wSl9aWoqtW7di69atWLhwIT744AM0atTIYd527tyJV199VaoDhYWFSEhIcHk9oijixRdfxMqVKwEA48ePxyOPPGKVJlRlbbZ3717s3bsXANCnTx+HaebMmYMPPvjAap1Op8O+ffuwb98+fP311/j444/Rvn17qzS+1vGtW7di9OjRKCoqslpvrsPLli3De++9h86dO9vt2717d8TFxaGwsBC///47+vXrV3thEFHIsaWLiHzSpUsXJCcnu+xieP78eWzfvh0qlarWroVLly7FtGnTYDAY0LFjRyxYsACbNm3C2rVr8frrryMrKwtnzpzB448/juLiYqt9jx8/jmHDhqG6uhoNGzbEnDlzsHHjRqxatQqjR4+GTqdz2orgyowZM6SAa/DgwVi+fDk2bdqE33//HW+//bYU/G3evNnpr/AFBQWYMmUKmjdvjrlz52Ljxo34448/MHLkSMjlcmi1WkyePNntWcmeeuop7NixQwoiO3TogB07dmDHjh1Ov1w7MmHCBGRnZ+Odd97Bhg0b8P3336Njx44ATL/iDxw4EFu3bsX//d//4ZdffsGmTZswe/ZspKSkwGg0YsaMGXbHfPfdd6Uvo71798bChQuxZcsWrFq1ChMmTEBiYiL279+Pp556ChqNxu28WnrrrbeQn5+PQYMG4eeff8bGjRsxa9YsZGVlQa1WY+TIkTh27JjVPtXV1Xjsscdw6NAhxMfHY8yYMfjtt9+wZcsWfPPNN+jfvz8A4KOPPsJHH33k9NyLFy9Gnz598Ntvv2H16tV45ZVX7L6U2+rRowcA048H27dvt9tuGXRt3brVbvv27dtRXV0NmUwmHQsAnn/+eSng6tu3L5YsWYItW7bg559/xpNPPgmFQoFDhw7hscceQ2VlpcO8ffvtt8jIyMCCBQuwceNGvPfeexgyZIjL65kyZYo0mcPYsWPx2GOPWW0PZVmbmX/oueyyy9CgQQO77f/8848UcN1+++349ttvsWnTJqxatQovv/wy4uPjUVlZiRdeeAFGo9FqX1/q+KlTp/DEE0+gqKgIGRkZePXVV7FmzRqsX78e06dPR2ZmJqqqqjBq1Ci7oAwAVCoVOnXqBMD9SYyIKAyIREQeysnJEXNycsR3331XFEVRnDBhgpiTkyPef//9DtMvWLBAzMnJEYcPHy6Koii+++670jEslZeXi+3btxdzcnLEYcOGiQaDwe5Yp06dEq+55hoxJydHfO2116y2Pf3002JOTo7YsWNH8dy5c3b7/v7779J5bc996tQpaf13331nlac2bdqIOTk54sSJEx1eX35+vti2bVsxJydHfP755622fffdd9Jxe/bsKVZUVNjt/9Zbb0lp9uzZ4/AczowfP17MyckRBw8ebLfNWTlv3rxZWt++fXsxLy/PanteXp54xRVXSGnmz59vd+yvv/5a2n7mzBlp/fHjx8XWrVs7fH3Mdu3aJV5++eViTk6OuGDBAo+u1/L1mzt3rt32EydOSPXjmWeesdr29ttvizk5OWKbNm3EnTt3Ojz+a6+9JqXJz8+X1luW2U033STqdDqP8i2KotitWzcxJydHnDlzptX6kydPijk5OWKHDh2ksjt+/LhVmmnTptm9x1avXi3ladq0aQ7P+euvv0pppk+fbrXtpptukrb99ddfDvd3VIdmzJghrfvggw8c7hfqshZFUbztttvEnJwc8cUXX3S43Vymt9xyi2g0Gu22L1u2TMrH3r17pfW+1nHz59Q111wj5ubm2u27e/du6fj//e9/HR5/3rx50jEcfU4SUfhhSxcR+czcvWXnzp0OxyL89NNPACD9uu3MihUrpBm5JkyYAJnM/iOqcePGGDx4MADTeCm9Xg/A+h5gjz76qMNujr169ULXrl3dvSzpuI888gj69u2LRx991GGarKwsNG/eHADsWt8sDR482OF09JbdxWrreulvd9xxh133p+zsbDRp0gSAadze0KFD7fazbG2wnLZ+yZIlMBqNiIuLw3PPPefwnG3btpXqzDfffONVvi+55BI888wzDtebu7mtX78eFRUVAEzd4ZYsWQLAVA/btWvn8LijR49GbGwsdDqdNF7O1i233AKFwvPe+d27dwcAbNy40Wq9uZWrU6dOUqupbWvXn3/+CQBWrVzmssvKysLzzz/v8Jx9+vSRun9+8803Vl1bzVJTU93u1vr+++9LLVOjRo3CU089ZZcmHMq6srISR44cAXBxllVb5u6U1dXVDmcCvPXWW/Hee+/hhx9+QMuWLaX1vtTx8vJy6bV89NFH0bRpU7t9r7rqKvTt2xcdOnRAbGysw+Obr6mystLj8aBEFBoMuojIZ9dffz1SU1MhiiJ+++03q20nT57Enj17EB8f73CMjaUtW7YAANLS0pCeno6qqiqHj7Zt2wIAqqqq8O+//wIwjYMxB2A33nij03PYjj+qTcOGDTF27Fi8/fbbaNGihd12rVaLv//+W5rcw5wHR8z5tmU5sYharfYof75y9oU4IyMDAHDppZc6/OJnGTxqtVpp2RwsXHrppQDg9DU0n/fYsWMoKSnxON99+vRxOnbOXM90Op2UnyNHjkhdtS6//HKn+RIEQZpgZMeOHQ6Pf/nll3ucX+BiwLR//36UlZVJ681BV8eOHaU6Yhl0nTt3DocPH7a6NsBU5wHThBhKpdLpec1f/isqKqT3i6XWrVs7HVNl6csvv8Tbb78NwBRMDRs2zGG6cCjr3NxcqUuguS7auvbaawGYuv7efffdWLBggVUAExcXh549eyInJ8fqhsq+1PFt27ZJwZ5lAG1r9uzZWLhwIR5//HGH280/8gCmbtVEFP44kQYR+UyhUKBPnz5YsmQJfvnlF6sB9eZWrl69ejn91dbM3MpTUlKCa665xq1z5+Xl4corr7RqYXN1M2BHgZO7jh49ir///hvHjx/HyZMnceLECeTm5roMtCw5m7XR8gud7diRQEtLS3O43tzKmJSU5HC7sy/p5tdw3759Hr2GzvLhjLPWCwBo1qyZ1bEBWM1cOXXqVEydOrXWczibQc6d2Tcduf7666FSqaDVarFlyxb07t0boihKk5R06tQJSUlJ+Pbbb6WACoDUgtuoUSMpSKmsrJRa8Wqr05bbz507hzZt2nh1PW+88Ya0/OeffyI/Px/16tWzSxcOZW1538CUlBSHaXr37o1evXph1apVyM3NxfTp0zF9+nQ0bNgQXbt2xU033YSuXbtavT8B3+q4ZauwZT31VGpqqtWxiSj8saWLiPyib9++AIBdu3ZZfYGynLWwNs4G+ruzj/kLKGD6hdoZZ0GEK2fOnMHDDz+Mfv36YfLkyfj000/xxx9/4MiRI0hMTESfPn2k7niueNNNKtBclZU3fHkNPZGcnOx0m2Vwbz62P/NlO/Ogu+Lj46VJSsytWwcPHkRxcTFSU1PRqlUraYKEvLw8aep7c9Bl2cpl2R0uPj7e5XktX2NH3ejcvR5RFHHbbbdBoVCgvLwcU6ZMcZguHMracvZSR116AdMPC3PnzsXrr79uFYiePXsW33zzDZ555hl0795dmjCktry6Yt7HsoXTl/ee5eeY5bUSUfgKv28ARBSROnXqhKysLBQUFODXX3/Fo48+iiNHjuDQoUNITU3FDTfcUOsxzF+W27Vr5/FYH8tfs2tqapx+0bLsCueO8vJyDBkyBGfOnIFMJsONN96Ia6+9Fi1btkSLFi2kYOvBBx90eR+waBEbG4vKykr069cPc+bMCdh5XM16aPkl1NyCZvkF96OPPnLZBTWQevTogQ0bNkjjuixbuQRBQKNGjdC4cWOcPn0aW7ZsQYMGDRxOFW8ZaNX2pduTAM2Vxx57DOPHj0e9evWwYMEC/P7771i5ciV69+5tlS4cytqyJdZZN1TAFHjde++9uPfee5GXl4c///wTmzZtwl9//YXS0lIUFxdj/PjxSE5Olu6150sdtywbV59TtbEcm+dO11AiCj22dBGRX8hkMuleOOZpjM2tXH369HE55sSsYcOGAEwtS66IDqZVN+8LwG6qcEueBkYLFy6U8vPOO+9g/vz5eOKJJ9CjRw+r1i1vxiXVRb68hp5w9TpajnExT59vOWV4oPPminkcT25uLs6dOyeNYzS3cAGQ7s20detWbN++HVVVVVatZICp9cbc2lfbRArmCSUA6/eJp8aPHw8AGDlypHScKVOmoLy83CpdOJS15X3GaruZuln9+vVx7733Yvbs2dItCMxdC7/88kspnS913LJsXNXhTZs24b///S/+97//Odxu+XlT2z3ViCg8MOgiIr+x7GKYl5eHX375BYB7XQuBiwPbCwsLsWvXLqfp5s+fj2uvvRZ33HGH1AWrY8eO0q/Iq1atcrqvuauWu/755x8AphYT21/0zfLy8pCbmwsg+GOywu1XbvNruG/fPpdjTSZPnoxOnTrh7rvv9qq7lqvX0XzfqoSEBFx33XUATJNFmFsVbG8kbKmqqgpdunTBTTfdhJkzZ3qcr9o0adJEGmP1119/SRNIWAZU5gBs27Zt0nXaji0SBAEdOnQAAKxevdrq5ta2zD+CJCQkuBwL5674+Hi89NJLAEyTULz11ltW28OhrC2DG8txVJZGjRqFnj17OrzXnFwux2233SbNdmp5DF/qePv27aX3rHkWQ0cWLVqEuXPn4v3333e43TLocnQPMiIKPwy6iMhvOnTogPr160MURbz33ns4duwYsrOzpS8ptRkwYID0xfK1115z+Av1yZMn8emnn6KiogJarVZqbYqLi5OCvi+//FKa7c3Szp07pYk93GXumlRWVoaCggK77RqNBhMnTpR+1Xb15TcQzPkL9nmdue+++wCYZnF89dVXHU5RvmvXLixfvhylpaVITU31qouV+Ua0to4ePSq1SvTv31+qTwqFAgMHDgRg+rJr/kHA1pw5c1BUVISzZ8+idevWHufLHeap47/66iuUlpYiIyPDakpyc0vXuXPnpKnUHc38aS7rgoICzJ492+G5Vq1aJQU+AwYMcKvF2R09e/aUZgL99ttvpRY7IDzKulmzZtJkMOYfZmyp1WqcPn0aK1ascHirB61WK7UiWk7O40sdr1evnhTILViwwGFAuGfPHqxevRqA89tsnD17Vlp2NjsjEYUXBl1E5DeCIODWW28FcPHeNP369XN4vy1HMjMzMWrUKACmLx733XcfVq5cicLCQpw9exbLly/HkCFDUFpaCkEQMHHiRKuWnnHjxiE1NRXV1dUYPHgwvvnmG+Tn5+P8+fP46quv8Pjjj3vcncn8BcloNOKpp57Cpk2bUFRUhFOnTuF///sf7rnnHmzYsEFK72iigkAyz2J28OBB7Nu3DyUlJSENwC6//HIMGjQIgKkFZujQodiwYQOKi4tx8uRJfPnll3jiiSeg0+kQExODcePGeX2u5557Dh9++CHOnj2LwsJCLFu2DEOGDEF1dTVSU1Pt7qE0bNgwqVVgzJgxmD59Og4dOoSSkhLs2bMH48ePlwK2Dh06SFOt+5u5i+GBAwcAWLdyAaYv5pb3fZPJZA6nF7/55pulcUYLFizAc889h127dqG0tBTHjh3D22+/jdGjRwMwtbA5u5eXt1566SXEx8dDFEVMmjTJ6nYHoS7ruLg4abr5/fv3O0xjno49Pz8fDz/8MFauXIkzZ86gsLAQW7duxdNPP40TJ04AMI3ZNPO1jo8fPx6xsbEoKSnBAw88gBUrVqCgoACnT5/G0qVL8eSTT0Kn0yE7O9tqJlhL5hbSjIwMq+njiSh8cSINIvKrfv364bPPPpOCm9puiGzr8ccfR1VVFd5//30cOnQII0eOtEujVCrx8ssvo1u3blbr09PT8dlnn+GJJ55AQUEBJk2aZLU9JiYGo0ePxqxZs9zOzz333IOff/4Z27Ztw759+xx+Cbr00ktxxRVX4Mcff8SZM2eg0+n81qJQm06dOuGjjz5CdXW11LrwxRdfWI0RCrYXX3wRWq0W3377Lf7++2+H9xpKSEjA7NmzvW7huOmmm/D3339j1qxZdq9n/fr1MX/+fLvpxtPS0rBgwQI888wzyM3NxYIFC7BgwQK7Y7dt2xb//e9/3f6xwFMdOnRAUlKSNOOmo9eqc+fO0ti0du3aOZ06febMmRg7dixWr16Nn3/+WRpHaalNmzZ45513vJ60wZkGDRrg2WefxbRp03DixAnMnTtXCjDCoay7dOmCffv2WU2/b6ljx44YM2YMZs+e7fSzRiaTYdSoUVLrpJkvdbxly5Z477338Oyzz+Ls2bMOf3jIzs7GRx995HS2VXPQ1aVLF8cXT0Rhhy1dRORX7dq1kyYvaNq0Ka666iqP9hcEAaNGjcL333+Pe++9F02bNkVsbCxUKhWaNWuGBx54AP/73/9w7733Otz/8ssvx4oVKzB8+HDk5OQgLi4O6enp6NOnD5YuXWrXqlAblUqFBQsWYOzYsbjiiisQFxcHhUKB9PR0dOzYEZMnT8by5culX8JramqsWr4CrVu3bpg8eTKaNWsGpVKJ9PR0FBYWBu38jiiVSrzxxhv46quvcNttt6FRo0ZQqVSIjY1Fy5Yt8dhjj+Hnn392eXPY2rRp0wbff/89BgwYgIyMDMTFxaFVq1YYNWoUVqxY4TSYu/TSS7FixQpMmjQJHTt2RGpqKhQKBVJTU9GpUye8/vrrWLx4sdf3h3KHQqGw+rLsqE5aBmKuyikhIQHvv/8+5s2bh549eyIrKwtKpRL16tXDDTfcgOnTp2Px4sVu3dLAG0OHDpValD777DOrVqVQl7X5B59Tp05JYy5tPfnkk1i0aBHuuOMONG7cGCqVCjExMWjSpAnuvvtuLF26FE8//bTdfr7W8S5duuC3337D//3f/6Fly5aIj49HTEwMWrZsiWeeeQYrVqyQ7slmq7KyEvv27QPg/nhZIgo9QQzkNE1ERER+ZP4iOmLECIctE0SWBg0ahO3bt+PZZ5/F8OHDQ50dv1iyZAkmT56MSy65BL/++qvLKfGJKHywpYuIiIjqpGHDhgEAvv/++6DPLBooS5cuBWBqpWPARRQ5GHQRERFRndS1a1e0bdsWJ0+exLp160KdHZ/9+++/2LNnDxo0aIABAwaEOjtE5AEGXURERFRnjR07FoIgOL3nVSSZO3cuAGD06NFBm6yHiPyDQRcRERHVWZ06dcJDDz2EXbt2ubxxerjbvn07Vq1ahR49erCViygCMegiIiKiOm3s2LFo2rQppk+fDq1WG+rseGXGjBlITk7GlClTQp0VIvICZy8kIiIiIiIKILZ0ERERERERBRCDLiIiIiIiogBi0EVERERERBRADLqIiIiIiIgCiEEXERERERFRADHoIiIiIiIiCiAGXURERERERAHEoIuIiIiIiCiA/h+OkaRrbKQWkwAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1000x900 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# needs df8 for violin, df9 for points, dictionaries (median_time, ci, ci_delta) for error plots, df6 for individual points\n",
    "\n",
    "modes = ['SQL', 'RD']\n",
    "patterns = [1, 2, 3, 4]\n",
    "y_tilt_section = [0, 1, 2, 3]\n",
    "\n",
    "# Define pre-settings\n",
    "figwidth = 10\n",
    "figheight = 9\n",
    "xlab_size = 20\n",
    "ylab_size = 20\n",
    "figfont_size = 20\n",
    "\n",
    "# Define consistent color maps\n",
    "my_cmap_sns_light = [(0.7764705882352941, 0.8588235294117647, 0.9372549019607843), (0.9921568627450981, 0.8156862745098039, 0.6352941176470588)]        # light orange, light blue\n",
    "my_cmap_sns_dark = [(0.19215686274509805, 0.5098039215686274, 0.7411764705882353), (0.9019607843137255, 0.3333333333333333, 0.050980392156862744)]      # dark orange, dark blue\n",
    "my_cmap_dark = sns.color_palette(my_cmap_sns_dark, as_cmap=True)\n",
    "my_cmap_light = sns.color_palette(my_cmap_sns_light, as_cmap=True)\n",
    "\n",
    "# Create empty figure and plot the individual datapoints\n",
    "fig, ax = plt.subplots(figsize=(figwidth,figheight))\n",
    "\n",
    "\n",
    "# 1. Violinplots\n",
    "axsns = sns.violinplot(x='median', y='pattern', data=df8,\n",
    "                       hue='mode',\n",
    "                       hue_order=['SQL', 'RD'],\n",
    "                       split=True,   # half violinplots https://stackoverflow.com/questions/53872439/half-not-split-violin-plots-in-seaborn\n",
    "                       inner='quartile',\n",
    "                       cut=0,               # 0 means ending sharp at end points\n",
    "                       width=.4,\n",
    "                       orient = 'h',\n",
    "                       zorder=20,\n",
    "                       palette = my_cmap_light,)\n",
    "\n",
    "# change the medium default line to full (https://stackoverflow.com/questions/60638344/quartiles-line-properties-in-seaborn-violinplot)\n",
    "for l in axsns.lines[1::3]:\n",
    "    l.set_linestyle('-')\n",
    "    l.set_linewidth(1.2)\n",
    "    l.set_color('black')\n",
    "    l.set_alpha(0.8)\n",
    "\n",
    "\n",
    "# 2. Plot individual points\n",
    "y_base = np.zeros(df6.values.shape[0])  # base vector to which to broadcast y-tilt values\n",
    "y_tilt_mode = [0.3, 0.39]\n",
    "for i, mode in enumerate(modes):\n",
    "    for j, pattern in enumerate(patterns):\n",
    "        column = (mode, pattern)\n",
    "        ax.plot(df9[column],\n",
    "                y_base + y_tilt_mode[i] + y_tilt_section[j],\n",
    "                '^',\n",
    "                alpha=1,\n",
    "                zorder=20,      # higher means more visible\n",
    "                markersize=10,\n",
    "                markeredgewidth=0,\n",
    "                # markerfacecolor='none',\n",
    "                markerfacecolor=my_cmap_sns_dark[i],\n",
    "                markeredgecolor=my_cmap_sns_dark[i],)\n",
    "        ax.plot(df9[column],    # white background behind the markers, but in front of the connecting lines\n",
    "                y_base + y_tilt_mode[i] + y_tilt_section[j],\n",
    "                '^',\n",
    "                markersize=10,\n",
    "                markeredgewidth=1,\n",
    "                markerfacecolor='white',\n",
    "                color ='white',\n",
    "                linewidth = None,\n",
    "                zorder=1,)\n",
    "\n",
    "\n",
    "# # 3. Plot lines connecting individual points\n",
    "# for idx in df9.index:\n",
    "#     for i, mode in enumerate(modes):\n",
    "#         for j in range(len(patterns)-1):\n",
    "#             start = (mode, patterns[j])\n",
    "#             end = (mode, patterns[j+1])\n",
    "#             ax.plot(df9.loc[idx, [start, end]],\n",
    "#                     [y_tilt_mode[i] + y_tilt_section[j], y_tilt_mode[i] + y_tilt_section[j+1]],\n",
    "#                     color =my_cmap_sns_dark[i], linewidth = 2, linestyle ='-', alpha = .2, zorder=0)\n",
    "\n",
    "\n",
    "# 4. CI Errorbars & numbers\n",
    "y_tilt_mode = [0.5, 0.57]\n",
    "for i, mode in enumerate(modes):\n",
    "    for j, section in enumerate(patterns):\n",
    "        column = (mode, section)\n",
    "        plt.errorbar(median_time[column],\n",
    "                     y_tilt_mode[i]+y_tilt_section[j],\n",
    "                     xerr=np.array([[ci_delta[column][0] ,ci_delta[column][1]]]).T,\n",
    "                     fmt='o', markersize=10,\n",
    "                     capsize = 0,\n",
    "                     lw = 3, alpha=1,\n",
    "                     zorder=100,        # higher means more visible\n",
    "                     color = my_cmap_sns_dark[i])    # 'black'\n",
    "        ax.text(median_time[column], y_tilt_mode[i]+y_tilt_section[j] + 0.22, f'{median_time[column]:.1f}',\n",
    "                horizontalalignment='center', color = my_cmap_sns_dark[i],\n",
    "                fontsize=figfont_size)\n",
    "\n",
    "\n",
    "# 5. Plot red line connecting medians\n",
    "for i, mode in enumerate(modes):\n",
    "    for j in range(len(patterns)-1):\n",
    "        start = (mode, patterns[j])\n",
    "        end = (mode, patterns[j+1])\n",
    "        ax.plot([median_time[start], median_time[end], ],\n",
    "                [y_tilt_mode[i] + y_tilt_section[j], y_tilt_mode[i] + y_tilt_section[j+1]],\n",
    "                color =my_cmap_sns_dark[i], linewidth = 3, linestyle ='-', alpha = .3, zorder=0)\n",
    "\n",
    "\n",
    "\n",
    "# # #Additional settings\n",
    "ax.set_xticks(range(0, 100, 5))\n",
    "ax.set_xlabel('Median time per worker (sec)', size = xlab_size)\n",
    "ax.set_ylabel(None)\n",
    "ax.set_yticklabels(['P1', 'P2', 'P3', 'P4'])\n",
    "ax.set_xlim(0, 50.05)\n",
    "ax.set_ylim(3.9, -0.25)\n",
    "leg = plt.legend(loc='lower right',\n",
    "                 borderaxespad= 0.2,\n",
    "                 frameon = True,\n",
    "                 labelspacing = 0.1)\n",
    "leg.get_frame().set_alpha(1)\n",
    "leg.get_frame().set_linewidth(0.0)\n",
    "\n",
    "for text, text2 in zip(leg.get_texts(), ['SQL', 'DX']):\n",
    "    text.set_text(text2)\n",
    "\n",
    "plt.grid(axis = 'x', linewidth = 0.5, color = 'lightgray')\n",
    "sns.despine()               # remove bounding box\n",
    "\n",
    "if savefig:\n",
    "    plt.savefig(fig_dir + f'/q3_figure_variant{VARIANT}.pdf', bbox_inches='tight')\n",
    "    plt.savefig(fig_dir + f'/q3_figure_variant{VARIANT}.svg', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "526669f1",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "# Question 4: Correctness\n",
    "\n",
    "(1) take mean correct over all questions and all users answered in SQL or RD (32/2*13) and the difference in correctness score\n",
    "(2) calculate 95% CI for each"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "9b281fc0",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dfq4b:\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>mode</th>\n",
       "      <th>RD</th>\n",
       "      <th>SQL</th>\n",
       "      <th>diff</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>worker_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.938</td>\n",
       "      <td>0.312</td>\n",
       "      <td>0.625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.875</td>\n",
       "      <td>0.125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.500</td>\n",
       "      <td>0.500</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0.875</td>\n",
       "      <td>1.000</td>\n",
       "      <td>-0.125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>0.875</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.812</td>\n",
       "      <td>0.188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.875</td>\n",
       "      <td>0.125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.750</td>\n",
       "      <td>0.250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.812</td>\n",
       "      <td>0.188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>0.938</td>\n",
       "      <td>0.688</td>\n",
       "      <td>0.250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>0.812</td>\n",
       "      <td>0.938</td>\n",
       "      <td>-0.125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>0.625</td>\n",
       "      <td>0.688</td>\n",
       "      <td>-0.062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>0.875</td>\n",
       "      <td>1.000</td>\n",
       "      <td>-0.125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>0.938</td>\n",
       "      <td>1.000</td>\n",
       "      <td>-0.062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>0.812</td>\n",
       "      <td>0.625</td>\n",
       "      <td>0.188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>0.875</td>\n",
       "      <td>0.312</td>\n",
       "      <td>0.562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>0.938</td>\n",
       "      <td>0.875</td>\n",
       "      <td>0.062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>0.938</td>\n",
       "      <td>0.750</td>\n",
       "      <td>0.188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.625</td>\n",
       "      <td>0.375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>0.875</td>\n",
       "      <td>0.188</td>\n",
       "      <td>0.688</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>0.625</td>\n",
       "      <td>0.438</td>\n",
       "      <td>0.188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.438</td>\n",
       "      <td>0.562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>0.938</td>\n",
       "      <td>0.375</td>\n",
       "      <td>0.562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.188</td>\n",
       "      <td>0.812</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.938</td>\n",
       "      <td>0.062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>0.938</td>\n",
       "      <td>0.875</td>\n",
       "      <td>0.062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>0.875</td>\n",
       "      <td>0.250</td>\n",
       "      <td>0.625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.938</td>\n",
       "      <td>0.062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>0.750</td>\n",
       "      <td>0.500</td>\n",
       "      <td>0.250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>0.875</td>\n",
       "      <td>0.188</td>\n",
       "      <td>0.688</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>0.812</td>\n",
       "      <td>0.500</td>\n",
       "      <td>0.312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>154</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>0.812</td>\n",
       "      <td>0.250</td>\n",
       "      <td>0.562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>0.938</td>\n",
       "      <td>1.000</td>\n",
       "      <td>-0.062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.938</td>\n",
       "      <td>0.062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.938</td>\n",
       "      <td>0.062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>166</th>\n",
       "      <td>0.938</td>\n",
       "      <td>1.000</td>\n",
       "      <td>-0.062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>0.875</td>\n",
       "      <td>0.250</td>\n",
       "      <td>0.625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "mode         RD   SQL   diff\n",
       "worker_id                   \n",
       "0         0.938 0.312  0.625\n",
       "2         1.000 1.000  0.000\n",
       "5         1.000 1.000  0.000\n",
       "6         1.000 1.000  0.000\n",
       "10        1.000 0.875  0.125\n",
       "17        0.500 0.500  0.000\n",
       "21        1.000 1.000  0.000\n",
       "28        0.875 1.000 -0.125\n",
       "32        0.875 0.125  0.750\n",
       "39        1.000 0.812  0.188\n",
       "42        1.000 0.875  0.125\n",
       "43        1.000 0.750  0.250\n",
       "50        1.000 0.812  0.188\n",
       "52        0.938 0.688  0.250\n",
       "57        0.812 0.938 -0.125\n",
       "58        0.625 0.688 -0.062\n",
       "60        0.875 1.000 -0.125\n",
       "66        1.000 1.000  0.000\n",
       "72        0.938 1.000 -0.062\n",
       "75        0.812 0.625  0.188\n",
       "77        0.875 0.312  0.562\n",
       "80        0.938 0.875  0.062\n",
       "81        1.000 1.000  0.000\n",
       "83        0.938 0.750  0.188\n",
       "87        1.000 0.625  0.375\n",
       "89        0.875 0.188  0.688\n",
       "91        1.000 1.000  0.000\n",
       "92        0.625 0.438  0.188\n",
       "96        1.000 0.438  0.562\n",
       "110       0.938 0.375  0.562\n",
       "115       1.000 1.000  0.000\n",
       "117       1.000 0.125  0.875\n",
       "119       1.000 0.188  0.812\n",
       "121       1.000 0.938  0.062\n",
       "125       0.938 0.875  0.062\n",
       "130       1.000 1.000  0.000\n",
       "136       1.000 1.000  0.000\n",
       "141       0.875 0.250  0.625\n",
       "143       1.000 0.938  0.062\n",
       "146       0.750 0.500  0.250\n",
       "148       0.875 0.188  0.688\n",
       "153       0.812 0.500  0.312\n",
       "154       1.000 1.000  0.000\n",
       "158       0.812 0.250  0.562\n",
       "159       0.938 1.000 -0.062\n",
       "162       1.000 0.938  0.062\n",
       "165       1.000 0.938  0.062\n",
       "166       0.938 1.000 -0.062\n",
       "168       0.875 0.250  0.625\n",
       "169       1.000 1.000  0.000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "7/50 (0.140) better with SQL.\n",
      "30/50 (0.600) better with RD.\n",
      "13/50 (0.260) equally good.\n",
      "mean RD correct = 0.924, 95% CI [0.891, 0.953]\n",
      "mean SQL correct = 0.718, 95% CI [0.632, 0.800]\n",
      "mean difference in correctness = 0.206, 95% CI [0.131, 0.285]\n"
     ]
    }
   ],
   "source": [
    "# dfq4a: Create two columns mode and mean, with 2 rows per worker\n",
    "dfq4a = df_filtered_data.groupby(['worker_id', 'mode']).correct.agg(['mean'])\n",
    "dfq4a.reset_index(inplace=True)\n",
    "# display(dfq4a)\n",
    "\n",
    "# dfq4b: Pivot to have one row per worker\n",
    "dfq4b = pd.pivot_table(dfq4a, values=['mean'], index=['worker_id'], columns=['mode'])\n",
    "dfq4b=dfq4b.droplevel(0, axis=1)\n",
    "dfq4b['diff'] = dfq4b['RD'] - dfq4b['SQL']\n",
    "print(\"dfq4b:\\n\")\n",
    "display(dfq4b)\n",
    "\n",
    "# Calculate fraction of those better in either mode\n",
    "num_SQLbetter = np.where(dfq4b['SQL'] > dfq4b['RD'], 1, 0).sum()\n",
    "num_RDbetter = np.where(dfq4b['SQL'] < dfq4b['RD'], 1, 0).sum()\n",
    "num_workers =len(dfq4b)\n",
    "print(f'{num_SQLbetter}/{num_workers} ({num_SQLbetter/num_workers:.3f}) better with SQL.')\n",
    "print(f'{num_RDbetter}/{num_workers} ({num_RDbetter/num_workers:.3f}) better with RD.')\n",
    "print(f'{num_workers-num_RDbetter-num_SQLbetter}/{num_workers} ({(num_workers-num_RDbetter-num_SQLbetter)/num_workers:.3f}) equally good.')  # (7/6/2023): fixed: was incorrectly deducing RD twice, instead of num-RD-SQL\n",
    "\n",
    "# Mean of mean correctness for each mode Plus 95% CI\n",
    "modes_diff = ['RD', 'SQL', 'diff']\n",
    "mean_correct = {}\n",
    "ci = {}\n",
    "ci_delta = {}\n",
    "for mode in modes_diff:\n",
    "    mean_correct[mode] = np.mean(dfq4b[mode])\n",
    "    ci[mode] = scipybootstrap((dfq4b[mode],), statistic=np.mean, n_resamples=BOOTSTRAPSAMPLES, confidence_level=BOOTSTRAPCONFIDENCE, method='percentile', axis=0).confidence_interval        #convert array to sequence\n",
    "    ci_delta[mode] = [mean_correct[mode] - ci[mode].low, ci[mode].high - mean_correct[mode]]\n",
    "\n",
    "print(f\"mean RD correct = {mean_correct['RD']:.3f}, 95% CI [{ci['RD'].low:.3f}, {ci['RD'].high:.3f}]\")\n",
    "print(f\"mean SQL correct = {mean_correct['SQL']:.3f}, 95% CI [{ci['SQL'].low:.3f}, {ci['SQL'].high:.3f}]\")\n",
    "print(f\"mean difference in correctness = {mean_correct['diff']:.3f}, 95% CI [{ci['diff'].low:.3f}, {ci['diff'].high:.3f}]\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "c2506325",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 970x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Removing 'diff' from variables\n",
    "dfq4c=dfq4b.copy()\n",
    "dfq4c.drop('diff', inplace=True, axis=1)\n",
    "modes = ['RD', 'SQL']\n",
    "\n",
    "# Define pre-settings\n",
    "figwidth = 9.7\n",
    "figheight = 6\n",
    "xlab_size = 20\n",
    "ylab_size = 20\n",
    "figfont_size = 24\n",
    "\n",
    "# Define consistent color maps\n",
    "my_cmap_sns_light = [(0.9921568627450981, 0.8156862745098039, 0.6352941176470588), (0.7764705882352941, 0.8588235294117647, 0.9372549019607843)]        # light blue, light orange\n",
    "my_cmap_sns_dark = [(0.9019607843137255, 0.3333333333333333, 0.050980392156862744), (0.19215686274509805, 0.5098039215686274, 0.7411764705882353)]      # dark blue, dark orange\n",
    "my_cmap_dark = sns.color_palette(my_cmap_sns_dark, as_cmap=True)\n",
    "my_cmap_light = sns.color_palette(my_cmap_sns_light, as_cmap=True)\n",
    "\n",
    "# Create empty figure and plot the individual datapoints\n",
    "fig, ax = plt.subplots(figsize=(figwidth,figheight))\n",
    "\n",
    "# 1. Violinplots\n",
    "axsns = sns.violinplot(x='mean', y='mode', data=dfq4a,\n",
    "                       hue=True, hue_order=[False, True], split=True,   # half violinplots https://stackoverflow.com/questions/53872439/half-not-split-violin-plots-in-seaborn\n",
    "                       inner='quartile',\n",
    "                       cut=0,               # 0 means ending sharp at end points\n",
    "                       width=.6,\n",
    "                       orient = 'h',\n",
    "                       zorder=20,)\n",
    "\n",
    "# change the medium default line to full (https://stackoverflow.com/questions/60638344/quartiles-line-properties-in-seaborn-violinplot)\n",
    "for l in axsns.lines[1::3]:\n",
    "    l.set_linestyle('-')\n",
    "    l.set_linewidth(1.2)\n",
    "    l.set_color('black')\n",
    "    l.set_alpha(0.8)\n",
    "for i in [0, 2, 3, 5]:         # remove the 25% and 75% quartiles\n",
    "    l = axsns.lines[i]\n",
    "    l.set_linestyle('-')\n",
    "    l.set_linewidth(0)\n",
    "    l.set_color('red')\n",
    "    l.set_alpha(0.8)\n",
    "\n",
    "# Apply colorscheme to violinplots https://stackoverflow.com/questions/70442958/seaborn-how-to-apply-custom-color-to-each-seaborn-violinplot\n",
    "from matplotlib.collections import PolyCollection\n",
    "for ind, violin in enumerate(axsns.findobj(PolyCollection)):\n",
    "    violin.set_facecolor(my_cmap_light[ind])\n",
    "# plt.setp(ax.collections, alpha=.999)  # semi-transparent (https://stackoverflow.com/questions/62597959/seaborn-violinplot-transparency)\n",
    "\n",
    "\n",
    "# 2. Plot individual points [new dot plot]\n",
    "y_tilt = -0.5       # how far below each violinplot\n",
    "\n",
    "def dotplot(input_x, y0, delta, **args):\n",
    "    unique_values, counts = np.unique(input_x, return_counts=True)  # Count how many times does each value occur\n",
    "\n",
    "    # Convert 1D input into 2D array\n",
    "    scatter_x = [] # x values\n",
    "    scatter_y = [] # corresponding y values\n",
    "    for idx, value in enumerate(unique_values):\n",
    "        for counter in range(0, counts[idx]):\n",
    "            scatter_x.append(value)\n",
    "            scatter_y.append(y0+counter*delta)\n",
    "    plt.scatter(scatter_x, scatter_y, **args)\n",
    "\n",
    "for i, col in enumerate(dfq4c):\n",
    "    dotplot(input_x=dfq4c[col], y0=y_tilt + i, delta=0.03,      # y-axis tilt change with each column\n",
    "            marker='^',\n",
    "            alpha=1,\n",
    "            zorder=20,      # higher means more visible\n",
    "            color=my_cmap_dark[i],\n",
    "            s=150,\n",
    "            linewidth=0,)\n",
    "\n",
    "\n",
    "# 4. Plot red line connecting means\n",
    "ax.plot(np.mean(dfq4c, axis=0), [0, 1], color ='red', linewidth = 2, linestyle ='-',\n",
    "        alpha = .3,\n",
    "        zorder=4,\n",
    "        )\n",
    "\n",
    "# 5. CI Errorbars\n",
    "for i, mode in enumerate(modes):\n",
    "    plt.errorbar(mean_correct[mode], i, xerr=np.array([[ci_delta[mode][0], ci_delta[mode][1]]]).T,\n",
    "                 fmt='o', markersize=10,\n",
    "                 lw = 5, alpha=1,\n",
    "                 zorder=100,        # higher means more visible\n",
    "                 capsize = 0,\n",
    "                 # capthick = 4,\n",
    "                 capthick = 0,\n",
    "                 # color = 'black',\n",
    "                 color=my_cmap_dark[i],\n",
    "                 )   # my_cmap[1])\n",
    "\n",
    "    # ax.text(median_time[column], y_tilt_mode[i]+y_tilt_section[j] + 0.18, f'{median_time[column]:.1f}',\n",
    "    #         horizontalalignment='center', color = my_cmap_sns_dark[i], fontsize=20)\n",
    "    ax.text(mean_correct[mode],\n",
    "            i-0.2,\n",
    "            # f'{mean_correct[mode]:.2f}', horizontalalignment='center',\n",
    "            f'{100*mean_correct[mode]:.0f}%', horizontalalignment='center',\n",
    "            # color='black',\n",
    "            color=my_cmap_dark[i],\n",
    "            fontsize=figfont_size)\n",
    "    # ax.text(ci[mode].low, i+0.1, f'{ci[mode].low:.2f}', horizontalalignment='center', color='black', fontsize=20)\n",
    "    # ax.text(ci[mode].high, i+0.1, f'{ci[mode].high:.2f}', horizontalalignment='center', color='black', fontsize=20)\n",
    "\n",
    "\n",
    "#Additional settings\n",
    "# ax.set_yticklabels(modes, size= ylab_size)\n",
    "# ax.set_xlim(0.7499, 1.0003)\n",
    "# ax.set_xlim(0.74, 1.01)\n",
    "ax.set_xlim(0.1, 1.01)\n",
    "if VARIANT == 1:\n",
    "    ax.set_xticks(np.linspace(0.1, 1, num=10))\n",
    "if VARIANT == 3:\n",
    "    ax.set_xlim(0.8, 1.01)\n",
    "if VARIANT == 4:\n",
    "    ax.set_xlim(0.4, 1.01)\n",
    "ax.set_ylim(-0.6, 1.5)\n",
    "ax.set_xlabel('Mean accuracy per worker', size = xlab_size)\n",
    "ax.set_ylabel(None)\n",
    "ax.set_yticklabels(['DX', 'SQL'])\n",
    "ax.legend_.remove()\n",
    "plt.grid(axis = 'x', linewidth = 0.5, color = 'lightgray')\n",
    "sns.despine()               # remove bounding box\n",
    "\n",
    "# import matplotlib.ticker as mtick\n",
    "ax.xaxis.set_major_formatter(mtick.PercentFormatter(1.0))           # show in percentage\n",
    "\n",
    "if savefig:\n",
    "    plt.savefig(fig_dir + f'/q4_figure_variant{VARIANT}.pdf', bbox_inches='tight')\n",
    "    plt.savefig(fig_dir + f'/q4_figure_variant{VARIANT}.svg', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e212f759",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "# Question 4. Figure 4b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "dc15e9e6",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Define figure settings\n",
    "figwidth = 10\n",
    "figheight = 3\n",
    "xlab_size = 20\n",
    "ylab_size = 20\n",
    "figfont_size = 24\n",
    "\n",
    "# Define consistent color maps\n",
    "my_cmap_sns_dark = [(0.8392156862745098, 0.15294117647058825, 0.1568627450980392)]\n",
    "my_cmap_sns_light = [(0.984313725490196, 0.6039215686274509, 0.6)]\n",
    "my_cmap_dark = sns.color_palette(my_cmap_sns_dark, as_cmap=True)\n",
    "my_cmap_light = sns.color_palette(my_cmap_sns_light, as_cmap=True)\n",
    "\n",
    "# Create data frame for split violinplot\n",
    "sample = np.array(dfq4b['diff'])                                                     # extract the sample and then create the boostrapped medians\n",
    "data_difference = dfq4b['diff']\n",
    "dfvp = pd.DataFrame()\n",
    "dfvp[\"values\"] = sample\n",
    "dfvp[\"all\"] = \"\"                                        # attribute that is shared by all entries\n",
    "\n",
    "# Create empty figure and plot the individual datapoints\n",
    "fig, ax = plt.subplots(figsize=(figwidth,figheight))\n",
    "\n",
    "\n",
    "# 1. Violinplot\n",
    "axsns = sns.violinplot(x='values', y='all',    # y='all' just need to group both types together\n",
    "                       data=dfvp,\n",
    "                       hue = True, hue_order = [False, True],\n",
    "                       split = True, inner = 'quartile',\n",
    "                       cut=0,                  # 0 means ending sharp at end points\n",
    "                       width=.6, scale = 'width',\n",
    "                       # dodge = False,        # When using ``hue`` nesting, setting this to ``True`` will separate the strips for different hue levels along the categorical axis.\n",
    "                       orient = 'h',\n",
    "                       color=my_cmap_light[0],\n",
    "                       zorder = 3)\n",
    "\n",
    "# change the medium default linke to full\n",
    "for l in axsns.lines[1::3]:\n",
    "    l.set_linestyle('-')\n",
    "    l.set_linewidth(1.5)\n",
    "    l.set_color('black')\n",
    "    l.set_alpha(0.8)\n",
    "\n",
    "\n",
    "# 2. Plot individual points [new dot plot]\n",
    "y_tilt = -0.3                                      # Set some delta for the points below the violinplot\n",
    "\n",
    "def dotplot(input_x, y0, delta, **args):\n",
    "    unique_values, counts = np.unique(input_x, return_counts=True)  # Count how many times does each value occur\n",
    "\n",
    "    # Convert 1D input into 2D array\n",
    "    scatter_x = [] # x values\n",
    "    scatter_y = [] # corresponding y values\n",
    "    for idx, value in enumerate(unique_values):\n",
    "        for counter in range(0, counts[idx]):\n",
    "            scatter_x.append(value)\n",
    "            scatter_y.append(y0+counter*delta)\n",
    "    plt.scatter(scatter_x, scatter_y, **args)\n",
    "\n",
    "dotplot(input_x=data_difference, y0=y_tilt, delta=0.02,\n",
    "        marker='^',\n",
    "        alpha=1,\n",
    "        zorder=20,      # higher means more visible\n",
    "        color=my_cmap_dark[0],\n",
    "        s=150,\n",
    "        linewidth=0,)\n",
    "\n",
    "\n",
    "# 3. CI Errorbars & show numbers\n",
    "# axeb = plt.errorbar(median_ratio, 0, xerr=np.array([[ci_ratio_delta[0], ci_ratio_delta[1]]]).T,\n",
    "axeb = plt.errorbar(mean_correct['diff'], 0, xerr=np.array([[ci_delta['diff'][0], ci_delta['diff'][1]]]).T,\n",
    "                    fmt='o',\n",
    "                    markersize=10, alpha=1,\n",
    "                    # lw = 3,\n",
    "                    lw = 5,\n",
    "                    zorder=100,      # higher means more visible\n",
    "                    capsize = 0,     # 10\n",
    "                    # capthick = 4,\n",
    "                    capthick = 0,\n",
    "                    # color = 'black',\n",
    "                    color = my_cmap_dark[0],\n",
    "                    )\n",
    "\n",
    "\n",
    "\n",
    "# 4. vertical bar for x-axis = 1\n",
    "plt.plot([0, 0], [-10, 10], color = 'black', zorder = 0, linewidth=2)\n",
    "\n",
    "\n",
    "meandiff = np.mean(sample)           # rename\n",
    "# ax.text(meandiff, -0.12, f'{meandiff:.2f}', horizontalalignment='center',\n",
    "ax.text(meandiff, -0.12, f'{100*meandiff:.0f}%', horizontalalignment='center',\n",
    "        # color='black',\n",
    "        color = my_cmap_dark[0],\n",
    "        fontsize=figfont_size)\n",
    "# ax.text(ci_ratio.low, 0.04, f'{100*ci_ratio.low:.1f}%', horizontalalignment='center', color='black', fontsize=20)\n",
    "# ax.text(ci_ratio.high, 0.04, f'{100*ci_ratio.high:.1f}%', horizontalalignment='center', color='black', fontsize=20)\n",
    "\n",
    "\n",
    "\n",
    "# Additional settings\n",
    "ax.set_ylim(-0.4, 0.35)\n",
    "ax.set_ylabel(None)         # remove the 'all'\n",
    "\n",
    "ax.set_xticks(np.linspace(-1, 1, num=11))\n",
    "# ax.set_xticks(np.linspace(-1, 1, num=21), minor=True)\n",
    "ax.set_xlim(-0.201, 0.901)\n",
    "\n",
    "if VARIANT == 3:\n",
    "    ax.set_xticks(np.linspace(-0.2, 0.2, num=9))\n",
    "    ax.set_xlim(-0.151, 0.201)\n",
    "if VARIANT == 4:\n",
    "    ax.set_xlim(-0.201, 0.601)\n",
    "ax.set_xlabel('Difference in accuracy per worker (DX - SQL)', size = xlab_size)\n",
    "\n",
    "sns.despine(trim=False, left=True)               # remove bounding box\n",
    "plt.grid(axis = 'x', linewidth = 0.5, color = 'lightgray', which='both')\n",
    "ax.legend_.remove()\n",
    "ax.xaxis.set_major_formatter(mtick.PercentFormatter(1.0, decimals=0), )           # show in percentage\n",
    "\n",
    "if savefig:\n",
    "    plt.savefig(fig_dir + f'/q4_figure2_variant{VARIANT}.pdf', bbox_inches='tight')\n",
    "    plt.savefig(fig_dir + f'/q4_figure2_variant{VARIANT}.svg', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5489ff85",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "# Print user feedback\n",
    "\n",
    "Prints all comments received from participants who passed the requirements (0.5 correctness)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "f9864109",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0: \"I found the tutorial to be very helpful in understanding the queries. In particular, the side-by-side examples were very helpful. \"\n",
      "1: \"Good\"\n",
      "2: \"good little bit in deficit in the coding\"\n",
      "3: \"the instructions are quite clear and easy to follow, the diagrams are easy to understand and questions are not very difficult. And the tutorials especially the last summary page is quite helpful. I think the diagrams can be applied to my own usage but I am not quite sure now. \"\n",
      "4: \"risky \"\n",
      "5: \"It's very understanding and easy to play, also interesting too.\"\n",
      "6: \"good survey . i really love it. this is very helpful for some database studies. \"\n",
      "7: \"the example was very helpful\n",
      "None\"\n",
      "8: \"GOOD\"\n",
      "9: \"nan\"\n",
      "10: \"Really very nice survey I enjoyed lot to do SQL very easy. Thank you so much. I m very Happy to complete This survey.\"\n",
      "11: \"It is easy and creates curiosity while answering. It's all good.\"\n",
      "12: \"nan\"\n",
      "13: \"I will learn more and feel good for contain this survey.\"\n",
      "14: \"GOOD\"\n",
      "15: \"Hard \"\n",
      "16: \"It is easy and very curious to participate.\"\n",
      "17: \"none\"\n",
      "18: \"The SQL queries was understanding, No that's are good, it's most useful of query patterns.\"\n",
      "19: \"1.your diagrams and explanation are awesome to understand.\n",
      "2.All of your queries and content are interesting that just like daily activities.\n",
      "3.I fix my mid itself some logical pattern to do your hit  fast .\n",
      "3. your organizational chart based explanation is very useful. Diagram easily understand compare to text.\"\n",
      "20: \"nan\"\n",
      "21: \"nan\"\n",
      "22: \"All Diagrams of SQL queries helpful for understanding the queries.\"\n",
      "23: \"Yes, the diagrams were helpful. It was very easy to understand both code and diagram. I found the exercise to be very straightforward and understandable. I noted the visual patterns in the diagrams, and the syntax patterns in the code. The last page of the tutorial with a summary of the diagrams was most helpful. I don't know how these diagrams would be helpful in real life. I can't imagine a scenario wherein that would be true.\"\n",
      "24: \"none\"\n",
      "25: \"Very hard \"\n",
      "26: \"nan\"\n",
      "27: \"good\"\n",
      "28: \"NONE\"\n",
      "29: \"I found the Diagram of the SQL to be extremely helpful in determining the answer. \n",
      "\n",
      "The overview of the four query patterns I liked.\n",
      "\n",
      "The Diagrams I think were much easier to understand than the SQL.\n",
      "\n",
      "After answering a few questions I felt I got the patterns down quicker. The diagrams felt much easier to answer than the SQL questions.\n",
      "\n",
      "Overview of the four patterns was helpful.\n",
      "\n",
      "Yes.\"\n",
      "30: \"your SQl queries very helpful that has presented with daily activities.\n",
      "Your diagrams awesome that has great explanation to understand even beginner.\n",
      "I believe your both pattern useful.\n",
      "Tutorials helped very much for my knowledge.\n",
      "It is interesting task to do in future also\"\n",
      "31: \"NONE\"\n",
      "32: \"NONE\"\n",
      "33: \"The diagrams were very easy to understand. The dotted lines helped group the queries. \"\n",
      "34: \"It was great study for me.\"\n",
      "35: \"The instructions/diagrams were quite clear.\n",
      "The logical patterns became more apparent with practice, even making mistakes helped firm up future comprehension.\n",
      "\"\n",
      "36: \"nan\"\n",
      "37: \"none\"\n",
      "38: \"good survey. i really enjoyed it . i know some database knowledge from this study.\"\n",
      "39: \"It's very interesting to participate and easy to understand.\"\n",
      "40: \"I found it helpful when using the guide. They were all useful. I could see the pattern. The examples helped the most. Thank you.\"\n",
      "41: \"The SQL  PDF really help to me to do this survey.\n",
      "It's really understandable.\n",
      "First 5 question I unable to understand after that I gave all answer correctly. I take some practice and time.\n",
      "Thank You!!\"\n",
      "42: \"This survey understandable I adapted quickly.\n",
      "Instruction PDF clearly explained\n",
      "I suggest Add extra 3 question to practice to make any mistake to do complete this task\"\n",
      "43: \"good\"\n",
      "44: \"I found your diagrams very helpful in understanding the queries. At first I didn't get it, but after staring at the diagrams for a few minutes it clicked and everything became super simple. I saw the patterns and it became just looking for the correct pattern to know which query was being used.\"\n",
      "45: \"good\"\n",
      "46: \"Hard one \"\n",
      "47: \"nan\"\n",
      "48: \"good\"\n",
      "49: \"nan\"\n"
     ]
    }
   ],
   "source": [
    "feedback = dfresults.loc[dfresults.worker_id.isin(dftemp.index),'feedback']\n",
    "for i, text in enumerate(feedback):\n",
    "    print(f'{i}: \"{text}\"')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "b6e5b507",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "# end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "fc467f89",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    },
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "# end"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}