notebook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "ca95b9b4",
   "metadata": {},
   "source": [
    "# Notebook - Statistical Arbitrage Example\n",
    "\n",
    "\n",
    "*Notebook for the paper [Reinforcement Learning with Dynamic Convex Risk Measures](https://arxiv.org/abs/2112.13414), by [Anthony Coache](https://anthonycoache.ca/) and [Sebastian Jaimungal](http://sebastian.statistics.utoronto.ca/).*\n",
    "\n",
    "This code notebook showcases how to use the Python files in order to reproduce some of the main figures of the statistical arbitrage example. Note that the code architecture is similar between all sets of experiments, and the code for each example is available in separate folders on our [Github repository](https://github.com/acoache/RL-DynamicConvexRisk):\n",
    "- **StatArbitrage**: Statistical Arbitrage Example\n",
    "- **CliffWalking**: Cliff Walking Example\n",
    "- **Hedging**: Hedging with Friction Example"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c2dfeeb9",
   "metadata": {},
   "source": [
    "## 1. Importation of Python Libraries\n",
    "\n",
    "First, our code requires us to import several Python packages, such as NumPy, PyTorch, SciPy, Matplotlib, os, time, as well as some of our classes as modules. Note here that we would need to change the prefix of the modules for the folder name of the corresponding example we want to reproduce."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "25693973",
   "metadata": {},
   "outputs": [],
   "source": [
    "# numpy\n",
    "import numpy as np\n",
    "import numpy.matlib\n",
    "\n",
    "# plotting\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.stats import gaussian_kde\n",
    "from matplotlib.colors import LinearSegmentedColormap\n",
    "\n",
    "# pytorch\n",
    "import torch as T\n",
    "import torch.optim as optim\n",
    "\n",
    "# personal files\n",
    "import StatArbitrage.utils as utils\n",
    "import StatArbitrage.hyperparams as hyperparams\n",
    "from StatArbitrage.models import PolicyApprox, ValueApprox\n",
    "from StatArbitrage.risk_measure import RiskMeasure\n",
    "from StatArbitrage.envs import TradingEnv\n",
    "from StatArbitrage.actor_critic import ActorCriticPG\n",
    "\n",
    "# misc\n",
    "import time\n",
    "import os"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19284291",
   "metadata": {},
   "source": [
    "## 2. Initialization of Training Parameters\n",
    "\n",
    "This section deals with the initialization of all hyperparameters, located in the **hyperparams.py** file. We can set the name of the repository, parameters that modify the environment -- e.g. number of periods, asset price dynamics model, etc. -- and hyperparameters that affect our actor-critic algorithm -- e.g. number of epochs, learning rates, etc. We also print the parameters prior to the training in order to verify their correctness."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f307553d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "*** Name of the repository:  StatArbitrage_ex1  ***\n",
      "\n",
      "*  T:  1  Ndt:  6  kappa:  2  sigma:  0.2  theta:  1  phi:  0.005  psi:  0.5  max_q:  5  max_u:  2\n",
      "*  Ntrajectories:  500  Mtransitions:  500  Nepochs:  300  Nsims_optimal:  1000\n",
      "*  Nepochs_V_init:  500  Nepochs_V:  50  lr_V:  0.001  batch_V:  200  hidden_V:  32  layers_V:  4\n",
      "*  Nepochs_pi:  10  lr_pi:  0.001  batch_pi:  200  hidden_pi:  32  layers_pi:  3\n"
     ]
    }
   ],
   "source": [
    "# initialize parameters for the model and algorithm\n",
    "repo, envParams, algoParams = hyperparams.initParams()\n",
    "\n",
    "# print all parameters for reproducibility purposes\n",
    "print('\\n*** Name of the repository: ', repo, ' ***\\n')\n",
    "hyperparams.printParams(envParams, algoParams)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7f67070",
   "metadata": {},
   "source": [
    "We also specify the risk measures that we want to evaluate. For instance, here, we will perform the optimization on three risk measures:\n",
    "\n",
    "1. the expectation;\n",
    "2. the dynamic CVaR at level $\\alpha=0.5$;\n",
    "3. the dynamic CVaR at level $\\alpha=0.2$.\n",
    "\n",
    "One can also add other (dynamic) risk measures in the **risk_measure.py** file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "24d78cc9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# specify the risk measures\n",
    "rm_list = ['mean', 'CVaR', 'CVaR']\n",
    "alpha_cvar = [-99, 0.5, 0.2] # threshold for the conditional value-at-risk\n",
    "\n",
    "print_progress = 200 # number of epochs before printing the time/loss\n",
    "\n",
    "utils.directory(repo) # creation of the main repository to store results"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d81918ce",
   "metadata": {},
   "source": [
    "We initialize the environment object, that is common to all optimization problems. Finally, we define labels for all risk measures that we will use later, whether for folder names when storing results or legends when plotting figures."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "3df93e21",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['mean', 'CVaR0.5', 'CVaR0.2']\n"
     ]
    }
   ],
   "source": [
    "# create the environment and risk measure objects\n",
    "env = TradingEnv(envParams)\n",
    "\n",
    "# labels to use for the figures in the testing phase\n",
    "rm_labels = []\n",
    "\n",
    "for idx_method, method in enumerate(rm_list):\n",
    "    \n",
    "    if(method == 'mean'):\n",
    "        rm_label = method\n",
    "    if(method == 'CVaR'):\n",
    "        rm_label = method + str( round(alpha_cvar[idx_method],3) )\n",
    "    \n",
    "    rm_labels.append(rm_label)\n",
    "    \n",
    "    # create a subfolder to store results\n",
    "    utils.directory(repo + '/' + rm_label)\n",
    "\n",
    "# print labels\n",
    "print(rm_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f996821b",
   "metadata": {},
   "source": [
    "## 3. Training our Agents\n",
    "\n",
    "Once we validate all the parameters, we can start the training phase of our actor-critic algorithm. This corresponds to the Algorithms 2 and 3 presented in Section 6 of our [paper](https://arxiv.org/abs/2112.13414). We loop through all the risk measures and perform the optimization by alternating between\n",
    "\n",
    "1. Estimating the value function of the current policy;\n",
    "2. Updating the policy by policy gradient.\n",
    "\n",
    "When the training is completed, we save the ANNs for both the policy and the value function in the corresponding subfolder. In the corresponding folders, one can also print progress in the console and plot some diagnostic figures during the training phase."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "778e13c2",
   "metadata": {},
   "outputs": [],
   "source": [
    "for idx_method, method in enumerate(rm_list):\n",
    "    # print progress\n",
    "    print('\\n*** Method = ', method, ' ***\\n')\n",
    "    start_time = time.time()\n",
    "\n",
    "    # create the risk measure object\n",
    "    risk_measure = RiskMeasure(Type=method, alpha=alpha_cvar[idx_method])\n",
    "\n",
    "    # create ANNs for both the policy & value function\n",
    "    policy = PolicyApprox(3, env,\n",
    "                            n_layers=algoParams[\"layers_pi\"],\n",
    "                            hidden_size=algoParams[\"hidden_pi\"],\n",
    "                            learn_rate=algoParams[\"lr_pi\"])\n",
    "    value_function = ValueApprox(3, env,\n",
    "                            n_layers=algoParams[\"layers_V\"],\n",
    "                            hidden_size=algoParams[\"hidden_V\"],\n",
    "                            learn_rate=algoParams[\"lr_V\"])\n",
    "    \n",
    "    # initialize the actor-critic algorithm\n",
    "    actor_critic = ActorCriticPG(repo=repo,\n",
    "                                    method=rm_labels[idx_method],\n",
    "                                    env=env,\n",
    "                                    policy=policy,\n",
    "                                    V=value_function,\n",
    "                                    risk_measure=risk_measure,\n",
    "                                    gamma=algoParams[\"gamma\"])\n",
    "    \n",
    "    ##############################################################################\n",
    "    ###########################     TRAINING PHASE     ###########################\n",
    "    \n",
    "    # first estimate of the value function\n",
    "    actor_critic.estimate_V(Ntrajectories=algoParams[\"Ntrajectories\"],\n",
    "                                Mtransitions=algoParams[\"Mtransitions\"],\n",
    "                                batch_size=algoParams[\"batch_V\"],\n",
    "                                Nepochs=algoParams[\"Nepochs_V_init\"],\n",
    "                                rng_seed=algoParams[\"seed\"])\n",
    "\n",
    "    for epoch in range(algoParams[\"Nepochs\"]):\n",
    "        # estimate the value function of the current policy\n",
    "        actor_critic.estimate_V(Ntrajectories=algoParams[\"Ntrajectories\"],\n",
    "                                    Mtransitions=algoParams[\"Mtransitions\"],\n",
    "                                    batch_size=algoParams[\"batch_V\"],\n",
    "                                    Nepochs=algoParams[\"Nepochs_V\"],\n",
    "                                    rng_seed=algoParams[\"seed\"])\n",
    "        \n",
    "        # update the policy by policy gradient\n",
    "        actor_critic.update_policy(Ntrajectories=algoParams[\"Ntrajectories\"],\n",
    "                                    Mtransitions=algoParams[\"Mtransitions\"],\n",
    "                                    batch_size=algoParams[\"batch_pi\"],\n",
    "                                    Nepochs=algoParams[\"Nepochs_pi\"],\n",
    "                                    rng_seed=algoParams[\"seed\"])\n",
    "\n",
    "        # print progress\n",
    "        if epoch % print_progress == 0 or epoch == algoParams[\"Nepochs\"] - 1:\n",
    "            print('*** Epoch = ', str(epoch) ,\n",
    "                    ' completed, Duration = ', \"{:.3f}\".format(time.time() - start_time), ' secs ***')\n",
    "            start_time = time.time()\n",
    "\n",
    "    # save the neural network\n",
    "    T.save(actor_critic.policy.state_dict(),\n",
    "            repo + '/' + rm_labels[idx_method] + '/policy_model.pt')\n",
    "    T.save(actor_critic.V.state_dict(),\n",
    "            repo + '/' + rm_labels[idx_method] + '/V_model.pt')\n",
    "\n",
    "    # print progress\n",
    "    print('*** Training phase completed! ***')\n",
    "    \n",
    "    ###########################     END OF TRAINING     ##########################\n",
    "    ##############################################################################"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f9be93cb",
   "metadata": {},
   "source": [
    "## 4. Initialization of Testing Parameters\n",
    "\n",
    "This section allows us to set the number of trajectories, the seed for the random generator of both NumPy and PyTorch, and the style of the figures created in the testing phase. We perform the training and testing of our agents on the same environment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "cebc6d0c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# testing phase parameters\n",
    "Nsimulations = 30000 # number of simulations following the optimal strategy\n",
    "seed = 1234 # set seed for replication purposes\n",
    "\n",
    "# font sizes for figures\n",
    "plt.rcParams.update({'font.size': 16})\n",
    "plt.rc('axes', labelsize=20)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae10e7f8",
   "metadata": {},
   "source": [
    "We next instantiate objects to store results of the testing phase. Note here that we have placeholders for the policy $\\pi^{\\theta}$ and the value function $V^{\\phi}$, but we will load the trained model when creating figures."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "f958d1e0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# matrix to store all testing trajectories\n",
    "costs = np.zeros((Nsimulations, envParams[\"Ndt\"], len(rm_list)))\n",
    "\n",
    "# create the risk measure object\n",
    "risk_measure = RiskMeasure(Type='mean', alpha=-99)\n",
    "\n",
    "# create ANNs for both the policy & value function\n",
    "policy = PolicyApprox(3, env,\n",
    "                        n_layers=algoParams[\"layers_pi\"],\n",
    "                        hidden_size=algoParams[\"hidden_pi\"],\n",
    "                        learn_rate=algoParams[\"lr_pi\"])\n",
    "value_function = ValueApprox(3, env,\n",
    "                        n_layers=algoParams[\"layers_V\"],\n",
    "                        hidden_size=algoParams[\"hidden_V\"],\n",
    "                        learn_rate=algoParams[\"lr_V\"])\n",
    "    \n",
    "# initialize the actor-critic algorithm\n",
    "actor_critic = ActorCriticPG(repo=repo,\n",
    "                                method=rm_labels[idx_method],\n",
    "                                env=env,\n",
    "                                policy=policy,\n",
    "                                V=value_function,\n",
    "                                risk_measure=risk_measure,\n",
    "                                gamma=algoParams[\"gamma\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0b3022da",
   "metadata": {},
   "source": [
    "## 5. Testing our Agents\n",
    "\n",
    "All our agents were trained with different dynamic risk measures, and now is the time to compare their optimal strategies. In this first part of the testing phase, we load the trained models, simulate trajectories using the best policy, and store the costs and the states observed. While we have the optimal policy, we also make sure to plot what the optimal behavior look like, for instance with a heatmap."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "145f03fe",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "*** Method =  mean  ***\n",
      "\n",
      "*** Loading trained model completed! ***\n",
      "*** Testing phase completed! ***\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "*** Method =  CVaR  ***\n",
      "\n",
      "*** Loading trained model completed! ***\n",
      "*** Testing phase completed! ***\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "*** Method =  CVaR  ***\n",
      "\n",
      "*** Loading trained model completed! ***\n",
      "*** Testing phase completed! ***\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for idx_method, method in enumerate(rm_list):\n",
    "    # print progress\n",
    "    print('\\n*** Method = ', method, ' ***\\n')\n",
    "    start_time = time.time()\n",
    "    \n",
    "    # load the trained model\n",
    "    actor_critic.policy.load_state_dict(T.load(repo + '/' + rm_labels[idx_method] + '/policy_model.pt'))\n",
    "    actor_critic.V.load_state_dict(T.load(repo + '/' + rm_labels[idx_method] + '/V_model.pt'))\n",
    "\n",
    "    # print progress\n",
    "    print('*** Loading trained model completed! ***')\n",
    "\n",
    "    ##############################################################################\n",
    "    ###########################     TESTING PHASE     ############################\n",
    "    \n",
    "    # set seed for reproducibility purposes\n",
    "    T.manual_seed(seed)\n",
    "    np.random.seed(seed)\n",
    "\n",
    "    # initialize the starting state\n",
    "    s, q = env.reset(Nsimulations)\n",
    "    \n",
    "    for timestep in env.spaces[\"t_space\"][:-1]:\n",
    "        # simulate transitions according to the policy\n",
    "        u, _ = actor_critic.select_actions(s, q, timestep*T.ones(Nsimulations), 'best')\n",
    "        s, q, cost = env.step(s, q, u)\n",
    "\n",
    "        # store costs\n",
    "        costs[:,timestep,idx_method] = cost.detach().numpy()\n",
    "\n",
    "    # get terminal reward\n",
    "    costs[:,-1,idx_method] = env.get_final_cost(s, q).detach().numpy()\n",
    "    \n",
    "    # print progress\n",
    "    print('*** Testing phase completed! ***')\n",
    "    \n",
    "    ###########################     END OF TESTING     ###########################\n",
    "    ##############################################################################\n",
    "    \n",
    "    ##############################################################################\n",
    "    ###########################     PLOT: OPTIMAL POLICY     #####################\n",
    "    plt.figure(figsize=(10,4))\n",
    "    for time_idx, time_val in enumerate([0, 2, 4]):\n",
    "        # allocate the subplot\n",
    "        plt.subplot(1,3,time_idx+1)\n",
    "        \n",
    "        # initialize 2D histogram\n",
    "        hist2dim_pi = np.zeros([len(env.spaces[\"s_space\"]), len(env.spaces[\"q_space\"])])\n",
    "\n",
    "        for s_idx, s_val in enumerate(env.spaces[\"s_space\"]):\n",
    "            for q_idx, q_val in enumerate(env.spaces[\"q_space\"]):\n",
    "                # best action according to the policy\n",
    "                hist2dim_pi[len(env.spaces[\"s_space\"])-s_idx-1, q_idx], _ = \\\n",
    "                        actor_critic.select_actions(T.Tensor([s_val]).to(actor_critic.device),\n",
    "                                                    T.Tensor([q_val]).to(actor_critic.device),\n",
    "                                                    T.tensor([time_val]).to(actor_critic.device),\n",
    "                                                    'best')\n",
    "\n",
    "        # plot the 2D histogram\n",
    "        plt.imshow(hist2dim_pi,\n",
    "                interpolation='none',\n",
    "                cmap=utils.cmap,\n",
    "                extent=[np.min(env.spaces[\"q_space\"]),\n",
    "                        np.max(env.spaces[\"q_space\"]),\n",
    "                        np.min(env.spaces[\"s_space\"]),\n",
    "                        np.max(env.spaces[\"s_space\"])],\n",
    "                aspect='auto',\n",
    "                vmin=-env.params[\"max_u\"],\n",
    "                vmax=env.params[\"max_u\"])\n",
    "        \n",
    "        plt.title('Learned; Period:' + str(time_val+1))\n",
    "        plt.xlabel(\"Inventory\")\n",
    "        plt.ylabel(\"Price\")\n",
    "        plt.colorbar()\n",
    "        plt.tight_layout()\n",
    "    \n",
    "    plt.show()\n",
    "    \n",
    "    ###########################     PLOT: OPTIMAL POLICY     #####################\n",
    "    ##############################################################################"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "661feb7c",
   "metadata": {},
   "source": [
    "## 6. Plotting Additional Figures\n",
    "\n",
    "In the second part of the testing phase, we now create figures that compare all agents at the same time. In the statistical arbitrage example, a plot of interest is the distribution of terminal wealth when following the optimal policy for all dynamic risk measures."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a320b6d4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArsAAAELCAYAAADZQU5qAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAB1DUlEQVR4nO3dd3wb5f3A8c9Xy/LedjYJgYSwRyBQVkpYZaSlUFp+jDBaCoUWyqalQIEWWspqS5kFyqaFMlJoygw0ECChzAxCyI4zvKdsazy/P+6kyLZsSx6SbX3fr5dyp+eeu/vKF52+evTcc2KMQSmllFJKqZHIkeoAlFJKKaWUGiya7CqllFJKqRFLk12llFJKKTViabKrlFJKKaVGLE12lVJKKaXUiKXJrlJKKaWUGrE02VUqSUTkTBExInJmCvY909739Z3KjYjMT3Y8UfufLyLDdvxDETlWRD4QkQb7b3lnP7YV8xip2FL1fkr1e2ag9OX/WyrPYUr1hya7SiVARCbaJ/voR7OIbBSR10TkGhEZN0j7HnYfsiJyvR33zFTHMtBEZDLwT2A8cD/wa2BeL+usEZE1gx9d/KL+Tz+S6lhU6g3H84xSvXGlOgClhqmvgCfteS8wCjgAuBG4RkSuNsbc0Wmd54H3gU1Ji3KbD4FpQFUK9t2TM4CsVAfRR4cBHuASY8zTqQ4mDaXy/aSUGkY02VWqb1YYY67vXCgixwIPA7eLSJMx5oHwMmNMPVCfvBC3Mca0AMtTse+eGGPWpTqGfhhtTzenNIo0lcr3k1JqeNFuDEoNIGPMy8CJ9tObRSQ7vKy7/m4icriIvCoim0SkTUQ2i8hbIvIDe/nMqH6th3bqQnFm522LyLdF5D0RaRKRT6K30V3/PBHZTkT+ISI1dreMN0Vkvxj1uv0ZvnP/W/un0Ovsp29Fxbymu3WiystE5M8islZE2u2/zd9EZFJ3MYlIjojcJSIV9t/xMxE5KVas3Ylnv+Gf/bG6LXR+bRO72W54ne2A7Todw+tj1J9ud4tpFJF6EXm+h23vZR+7Lfbr/lpEbhaRnDhe75nAavvpnE5xzYyqlyEil4vIpyLSYsf0uogcGmOb8+31vSLyWxFZLSJ+Ebk4usuEiOwsIq/Y26oSkQfC7xcROV5EPrT3tVFEfhkrdun0fuq0/R3sv1ut/X/6dRHZI8Z2DhORh0VkhV2v0X7/fL+3v18vf9uL7ViO7lT+uF3+eKfy4+zyizuVT7bj22D/n9wgIn8RkdIY+zxbRF6y//+22X/XF0Vkehzx9nqe6VT/SPvv1Cwi1fb7pDiOP41SSactu0oNMGPMf0XkHeAQYBbwUnd1ReQ4e/lme1qN1SViOnAS8DSwBiuxug5YCzwStYlPOm3y+8Dh9rb+S3zv8UK7bgVwHzDO3s7bIjLLGPNeHNuIJRznocDfsF4HQF1PK4lIGfABMBF4Dau7yFTgdOBYETnIGNO5ldoNvGq/luewukb8APi7iBxtjHm1t2AT2G8d1vGYmcBrC69zsf38zqhl8zvV3Re4AngL63jsBXwH2E1EdjXGtEbFfALW/5F24AWs/0d7A1cB3xSRQ4wx7T287E+Au4CLgE/tbYStsffhxfrbHgwswuqfnA18G3hDRE42xvwzxrafB3bF6sfcAGyIWjYJeBdYCDyA9V75IZAnIs9h/TryPPAecAJwk4hUGGMe7uG1RJuI1cVhCfAQMNmO9y0RmWaM2RJV9wpge6xjvxEoAmYDT4vIaGPMnXHus7P59nQmHftyz+w07VweXg8ROcBe14v1nl4N7AScBxwlIvsaY2qitnE38DHW8arG+jt826470xjzfg/xriH+88xs4FhgLtYxOgSrS9Jk4KAe9qFUahhj9KEPfcT5wPrwMMC/eql3g13vhqiyM+2yM6PK/gm0AWUxtlHc6bkB5nezv/C2g8DMGMtn2suvj7FNA/ytU/ksu/yzTuVrgDXdxDDfOqV0KLve3k6XmHpY5xF7nWs7lZ9tl78VIyaDlah5YryGeXEe20T32+Nr62YfPf39wsfIAN/vtOxRu/wHUWUlWEnkKmBMp/qX2/UvS+D/9CPdLL/ZXn5Vp/JS+/VUApmdjynwEVDQzb4M8JOochdWkhYCtgJ7RS0bC7QCX3Tzf/7MbrZ/Zaf6N3bzOibGeM3ZWMl/PZAV7/uwUz0HUAN8EFW2o73+G/Z0x6hli+36Dvu5ByvprAGmddr29+z1/9ypfFKMOKYBjcDr3fx/i3VO6O084wcOjCp3Yn05M8D+8b4f9KGPZD20G4NSg6PCnpbEUddvPzowxlT3Yb/PG2PmJ7hOELim077fwGrd3E1E9upDHH0iIhlYrcpbgd93Wvww8AUwU0QmxFj95yaqFdN+DWuxWkoHc78D7R1jzDOdyh6yp9Gv5QwgFyt5q+hU/zasJPQH/QlERBxYrYhLjTG3RC8zxlQCf8D6Pz4rxurXGWPqutn0SuCeqG0FsFrkBZhrjPk4atlGYAEwTUTi/TVyNXBrp7K/2tMO/x+MMWs6r2yMacZqsc8DunTniYcxJgS8A+wjInl28Tft6fXRz0UkH9gT69iH7GXHAROAW4wxyzpt+x9YXya+36l8NZ3Y674FHCwinr68lhieNMa8G7WPINbfC+J4vymVbNqNQanUegbrZ9ovROQp4E3gXWNdfNMXi/uwzlpjzPoY5QuAI4A9sFrdkmEq1k+2r5mon+vBav4VkbexfhrfA4i+uK0u1gc91k/nBwzifgfDRzHKwl0ACqLKZtjTA0Vk5xjr+LF+8u6PqfY+10ns/t472tOdgH91WtbT/8XPjTGd+2qHR1X4NEb9zVgtpeVYXQ1680lU0hgW62+InYhegfVz//Z0HR1kNH03397uwcDLWK2pq4zV1Wk1VrJ7v73cSccuLeHju0s3f/tMoERESowxVfZr2QH4hb3dMVitw9GKGZjRK+L9P6rUkKDJrlKDY4w9reypkjHmGRHxA5cAPwcuBQIiMg+rpXJlgvvd0nuVLrb2sq38Pmyzr8ItYN29js2d6oV19+UgQHwX4vZ1v4OhIUZZwJ46o8qK7OnPBjGW8D52tx/dyY5R1tP/xVivMdjDsvDrd/ewzR63b4wJiAhE/Q3tls63sVpVP8LqylJjx7InVqKaEec+Y5lvT7/JtmT331HLvmXPz+xUH7b97c/oZR/ZQJWI7Ig1xGAu8DpWn+cmrK4h38H6otaf1xIt3v+jSg0JmuwqNTjCV6n32tJqrIt7/mn/lHkQcDLWB9xOIrKL6fkCoy6bSzhSKOumvNyeRieSIbq2FoUNRFIc/hAt72Z5ead6AyVV++2PcCzTTNcL9gZ6H88YYxLqEhGj5XYo+jZWUvuAMebc6AUicqW9vD8+A2qxusBMxWolnm8vmw+cJSI7YSW7tXb9sPDf/lvGmB5vVmK7GKtV9VRjzJPRC0RkBlayq1Ra0j67Sg0wETkY6+rkKqxuCXExxtQbY142xswBXgR2wLq4JCzE4LSabCci42OUh6+qjv5ZuRYo69x30h4yake6CrfWxRv3l1gXI+1n96Pt7JAYMQ2EZO03yMAdww/t6f793E5Px2gZ1sVN00VkJLbYTbansUZMObC/G4/qtxseUQOs/rPR0+/Qtb8uJH58Y74WezSNveONmcE7zyiVMprsKjWAROQYrAttAH5hX+jSU/2DOycR9kVB4QvbovuP1mANCzbQnMBNnWKYhdVf94voi4Wwhp5yA6dG1RWsK/Zj/ZQdHhYpVjLdhTGmDasfczlWl47omOZg/ZQ+3wzwzSiSuN8arH6W3n5uB6wL55qwxnOe2nmhiOTHeXFhLdYvAl2OkX3h2L1YidTNsRJeEZkhIsP1Lnjh49khsRWR7wLHD9A+5mN91l4KfG2M2QBg95NfhdWFqXN/XbBGF1kPXC4iXRJeEcm0W2zDuryWqPdmd7/exDJY5xmlUka7MSjVN1OiLhrJwBob9xvAFKyhxC4xUXdP68GfgFEi8i7WME6C1b9vT6zhzb6MqvsmcLKIvIB1wVgQeMkYE/3TZ198hjUm60KsD9xxWF0pWoEfd6r7Z+As4EEROQKrT/LBWD+ffkrXn0rDwxH9VkR2weoSUWeM+XMP8VyB1Q3kNyJyCPA/rAulTsAaO/T8Pr3K3iVjv29ijaH8bxH5L9b4uO8YY95JdEPGmK0icipWkv65iLyCdRvrbKxxbGdiXSF/Xi/baRKRRcAhIvKYvY0Q8JgxZi1wrR3z5cBsO+5wQrQP1t9oNNCS6GsYAuZiJYlX2v8/lwO7AEdj9Xk9YQD2Md+eltK1BXk+1tB20fUA6wuYiHwPq4/veyLyKrAU63N7Itb/1YV2rGCNyXwWVpeoZ7Deawdj/V+YT9dxfbszWOcZpVJGk12l+mZHtt0dzIfVOrYUazzUv4Vbb+JwM9Yd1/bBulilFau152dYH17RLrKnh2G1OjmwroDu74dQLdYg8bdhJbderAH/rzbGfBBd0RjzhVh3hLoZ66YXTcArwGXA3ztv2BizVETOwmrV+inWF4O1WElzTHYSNwMryZqN9XprgMexhrOKNepCvyVpvzdi3fjiOLZdgf9rrJ+6+xLzSyKyD1aiPgs4Bquv5zqsm0U8EuemTgfusOPKx/rStQBrpI5WETkS6//G6VjDXbmxrur/DPgtVpedYccY0ygih2ENoXYQ1hfNT7D+jqMZmGQ33G+3kK6tt29hJbud++uG4/tARPbEOr7fsuNrwXrfP2o/wnU/st+bN2KNw9tu7+8UoMsd6HowWOcZpVJGhsc1BEoppZRSSiVO++wqpZRSSqkRS5NdpZRSSik1Ymmyq5RSSimlRixNdpVSSiml1IiVNqMxlJSUmIkTJ6Y6DKWGpS2VtZSXFqY6jKQK1W/CkT8agIqGKsbklfSyRvIMtXiSLfrYKJWIdDyXDVcfffRRlTGmdCC2lTajMUyfPt0sXtzrnVuVUjGM2eM0Kj59PNVhJNWmCzyMvtu6U7NcPhNz6/zUBhRlqMWTbNHHRqlEpOO5bLgSkY+MMdMHYlvajUEppZRSSo1Y2rKrlFJKKaWGFG3ZVUol1ePPvpnqEJKuZcGDkfn735+bwki6GmrxJFv0sVEqEel4LlOa7Cql4nDFjQ+lOoSkq3/qJ5H5Hz93Wwoj6WqoxZNs0cdGqUSk47lMabKrlFJKKaVGME12lVJKKaXUiJU24+zGo76+nqqqKtrbdUibkcrj8VBSUkJ+fn6qQxlWHrnrklSHkHSF5/0zMv/SWb9NYSRdDbV4ki362CiViHQ8lylNdiNaW1vZsmUL48aNIzMzExFJdUhqgBlj8Pl8bNiwgYyMDLxeb6pDGjZ233lSqkNIOvf4vSPz+4ydEtc6PV04du7+x/c7prB44xmpoo+NUolIx3OZ0m4MEZWVlZSWlpKVlaWJ7gglImRlZVFSUkJlZWWqwxlW9j7ip6kOIem2/nJiZH7sTSelLpAYhlo8yRZ9bJRKRDqey5QmuxGtra3k5OSkOgyVBLm5ubS2tqY6DKWUUkolgSa7tkAggMulvTrSgcvlIhAIpDoMpZRSSiWBJrtRtPtCetDjnLhTT/xmqkNIuswDz4nM/2jGcSmMpKuhFk+yRR8bpRKRjucypbcLjli2bBnTpk1LYkQqlfR4q8GQrAvUlFJqpNPbBSulkuqoH1yT6hCSrvKWGZH5fe48t9f6xhi+3LqOQCg4mGHFHc9IFn1slEpEOp7LlA491qtU30f7tJMOS+n+lQL4fNmaVIeQdIH1H0fm/7dxRa/1X12xiNv/+w/GF5Rx1vSjGZtfOmixxRPPSBZ9bJRKRDqey5S27Cql1IBYtnUtADUtDfz2zSeY9+WHhEwoxVEppZTSZFcp1avy0oJUh5B0jvzRkfnRecW91l9VvQmvy8OvjzyL3Udvz/Nf/Jdb5z/NlsbaAY8tnnhGsuhjo1Qi0vFcpjTZHbGuv/56RITly5dz1FFHkZ2dzYQJE3j44YcBeOyxx9hpp53Iycnhm9/8Jl9//XWH9e+//3722GMPvF4vJSUlnHPOOdTU1HSo8+c//5kDDjiAoqIiCgoK2H///Xn55Zc71FmzZg0iwn333ce1117L6NGjKSgo4Pjjj2fDhg2D+0dQA+bj1/+c6hCSrvy3ayPzFb96rtf6X1dXUJqdT25GFufOOJ5z9j2GzY013PjGo7y18mNCoYFr5Y0nnpEs+tgolYh0PJcpTXZHvO9973sce+yxvPDCC+yzzz6cffbZ/OIXv+Cee+7hlltu4eGHH+bLL7/k//7v/yLrXHXVVVxwwQUcfvjhvPTSS9x6663MmzePb33rWwSD2y6+WbNmDT/84Q/5xz/+wTPPPMP06dM57rjjmDdvXpc4br75ZlauXMlDDz3EXXfdxcKFCznttNOS8jdQ/feHe9IvuWp8+YbI/PWvPtxr/VU1FZRkFwDW8Hb7TZjGtUfMYUrJOJ7+9E3+svCFAYstnnhGsuhjo1Qi0vFcpvQCtRHv8ssv54wzzgBg+vTpzJ07l/vuu4/Vq1eTl5cHwKZNm7joootYu3YtxhhuvfVWrrvuOq699trIdqZMmcJBBx3E3Llz+c53vgPAH/7wh8jyUCjErFmzWLFiBffccw9HH310hzgmTpzIk08+GXleWVnJ5ZdfTkVFBWPGjBmsl68GyO33Ps9l55+Y6jCSqumVm8g91noP/Pq1v3H9kWd1WzcUCrG6ZhOHbr9Hh/LCzFx+euB3ufrfD/DemiVceOB3ByS23uIZ6aKPjVKJSMdzmdKW3RHvW9/6VmS+sLCQsrIy9t9//0iiC7DTTjsBsH79el577TVCoRCnnnoqgUAg8pgxYwa5ubm88847kfU++ugjjjvuOMrLy3G5XLjdbl577TW+/PLLLnEcc8wxHZ7vtttuAKxbt25AX69SqVDRUEVbwB9p2Y0mIozKLWJF1frkB6aUUkqT3ZGusLCww3OPxxOzDKC1tZWtW7cCsMMOO+B2uzs8Ghsbqa6uBqzEeNasWdTU1PCnP/2J9957j0WLFnH00UfT2traJY6ioqIOzzMyMiL7VGq4W1WzCYDS7PyYy8tzC/mycj3pchMfpZQaSrQbg+qguNi6yvvVV1/tkhRHL583bx719fX8/e9/Z9y4cZHlLS0tyQlUJdW8p25MdQhJV3Ll+5H5xRfd12Pdr6srACiN0bILMCqniKY2H5saqhmTX9Lv2HqLZ6SLPjZKJSIdz2VKk13VyRFHHIHD4WDdunUcccQR3dYLJ7VutztStmLFCt59990Oya9S6WBVdQVOh4OirNyYy8tzrS+OX1auH5BkVymlVPy0G4PqYPLkyVx55ZVceOGFXHHFFbz88su88cYbPPLII5x66qm89dZbABx++OG4XC7OOOMMXn31Vf72t79x5JFHMmHChBS/AjUYjj7lV6kOIemqfrd/ZH76XT/use7X1RVMKCjH6XDGXF6eY3XjWVE5MP12e4tnpIs+NkolIh3PZUpbdnuVjrfr/e1vf8u0adO4++67ufvuuxERxo8fz6xZs9hxxx0B2GWXXXjiiSe49tprmT17NpMnT+aWW25h3rx5zJ8/P7UvQKkkW1VTwfZF3d/ooDArl0x3Bl8OULKrlFIqfprsjlDXX389119/fZfyNWvWdCmbOXNmlwtnTj/9dE4//fQe93HyySdz8skndyj7wQ9+0OH5xIkTY16UE2ufSg1XX1dX8N1dD+l2uUOEHUvGabKrlFIpoN0YlFK9uuS8E1IdQtLlHHNNZP66I+Z0W6+htZmq5nq2L+75FrZTS8fzZeXADLXXUzzpIPrYKJWIdDyXKU12lVJxSMdB2KNvWtDTDRxW28OObV/U881RppSOY3XNZtoD/n7Hls43lAD0hhKqz9LxXKY02VVKxWGvwy9MdQhJt+UX20Xmx9zY8QPy/vfnRh73LHwJgE8rVva4vamlEwiZUGSYsv7oHE+6iT42SiUiHc9lSpNdpVQctlTWpTqEpAvVb4rMb2qo7rZeVXMdACXd3FAibGrpeIAB6crQUzzpIPrYKJWIdDyXqSGQ7IrIOBH5k4gsFJEWETEiMjFGPdPNY8/kR62UUpbKpnqyPV6yPN4e621LdvUiNaWUSqahMBrDDsDJwEfAf4Eje6j7CND51kErBicspVTYbtMmpjqEpHON3ysyv/fYKd3Wq2qu67VVFyA/M4fynEJWVG7od2w9xZMOoo+NUolIx3OZGhrJ7jvGmHIAEfkhPSe7G40xep9IpZLsP0/flOoQkq70qg8i8x9dfH+39Sqb69musDyubU4ZoBEZeoonHUQfG6USkY7nMjUEujEYY0KpjkEp1bPLb/hrqkNIuronz4/Mn/vsH2LWCYZCVLc0xNWyC+Hhx/rfjaG7eNJF9LFRKhHpeC5TQyDZTdD5ItJm9+19U0QOTnVASqWDJ557K9UhJJ3v3W0fig988K+YdWp9jYRMiNIEkt2q5npqWhr6FVt38aSL6GOjVCLS8Vymhley+zjwE+Bw4FygGHhTRGZ2t4KInCsii0VkcWVlZVKCVEqlj0p7JIbS7IK46k8tmwDAl1v1IjWllEqWodBnNy7GmOh71/5XRF4EvgBuAg7qZp37gfsBpk+f3qd707YseLAvqw2YrIN+2O9tLFy4kDvuuIMFCxZQVVVFbm4ue++9N6eddhqLFy/mnnvuYf369Ywe3fUOUMYYJk6cyKRJk5g/f35c+zvzzDP529/+FnmemZnJTjvtxIUXXsjZZ5/dpf6SJUv4+c9/znvvvUdGRgazZ8/mtttuo6ioqMf9rFmzhkmTJsVcVltbS0FBQVzxKtVXVc31AJTEmexOKRkHwIqq9RwwcZfBCksppVSU4dSy24ExphF4Gdg31bEMZXfeeScHHnggNTU1/O53v+P111/noYceYsqUKZx//vmUlJQQDAZ58sknY64/f/581q1bx5w5id2etLS0lIULF7Jw4UKeeuopCgoKOOecc3juuec61KuoqGDmzJn4fD6effZZ7r77bl5//XWOO+44QqH4unNfffXVkX2FH7m5uQnFq3r2v9f+lOoQkq7sN2si8xuveTZmncqmOpzioDArJ65tbl88BpfD2e9+u93Fky6ij41SiUjHc5kaRi27PehTi206eOedd7jkkku48MIL+eMf/9hh2be//W0uueQSmpubee6553j00Ue59NJLu2zj0UcfJSsri5NOOimhfXs8Hvbff//I81mzZjF+/HgeeOABTjxx292fbr31Vvx+P3Pnzo20xI4ZM4ZDDz2UF154ge9+97u97mv77bfvsC818D5buppRZYWpDiOp/Ov/h7PAugXwRxtXMCa/pEudquZ6irPzcUjv7Qb3vz8XgOKsPOYt/5CJhaMAOHf/4xOOrbt40kX0sVEqEel4LlPDuGVXRPKA44APUx3LUPW73/2OoqIifv/738dcPnnyZHbffXfmzJnDZ599xmeffdZheUtLC8899xwnnHACubm5LFq0iJNOOolx48aRmZnJ1KlT+cUvfoHP5+s1lpycHKZMmcK6dR2HXXrppZc49thjO3Q5OOSQQ5gwYQIvvvhi4i9aDYozL7o91SEkXe29275ozX74FzHrVDbXxX1xWlh5biFbmmr6FVt38aSL6GOjVCLS8VymhkiyKyInichJwD520bfsskPt5ZeJyAMi8n8iMlNE5gDvAqOAX6Yo7CEtGAzy1ltvceSRR+L19nxnp1NPPRWn08ljjz3WofyFF16gsbEx0oVh3bp17Lnnntx7773MmzePiy66iIceeoizzjorrnjWr1/P5MmTI2U+n4/Vq1ez6667dqm/yy67sHTp0nheKldffTUul4v8/Hxmz57N559/Htd6SvVXVXN93MOOhZXnFLG1qY6QjrqolFJJMVS6Mfyj0/O/2NO3gZnAl8AJ9iMfaMBKds8xxmjLbgxVVVX4fD622267XuuOGjWKo446iieffJJbbrkFp9MJWF0Yxo4dy6xZswA6dD8wxnDggQeSl5fHGWecwd13301xcXGH7QYCAQC2bt3KzTffTH19Pddcc01keW1tLcYYCgu7/qRUVFTEl19+2WPcGRkZ/PjHP+bII4+ktLSU5cuX89vf/pZvfOMbfPjhh0ybNq3X165UXzW3+2jxt8U9EkNYeW4hgVCQmpbGhBNlpZRSiRsSya4xRnpZPheYm6Rw0tKcOXN45ZVXeOONNzjyyCPZtGkTr7/+OpdddhkOh/UDQENDA7/5zW949tlnWb9+PX6/P7L+V1991SHZ3bhxI263O/JcRHjqqaeYMWPGgMU8evRo7r333sjzgw8+mKOPPppddtmF3/zmNzz++OMDtq909/tfdR1FY6TLP+Uvkfn7Tuzan73SHokh4W4MOdYoI5sba/qc7MaKJ51EHxulEpGO5zI1RLoxqIFXXFxMZmYma9eujav+7NmzKSgo4NFHHwXgiSeeIBgMdhiF4ayzzuLee+/lZz/7Ga+99hqLFi3i7rvvBqC1tbXD9srKyli0aBEffPABTz75JJMmTeLss89m+fLlkToFBQWICLW1tV3iqamp6XXosVjGjx/PQQcdxKJFixJeV3XvtJMOS3UISRc97F+si8giw47lFCS03VG51i8ZWxq7/r+PV18uahtJBmJIRpWe0vFcpjTZHbFcLhczZ87ktddeo62trdf6Xq+X73//+zz//PM0NTXx2GOPMX369EhXgNbWVl588UUuv/xyLrroIg499FCmT59OZmZmzO253W6mT5/OfvvtxymnnMIrr7xCIBDoMOJDVlYWEydOZMmSJV3WX7p0KTvvvHMfX73VkqwGzpg9Tkt1CEm36QJPZF4un9lleWVTHUDCrbO5GVlkujP6dZFarHjSSfSxUSoR6XguU5rsjmhXXXUV1dXVXHHFFTGXr169usMIDHPmzKGlpYXrrruOzz77rEOrbltbG8FgsEPXBIBHHnkkrlimTp3KBRdcwCuvvNKh1XX27Nm8/PLL1NfXR8oWLFjA2rVrmT17dlzbjrZu3ToWLFjAfvvtl/C6SiWiqrme3IwsvK7EEi8RoTynsF8tu0oppeKnye4Idsghh3D77bfzpz/9iSOOOIInnniC//73v7z00ktcdNFF7LrrrqxevTpS/4ADDmDKlCnccccdeDweTjnllMiy/Px89t9/f2677TYeffRRXnnlFU466SQ2btwYdzxXXXUV2dnZ3HDDDZGyyy+/HKfTyezZs5k3bx7PPPMMp556KjNmzOCEE06I1Hv77bdxuVyRbhYAl156KT//+c/5+9//zltvvcW9997LwQcfjMPh4Je/1EE61ODqy7BjYQMx/JhSSqn4DIkL1Iay4d437OKLL2a//fbjjjvu4LLLLovcLnj69Oncd999HH98x75/Z5xxBtdccw3HHHNMl9EVnnrqKc4//3wuuOACMjMzOfnkk7nrrrs47rjj4oqlrKyMn/3sZ9x88818/PHH7LXXXowdO5a33nqLSy65hBNPPBGPx8O3v/1tbrvttsiFcWCN/hAMBjvcVW2XXXbhnnvu4ZFHHqGpqYni4mIOO+wwrrvuOqZOndqPv5rq7PBD9kp1CEmXsesxkfnjph3QZXlVcz2Ti8f2advlOUV8sG4ZbQF/75VjiBVPOok+NkolIh3PZQrEmPS4Adn06dPN4sWLu12+bNkyHaoqjejxVv3xl/de4MLn7+KYnWYwe5cDE17/ow1fcv8H/+KaWadz49HnDEKESik1vInIR8aY6QOxLe3GoJTq1Rk/vS3VISRdzT3ficwf/9DVHZZVNzdgMJQmOBJDWFmONSLD5sa+dWXoHE+6iT42SiUiHc9lSpNdpVQcXn/n41SHkHRtX7wSmf/XsoUdlkWGHetrn1072d3S1LeL1DrHk26ij41SiUjHc5nSZFcppRJW2VwHkPDd08I8LjdFWbls6WPLrlJKqfhpsquUUgmqaq7D7XCR583u8zbKc4r63LKrlFIqfprsKqV6VfFp+t16efTd7ZF5c+v8Dssqm+spyc7H0Y+bl5TnWmPt9uUi4c7xpJvoY6NUItLxXKY02VVKxeHxZ99MdQhJ17Lgwcj8/e/P7bCsyk52+6M8p5DWQHufLlLrHE+6iT42SiUiHc9lSpNdpVQcrrjxoVSHkHT1T/0kMv/j57ZdwW2MobKp7zeUCBuVWwTAl5XrEl43Op50FH1slEpEOp7LlCa7SimVkMrmOtqC/j4POxZWnmMnu1vXD0BUSimlupNwsisi3xUR52AEo5RSQ93XVRVA34cdCyvMysXtcLGiasNAhKWUUqobfWnZfRZYKyI3iMiEgQ5IKTX0PHLXJakOIekKz/tnZP6ls34bmV9VYyW7fR12LMwhQllOQZ+6MUTHk46ij41SiUjHc5kCVx/WuRs4FbgGuFpE5gH3AS+bEXjv4VRfCHLu/sf3exsLFy7kjjvuYMGCBVRVVZGbm8vee+/NaaedxuLFi7nnnntYv349o0eP7rKuMYaJEycyadIk5s+fH9f+zjzzTP72t79FnmdmZrLTTjtx4YUXcvbZZ3epv2TJEn7+85/z3nvvkZGRwezZs7ntttsoKirqcT9vvPEGDz/8MAsXLqSiooIxY8Zw5JFH8utf/5qysrK4YlXx2X3nSakOIenc4/eOzO8zdkpkflX1JgCKs/P6vY9RuUUs35p4shsdTzqKPjZKJSIdz2WqDy27xpifAmOAs4HFwLHAi1itvdeKyJiBDVH1x5133smBBx5ITU0Nv/vd73j99dd56KGHmDJlCueffz4lJSUEg0GefPLJmOvPnz+fdevWMWfOnIT2W1paysKFC1m4cCFPPfUUBQUFnHPOOTz33HMd6lVUVDBz5kx8Ph/PPvssd999N6+//jrHHXccoVCox33ce++9VFdXc8011zBv3jyuvvpqXnrpJfbff3+ampoSilf1bO8jfprqEJJu6y8nRubH3nRSZH5t3WbyMrLwON393seo3CJW12ymLZDYUFrR8aSj6GOjVCLS8Vym+tayizGmFXgEeEREdgV+jNXaez1wjYi8DNxnjJk3QHGqPnjnnXe45JJLuPDCC/njH//YYdm3v/1tLrnkEpqbm3nuued49NFHufTSS7ts49FHHyUrK4uTTkrsw9Xj8bD//vtHns+aNYvx48fzwAMPcOKJJ0bKb731Vvx+P3PnzqWgoACAMWPGcOihh/LCCy/w3e9+t9t9/OUvf6G0tDTy/NBDD2XKlCkceuih/P3vf4/ZiqxUf21qqKEgM2dAtjUqt4iQCbGyaiO7jNIWJ6WUGgz9Ho3BGPNFVGvvWcAWYDbwsoisFpHLRKTvtxlSffa73/2OoqIifv/738dcPnnyZHbffXfmzJnDZ599xmeffdZheUtLC8899xwnnHACubm5LFq0iJNOOolx48aRmZnJ1KlT+cUvfoHP5+s1lpycHKZMmcK6dR1/sn3ppZc49thjI4kuwCGHHMKECRN48cUXe9xmdKIbtu+++wKwcePGXmNSqi8qGqrI9w5csgv0qSuDUkqp+AzI0GN2MnsG8FNgLCDAp0Ax8HtguYjsORD7UvEJBoO89dZbHHnkkXi93h7rnnrqqTidTh577LEO5S+88AKNjY2RLgzr1q1jzz335N5772XevHlcdNFFPPTQQ5x11llxxbN+/XomT54cKfP5fKxevZpdd921S/1ddtmFpUuXxvNSO3j77bcBmDZtWsLrqu6deuI3Ux1C0mUeeE5k/kczjovMVzRUk9+P2wRHK8stBGB5ghepRceTjqKPjVKJSMdzmepjN4YwEdkLqwvDKUAu0AI8CPzFGPOJiOQAPwF+DfwROKR/4ap4VVVV4fP52G677XqtO2rUKI466iiefPJJbrnlFpxOa2S5Rx99lLFjxzJr1iyADt0PjDEceOCB5OXlccYZZ3D33XdTXFzcYbuBQACArVu3cvPNN1NfX88111wTWV5ba90qtbCwsEtMRUVFfPnllwm95sbGRi6++GKmTZvGd77znYTWVT279dr0Sy4K/u+eyPz9J10GgD8YYGtTLfuN32lA9uF1eRhfUJZwy244nnQVfWyUSkQ6nstU38bZzRKRc0TkQ6wL1M4FNgAXAWOMMecaYz4BMMY0GWN+j5UATx+4sNVAmzNnDhUVFbzxxhsAbNq0iddff53TTjsNh8P6b9LQ0MCVV17J5MmTycjIwO12c/rpp2OM4auvvuqwvY0bN+J2u3G73YwdO5a7776bhx56iBkzZgxK/IFAgFNOOYWNGzfy9NNP43L163uc6uSoH1zTe6URpvKWbf9X97nzXIDIrX0HqhsDwE6lExJOdsPxpKvoY6NUItLxXKb61o2hArgf2BN4DjjMGLOLMeZPxpiGbtbZCPT8W7oaUMXFxWRmZrJ27dq46s+ePZuCggIeffRRAJ544gmCwWCHURjOOuss7r33Xn72s5/x2muvsWjRIu6++24AWltbO2yvrKyMRYsW8cEHH/Dkk08yadIkzj77bJYvXx6pU1BQgIhQW1vbJZ6amppehx4LC4VCzJkzh9dff50XXniB3XffPa71VPw+X7Ym1SEkXWD9x5H5/21cAcCmhmoACjIH7jKEncomsLxyHYmM3BiOJ11FHxulEpGO5zLVt24MjcBtwAPGmM1xrvMX4Kk+7Ev1kcvlYubMmbz22mu0tbWRkZHRY32v18v3v/99HnvsMZqamnjssceYPn16pO9ra2srL774Itdffz0XXXRRZL3PP/885vbcbjfTp1uN+fvttx977703u+++O5deeikvv/wyAFlZWUycOJElS5Z0WX/p0qUceuihcb3W8847j2eeeYZnn3020uVCqcFQ0VAFQMFAtuyWTaCpzUdFQxVj87tedKmUUqp/+tKyu50x5sYEEl2MMQ3GmPiaGNWAueqqq6iuruaKK66IuXz16tUdRmCYM2cOLS0tXHfddXz22WcdWnXb2toIBoO43R3HFn3kkUfiimXq1KlccMEFvPLKKyxatChSPnv2bF5++WXq6+sjZQsWLGDt2rXMnj271+1eeumlPPjggzz88MPaT3cQlZcWpDqEpHPkb7vJyug8qz96hd2yO1AXqIGV7EJiIzKE40lX0cdGqUSk47lM9S3ZfV1EzuipgoicJiJv9jEmNUAOOeQQbr/9dv70pz9xxBFH8MQTT/Df//6Xl156iYsuuohdd92V1atXR+ofcMABTJkyhTvuuAOPx8Mpp5wSWZafn8/+++/PbbfdxqOPPsorr7zCSSedlNAQX1dddRXZ2dnccMMNkbLLL78cp9PJ7NmzmTdvHs888wynnnoqM2bM4IQTTojUe/vtt3G5XJFuFmANrXb77bdz1llnseOOO/L+++9HHl9//XVf/2wqho9f/3OqQ0i68t9u+35e8SvrZigVDVU4xEGuN2vA9rNTaeLJbjiedBV9bJRKRDqey1TfujHMBOb3Umc7IL7foIe4gbhdbypdfPHF7Lffftxxxx1cdtllkdsFT58+nfvuu4/jj+/4+s444wyuueYajjnmmC6jKzz11FOcf/75XHDBBWRmZnLyySdz1113cdxx8Q2DVFZWxs9+9jNuvvlmPv74Y/baay/Gjh3LW2+9xSWXXMKJJ56Ix+Ph29/+NrfddlvkwjiwRn8IBoMd7qr273//G4CHHnqIhx56qMO+5syZE3ers+rdH+55jsvOP7H3iiNI48s3kHvstQBc/+rDXH/kWVQ0VDMqtwiHDMiojYDVSpubkZXQ8GPheNJV9LFRKhHpeC5TIIlcFAEgIiHgemPMDT3UuQm43BjTc0fRJJo+fbpZvHhxt8uXLVumY7OmET3eiRmzx2lUfPp4qsNIqk0XeBh9t3UbX7l8JubW+XzrwSuoaq4f0HFuz93/ePb743nke7N57dzb4lonHE+6ij42SiUiHc9lw5WIfGSMGZCRvPraPBEzQxbLdsAxwPo+R6WUUkNQRUP1oPSX7cvwY0oppeITV7IrIiERCYpI0C66Pvw8+gEEgFVYw5I9PTghK6VUalQ0VDFmMJLdsglsqK+ksbVlwLetlFLpLt4+u++wrTX3EGAdsCZGvSBQDbyBdSMJpdQIMO+pG1MdQtKVXPl+ZH7xRffRHvBT1VzPmLySAd9XeESGFVXr2Wfc1F7rL77ovgGPYTiJPjZKJSIdz2UqzmTXGDMzPG/32X24pz67Sik10oTvnjYmr5hQgtc69CZ6+LF4kl2llFLx60uf3UnAXQMdyFCQ6MV6anjS45y4o0/5VapDSLqq3+0fmZ9+148jN5QYjJbdycVjcDoccffbnX7Xjwc8huEk+tgolYh0PJepPiS7xpi1xpj63msOL263G5/Pl+owVBL4fL4uN8dQqjfhG0oMxgVqGS4P2xeN4ctKva5XKaUGWq/dGETkWqz+uncbY2rs5/Ewxphh0zmmrKyMjRs3MnbsWDIzMxGRVIekBpgxBp/Px8aNGykvL091OGqY2dayOzh3L9upTEdkUEqpwRBPn93rsZLdZ4Aa+3k8DDBskt28vDwAKioq8Pv9KY5GDRa32015eXnkeKv4XHLeCb1XGmFyjrkmMn/dEXOoaKjG6XBQml0wKPvbqXQCr65YRDAUxOlw9lj3uiPm9Lh8pIs+NkolIh3PZSqOm0qISPhOaB8YY1qjnvfKGPN2rwGIjAOuBKYDewCZwCRjzJpO9bxYyfNpQAHwCXClMeadeGLp7aYSSinVk7OeuYXXv/qI9df8g/vfnzvg2393zec8+tGr3HTUOZTmFAz7uzcqpVR/JPWmEsaYt+1Ha6fnvT7ijGEH4GSgFvhvD/X+CvwIuBY4DtgE/EdE9oxzP0qpPtrr8AtTHULSbfnFdpH5MTeeSEVD9aBcnBY2KrcI2DbqQ0/G3JjetzuNPjZKJSIdz2Wq73dQG0jvGGPKjTHHAP+IVUFE9gD+D/i5MeYBY8wbWAnyOkCHQFNqkG2prEt1CEkXqt8Umd/UUE1FQxWj84oGbX/lOfEnu5vsi+XSVfSxUSoR6XguU31IdkVkoogcIyLZUWUuEfm1iHwqIu+JSNydYowxoTiqzQb8WP2Gw+sFsO7SdpSIZCTwEpRSKmGD3bKbk5FJjiczrmRXKaVU/PrSsnsd8BjQFlV2DfArYDdgf+DvIjKQAyHuAqw2xnS+l+YSwIPVFUIpNUh2mzYx1SEknWv8XpH5vcbsQE1Lw6CNxBA2KreIzU29J7t7j50yqHEMddHHRqlEpOO5TPUt2T0AeMNuWUVEHMBPgOXABGA/oBn4+UAFCRRh9entrCZqeRcicq6ILBaRxZWVlQMYjlLp5T9P35TqEJKu9KoPIvP/nGMNLDOYLbtgJbtb4mjZ/eji+wc1jqEu+tgolYh0PJepviW75cDaqOd7AiVY4/BuMMYsBl4E9u1/eP1jjLnfGDPdGDO9tLQ01eEoNWxdfsNfUx1C0tU9eX5k/qIX/wQM3hi7YaNyi2hs89HU1vMNbs599g+DGsdQF31slEpEOp7LVN+SXTfWGLphB9rP34wq2wCM7kdcndUChTHKwy262slNqUH0xHNvpTqEpPO9u+1D8aWl7wGDc/e0aOERGbb00pXhgQ/+NahxDHXRx0apRKTjuUz1LdndAOwe9fwYoMoYsyyqrAxo6E9gnSwBJolIVqfynYF2YOUA7ksppWJKRjcGiG9EBqWUUvHpS7L7L+AIEfmDiNwEHAG81KnOFDp2deivuVgtyt8LF4iIC/g+8Koxpq27FZVSaiC4nS6Kswb3znvF2Xm4HE5NdpVSagDFc7vgzn4PfAe4xH6+EWuEBgBEpAzrIrY/xrtBETnJnt3Hnn5LRCqBSvsGFR+LyDPAnSLiBlYD5wOTgFP78BqUUgn432t/SnUISVf2mzWR+ZN2O5QP1y/H4Rjcockd4qAsp7DXZHfjNc8OahxDXfSxUSoR6XguU31Ido0xW0VkN2CWXfS2MaYxqkoJcDnwnwQ22/lmEn8JbxuYac+fBfwGuAnrdsGfAkcbY/6XwH6UUn3w2dLVjCqL1W1+5PKv/x/OgjEAfFW1YVBvKBFtVG4RG+p7Hj3mo40rGJM/uF0qhrLoY6NUItLxXKb6eAc1Y4zPGPMv+9HYadlSY8xdxpjlCWxPunnM7LTPS4wxo4wxXmPMDGPM/L7Er5RKzJkX3Z7qEJKu9t7vRuY/3fT1oPfXDRuVW0RVcx1tgfZu68x++BdJiWWoij42SiUiHc9lamjcLlgppYa8wR52LGxUbhEhY/i6uiIp+1NKqZGuL312EZEi4GysG0gUAs4Y1YwxZlaMcqWUGjZ8fuv613hadttXLuh2mWeHg+LaX3hEhuVb17Fz+cS41lFKKdW9hJNdEdkJmA+UAtJDVdPDMqXUMPL7X52d6hCSLv+Uv3D/+3OpbKoDYEXleu5/f26ft9ddItw5CS7PtfoTLt+6rttt3XfipX2OYyTIP+UvvVdSKoZ0PJepvnVj+APWOLq/A7YH3MYYR4xHrNZepdQwdNpJh6U6hKTLOuiHANS3NgGQ781Jyn69Lg+FmTk9Jrvn7n98UmIZqsLHRqlEpeO5TPUt2T0YeNkY8wtjzBpjTHCgg1JKDS1j9jgt1SEk3aYLPADUtTYDUJCZnbR9j8otYnll98muXD4zabEMReFjo1Si0vFcpvqW7AqwdKADUUqpoajeZyW7yWrZBSjPLWL51nUYo73BlFKqv/qS7H4ETB3oQJRSaiiqs7sxZHu8fVp/dM3XHP3Rg+z99Ws4QoG41hmVW0RjWwsVDVV92qdSSqlt+pLs3gAcIyIzBzYUpdRQdfghe6U6hKTL2PUYwOqz63a4EOnpetyu3IFWpq+cxze+fAlEmLT1Cw794u9ktjX2um545Iclm9fEXH7ctAMSimWkCR8bpRKVjucy1behx8YDLwKvishTWC29dbEqGmMe7XtoSqmh4tE/pd/V/0XnvwDvz6XO18z4gtKE1i2vW8PeX7+Gt72ZZWNnsGzcDEbXrWb6V/M47PMneX/KsVTnjet2/fCYvku2rOHIqft2WT737JsTimekKTr/hVSHoIapdDyXqb617D4CfBcrUT4duBN4uNPjEXuqlBoBzvjpbakOIelq7vkOAA2tTVS1NMS1jivYzt5fv8ZBy54n4PTw1m4/YOmEb2AcTiqKduCt3U7B78zgkKXPsf3mT6GbPrm5GVmUZhfwxebVMZcf/9DVfXpNI0X42CiVqHQ8l6m+teyeNeBRKKWGtNff+TjVISRd2xevwF7nUNfaHLmxRE8ymys54NPHyGpr4Msx+7B0/DcIOTqeYhuzinlzt1PYb+W/2Wv1mxQ2bWHJ9jMIObuOLrDrqEks2RI72f3XsoV9e1EjRNsXr6Q6BDVMpeO5TPUh2TXG/G0wAlFKqaGmLeCPK9EF2PGzZ/G2N/P2LidTnTe223oBVwbvTf02O69fyLSNH5A1/3d8cNg10KlP8C6jJvK3xf/BGJNwf2GllFLb9Ol2wUoplQ7CN5TorPPd0DLbGhn/9ZusLtutx0Q3QoSlE75Bu9vLHmveZvS699m0XceLznYpn0hjWwvr67YyobC8z69BKaXSXV/67AIgIqUicp6I3CUiD3Yq309EMgcmRKVUqlV8+niqQ0i60Xe3U2ePsXvRQSf2WHdKxWIAVoydntA+vh61Jw0FE9j5o0dxBNs7LNulfBJgXaTWmbl1fkL7GWlG393eeyWlYkjHc5nqY7IrIucAa4C7gZ/SsR9vObAQ+L/+BqeUGhoef/bNVIeQdC0LHoy07H5dXdFtvYz2ZiZt+Zy1JdNoychLaB9GHHwx/Wyymrey/dK5HZbtMmoiQMyL1O5/f26XsnTSsuDB3ispFUM6nstUH5JdETkCuB9YAZwA3BO93BjzBbAE+M4AxKeUGgKuuPGhVIeQdPVP/YR6+1bBPV0QtuOmj3CYECvGdh0iLB7Vo3dj0/gZ7PjFP8loqYmUF2XlMTqvmCUxkt0fP5feV5TXP/WTVIeghql0PJepvrXsXglsAg41xrwEbI1R5zNg5/4EppRSqVbns24o0R2P38fkzZ+xvmQqTZmFfdpH+8oFfFqyMxL0M/Wd22lfuSDSJ3iX8okxuzEopZSKX1+S3enAv4wxPQ08uQEY1beQlFJqaKhvbSLfm93t8h02fYwr5Gd5H1t1w5q9BXw1em+2q1xGYePmSPku5RNZumUtoVCoX9tXSql01pdk1wM091KnAAj2YdtKqSHokbsuSXUISVd43j+pa20mPzObnxzwnS7LXYE2Jm/+hI1FO9CYVdLv/X05bj987iz2WDM/crOJXUdNosXfyprazR3qvnTWb/u9v+Gs8Lx/pjoENUyl47lM9S3ZXQPs00udGcCXfdi2UmoI2n3nSakOIenc4/em3tdEgTeH7WIM/TV586d4gm0sH7vfgOwv4PSwZMJBFDdtYnzVcqD7ERn2GTtlQPY5XLnH753qENQwlY7nMtW3ZPdF4GAR+V6shSJyFrA78Fx/AlNKDR17H/HTVIeQdFt/OZH61mbyM3O48pX7OixzBv3suOkjNhVMoi5n4MbAXVu6MzXZ5ey6bgGhtmZ2Lt8OoMtFamNvOmnA9jkcbf3lxFSHoIapdDyXqb4lu78H1gFPicgzwAEAInKh/fx+4CvgTwMWpVJKJVkTDloD7RTE6LO7/ZbPyAi0snzcwLTqRojw6aSZZLU30fzareRn5jAuv1QvUlNKqX5IONk1xtQChwILgO8BRwIC/NF+/h4wyxjTW79epZQasraIB4B8b06HckewnR0rFrM1bzw1uWMGfL81uWNYVzyVptdvJ1i7gV1HTYo51q5SSqn49Ol2wcaYdcBMEdkdq2W3GKgH3jfGfDSA8SmlhoBTT/xmqkNIuvrdvgMr15Gfmc1BE3eLlI//+i0y/S0s2vGYQdv3ku0OYsKnq2iadzO7lO/C/K8/IRgK4nQ4AfjRjOMGbd/DQeaB56Q6BDVMpeO5TPXjdsEAxpjPjDH3GWN+a4y5WxNdpUamW69Nv+SiYb85ABR4czh9nyMj5eNXvkldVimVeeMGbd8tGXlkHfhDWt57mJ2ys2kNtLOqelNk+f0nXTZo+x4OCv7vnt4rKRVDOp7LVD+SXRHZTkSmi8g+IjJhIINSSg0tR/3gmlSHkHQrXrgWgILMHH7zxmMAZDVsorB6JetKdwKRQd1/zlFXgtPFpBX/AWDJlm1dGfa589xB3fdQV3nLjFSHoIapdDyXqQSTXREpEZHbRWQTsAr4APgQWC0iFSJyq4gUDUagSqnU+XzZmlSHkHSb6ivxOF14XR7W1Vk3ihy7xrqz2YbiqYO+f2fBGLIP+QnbffECQId+u//buGLQ9z+UBdZ/nOoQ1DCVjucylUCfXRHZEXgNGI91QVoAqLbni7DumHYJcKKIHG6MWTXw4SqlVHJsFjf53hwk3IJrDGNX/5fqsmn4MnIHff/3vz8XT9FuzHK6GOMwvLTkXcpy+nZLYqWUSmdxteyKiAN4ApgAvA0cDuQYY0YbY0YBuVijMrwDTAQeH5RolVIpUV5akOoQkm6rO4f8TGvYsXxvNnm1a8ht2MjGSQcnLYZ2bx6rph3Hzv5aNtdu67M7Oq84aTEMRY780akOQQ1T6XguU/F3YzgSmA78HWtYsTeNMe3hhcaYNmPM68BhwLPADBE5YsCjVUqlxMev/znVISTdlsLJFNjDjv3+2PMYu2YBIXGyacIBSY3j651nM1kCbG6uJxgKAVDxq/S+Z0/5b9emOgQ1TKXjuUzFn+yeCLQBPzXGvml7DPayCwE/kN63+FFqBPnDPemVXLX621hTU0GJv4n2lQt44b9PMfqrN9iSP4GmDZ8lJYb2lQtoX7mAlnUfk59Xhh+hedHTtK9cwPWvPpyUGIaqxpdvSHUIaphKt3OZssSb7O4NvGuMqeytojFmK9YNJ/Tm5UqNELff+3yqQ0iqFZUbCAGjMjIB+PfWCrLbG1lfMvgXpsXiKLf3u+lzAH792t9SEsdQ0fTKTakOQQ1T6XYuU5Z4k93xwJIEtrsE2C7xcJRSKvWWbbV+Ji/3eiNlQXGyqWhySuIpzcpFgM2+Fkrr16ckBqWUGq7iTXbzgLoEtluHddGaUkoNO8u2rsVhDGUZmUgoCEBF0WQCTk9K4vE4nBR7MljqymPn9e+mJAallBqu4k12PUAwge2G7HWUUiPAvKduTHUISbV0y1omFZTgcTgoq1/PPN+nbEhRF4aw0d4slngKKWncxLvHnpbSWFKt5Mr3Ux2CGqbS7VymLIncVKLbC9OUUmokWbZ1LVMLywEYX70cv9PD5oKJKY1ptDeTDQGo9hbS/MZdGH9rSuNRSqnhIpFk93oRCcbzAK4drICVUsl39Cm/SnUISRMIBlhRuYHtvnoNRzDAmOqVHO+ZRsgR9z14BsVobyYhDPPGHsSRrcU0vXFHSuNJparf7Z/qENQwlU7nMrVNIsmuJPgYUCIyU0RMjEfdQO9LKZW+VtVsoj3oZ0qohdF1q3CH/KkOCYBRXmtkiCVu6y5qTfNuIVCt480qpVRv4mqqMMYkkhQPtp8Bi6KeB1IViFJq5Fm2xUogdzQ+yqu+pNWdleKILOUZmTiAzW2+SFnjP6+g8EfPpC4opZQaBoZSEhuvZcaY96Mei1MdkFIj3SXnnZDqEJImPOzY7of+iFG1q9lQPIVvlY9NcVTgdjgoyfBS0drCdUfMIefoq2j95Hnalr2e6tCSLueYa1Idghqm0ulcprYZjsmuUirJLjv/xFSHkDTLtq5jTF4JxWN3wWmCrC/ZiWNGjUt1WIDVb3dzq4/rjzyLnFmX4Czdgfp/XIwJtPe+8giSe6xeFqL6Jp3OZWqb4ZjsPmFfCFctIk+KyIRUB6TUSLfX4RemOoSkWbplDTuXb0f90z+hOSOPmpxR/HLJ/1IdFmANP1bZ1sroG76LuDPIO+k2gltW0PzWH1MdWlJt+YXes0j1TTqdy9Q2wynZrQduA34IHAbcCBwOLBSRslgriMi5IrJYRBZXVvZ6p2OlVDe2VNalOoSkMMawvHIdOxWUQqCd9SU7gQgNgaFxkdpobyYG2NxYQ8uCBwnVbcQ1dnca//Vrml69NdXhJU2oflOqQ1DDVLqcy1RHwybZNcZ8bIy5zBgz1xjztjHmTuBooBzrorVY69xvjJlujJleWlqazHCVUsPQhvpKmtp8TG7ZDMC6kp1SHFFHo+0RGaJl7n0ShIK0fvx8CiJSSqmhb9gku7EYY/4HrAD2TXUsSo1ku02bmOoQkiI8EsOkDYvAnUljVjEA4zOHzogMeS43+S53pMyRU0LGzkfiX7eYthXzUxdcErnG75XqENQwlS7nMtXRsE52o+jd3ZQaRP95+qZUh5AUS7euAWD7TZ+Qe9x1kfIrpuyWoog6coiwe34hAWPwBbeNupgx7Ugku5j6J84j5GtIYYTJUXrVB6kOQQ1T6XIuUx0N62RXRKYDU4EPUx2LUiPZ5Tf8NdUhJMWyLesodLkolgD+9Z9Eyp9avyp1QXWye34hzcEAb1VujpSJy0PWAWcSrF5DwzMxe3WNKHVPnp/qENQwlS7nMtXRsEl2ReQJEblJRL4rIoeJyKXAPGAjkF6XIiuVZE8891aqQ0iKZVvXsmOohYwdDqF18dOR8vdqhs4Frjtm5wHw0qb1HcpdpZPJOeYafIuepOWDx1MRWtL43tWERfVNupzLVEfDJtkFvgBmAw8D/wEuBv4JzDDGVKUwLqXUCLF000p2bK8hc99TUh1Kt1wO67T9yuYNBEKhDstyjr4az+SDaHjmZwS2rkxFeEopNeQMm2TXGHOzMWZ3Y0y+McZtjBlvjDnXGKNj0Cil+q2yqY7q1hZ2xI93r++mOpxeVbe3sbBTi7M4nBSc+Qg43dQ9fHra3WxCKaViGTbJrlIqdf732p9SHcKgW7p5NQA7T9gVR1YBZb9ZE1l2085D6+r/T2fNJsPh6NKVoWXBg7QtfZXMvU/Cv+4jau49gZYFD6YoysETfWyUSkQ6nMtUV5rsKqV69dnS1akOYdB9/vmrAOy+7wkA+Ndvu2vaupbmlMTUnRVNDRxWOpq5m9ZjTNfBaNzj98Kzw0G0L3sN/6ZlKYhwcEUfG6USkQ7nMtWVJrtKqV6dedHtqQ5h0H3x5XtkmSA77HsyALX3buvKcP+aFakKK6bvfTCf40ePZ72vmU/ra2PW8e51Eo680fje/xvBxq1JjnBwRR8bpRKRDucy1ZUmu0qptPfggmdZWrmeCR4Xf/34De5/f26qQ+rVMaPG4UCY26krQ5i4PGR94yxMewv1j56DiRqXVyml0okmu0qptFe+8SNWkkFZwahUhxK30gwv3ygu7dJvN5qzcByZ+5xM29L/UP/UT2J2eVBKqZFOk12lVK9+/6uzUx3CoMr77EU2OTLIczpoX7mA9pUL+N/2syLLfzBuUgqj6+pPe8wA4PjR41naWMfXTY3d1vXscBA53/olvoWP0PjC1ckKcVDln/KXVIeghqmRfi5TsWmyq5Tq1WknHZbqEAZNqLmGxoYtAIzyZkXKV5fvHpk/sLgs6XH15qE1X+ELBgG4dunHPLTmq27r5hx7LVmHnEfz67fT9NofkhXioMk66IepDkENUyP5XKa6p8muUqpXY/Y4LdUhDJrWT57na/ECMCojM1J+4sI7IvM//fSDpMfVk3A8xZ4Mxnmz+LS+psf6IkLe9+7Eu8/JNL7wC1refSgZYQ6aTRd4Uh2CGqZG8rlMdU+TXaVUWvMteoolnkKcIpRkeFMdTsJ2zy9kTUsTDf6ebyAhDgcFZzxExs5HUf/UT/B98nySIlRKqdTSZFcplbYCVatoX/lfPvcUU+rx4hRJdUgJ2yO/CAN83lDXbZ2WBQ/SsuBBfO8/SsYuR+Ms3o66v55K2/I3kxanUkqliia7SqleHX7I0LqD2EBpfu12cLpZRQajvJkdlm0q3HZR2q55BUmOrGfR8Yz2ZlLiyei1K0OYuDLIOuQnOHLLqL3/JFo/f3mQohw8Gbsek+oQ1DA1Us9lqmeSLkPRTJ8+3SxevDjVYSilhohg/Sa2Xrsjjv1OZ/QnyziqfCzHjhqX6rD65PmKtbxdtYUN3/oeee74+rOGWurwffQMgQ2fkn3kFeQedz3idA1ypEopFR8R+cgYM30gtqUtu0qpXp3x09tSHcKAa37jLggGqNj7FAxQ3qm/7jeWvxCZv2/1l8kNrhed49kjv4igMfxnS0Xc23BkFVBy6TtkHngOza/+npo/fYtg/eaBDnVQ1NzznVSHoIapkXguU73TZFcp1avX3/k41SEMqFBzDS0L7sc7/fusaLPuLNa5G8Po2tWR+S966A+bCp3jmZiVQ67LzZPrVyW0HfFkUvB/95B/+oO0r/mQqlv2o+2rdwYw0sHR9sUrqQ5BDVMj7Vym4qPJrlIq7TS//RdMWxM5R17Osq1rEaAsI7PX9YYqhwgzS0bx6tYK3q3emvD6WfufQcnlCxBvLjV3HUnTq7diQqFBiFQppZJPk12lVFoJtTbRPP/PZOx+PO4xu7Js6zqKPRl4HMP7dDiztJzR3kx+teR/cd8WODxKQ8uCB/Gv/oDsQ87DPW5PGl/8JTV/PBJ/xZJBjloppQafXqCmlEorTW/cSeM/r6D4sgV4Ju3HbredhWmu5bztp6Y6tH4T4MJPP+CZ/Q7luNHj+7QNYwz+r9+ldcm/Ma2NZH/zZ+Qc80sc3tyBDVYppXqgF6gppZLq8WdHxnisxt9G8xt34Jl6GJ5J+xEIBlhRuaFLf12ASVs+i8z3pWvAYOountMnTGZKTh7XL/uEQB+7IYgInh0OIufoq3FP3I/mN25n6zXbU/fEeTT/94H+hD1gWhY8mOoQ1DA1Us5lKjGa7CqlenXFjcP79rJhvg8eI1S/iZyjrgTgk4qVtAf9jI6R7O696o3I/NMbVndZnkrdxeNyOLhu2p4sa6zniQQvVuvMkZFD1ozTyD7icsSbh++9h2h+64/4Ny/r13YHQv1TP0l1CGqYGinnMpUYHVRRKZUWTDBA02t/wD1xPzxTZgJw7/svkeX2slt+YWqD64NV67oOE/Z+TZCyUIip7iyu/ewj8tb6ICj4g4ZQCEIGQsbYUzAGRMAReUjk+T67bY/H5cDtHI3nGz/HvX4hwWX/ouqmvciccTo5x1yDq3i7FLxypZRKjCa7Sqm0UP/UBQSrVpEx7Qh87/6VmvY2nvzoP/xg3CSyhuDNFFat20woZGhrg9ZW8LVCq8/gawVGwwcfhmhvN7S3E3k8699MMAitBVnU7FvF+e+vIWdtH/vavlDdqUAocB3BWeOX873gYzS89xj/rt+ZF5r2pdVVSIuvFbfbhcftikw9bhduj4sMtxuPx0WGx43H4ybD48LjcfOdow8gNzeTvJwscnOy8Ga4kR5u2Rz+CXoWXX+OPu2kw/r2OpVSI97QO8MrpYacR+66JNUh9IsJhWhb+h8c+aNxjd0NgMfWfY0vGOTcSVNZXFvVZZ33ps6OzJ87ccrAx2QMzS0h6uuDNDQGaWgI0NAYpL4hSGNjkNq6IG1tVutrZyVbi9nSaPC4wZMBubngcQtuNzhd4HRm8nG7l7odGzlkVA4HZRdaLbcOq/U23JIrYm3fmK6tvsGQwR+EQMgQCIbnc1gTPJA/N+3JQZ5POLZgCUflL+fNpqn8Uw6hth2amltp9wfw+wO02w+/PxDzb/DA4/M6PHc4BG+Gh7KSfArycijIz6YgP4eCvBwKC3L4enUF2Vle2kt/StXmanKyM8nO8uJ0ao88FZ/hfi5TfaPJrlKqV7vvPCnVIfRL68fPEqqvIPOAMxFxEDKGB9es4BtFpeyeXxgz2a3NKY/MT8jK7tN+g0FDXX2AmtoAX6+upakJmpoNzU3Q1AzBYMf6LhdkZVmPMaOFzEzweiEzU/B6wZsJmV7wiZcch7PHfZcGCnmicRNbyxvZr3R0n+LvXi5wNKvb96e06gOOdizhiMLVbBp9BOvHHo8vq+Ntl0OhEO3+IO3tftqiH21+Wtv8tLa109rabj9vp6W1jZbWNqprG2jxtdHia6O1tZ1w3l/i8VHVviWy/Uyvhz/85Tlyc7PIzc4kN8d+ZGfynWO+QXlJAeVlhRTkZffYcqxGvuF+LlN9o8muUqpXex/xUyo+fTzVYfRJYOtX1D/5E5xFE3BP2AeA17ZWsKq5iWt32rPb9Y796AGeO+DnAFyz9GP+tMeMmPVCIUNdfZCamgBfraqhsQkaGw1NTdDS0rFl1uWC7GzIyYXyUUJONmRnC1lZVrnbTVzJ2AO1G/l5Yc/9ZctcHnbyZPFxayOVgXZKXZ5et5sov6eAijFHUd26D1mtmxi78WXGb3iR6qJ9WD92NtXF00GcOBwOvBkOvBnuPu8rFArha22nuaWVUz7/AXeOupemllaamn00N7fS2OyjscnH+k1VNDa10N5utSY/89J/I9vI8LgpKymgvKyAtlY/eblZ5NndKPJyra4UeblZ/Oi0ozUpHqGG87lM9Z0mu0qpESvU1kztAyeD003WQT9C7NbQ+1Z/SXmGl2+PiX8s2hZfkOrqANU1AVatraexwdDYaLXQRo/y5XZb3QqKi4WJEyEnB3JzhJwcyMiIL5kdKAd4C1jR3sKjDRVcWjRx0PbT5i2hzVtCQ85kCuu+oKD+c/as+Yh2dx61BbuxYsef0J5R3K99OBwOsrO8ZGd5Adht2sSeY2r309DYQlOzj4bGFhqarGljUwt1dU00NLbw1eoKWtvau6x78x//TllJAWWlBZSXFFgJcqk1LSvJp7Qkn/KSAooL83C5em5hV0qlnia7SqkRyRhD/ZPnE9i0lKILXiZYvQaAVc2NvLqlgqum7oanU1eAcCttdY0fgKefraCh0cDecOeft41+4HBAbg7k5cHYcWLPpyah7UmB080eGbm80lzFTp5sjs0pHdT9BV1ZVJXsR1XxPuQ2raKw9lPKK9+lrPI96vN3ZkvZwVSWHEibd3DjAKsVt7Q4n9Li/B7rtbf7I4lwQ1NLZNrY6KO+von1GytpbGrB19o1KRYRigtzKSstoKw4n7KSAkrtaVlJAZ8s+ZrcbKv1OMPT8eI7vaBOqeTRZFcp1atTT/xmqkNIWMv8u2ld/DS5x99AxrTDIzcieGD1Chwi7O8cxfMfbGbVlhbeXF1NdU2A2tpApB/tlB22Z2OFIS8Pyuuz2XNPIS9XyMuz+tQ6HKlLaHf15MRd9+DMQtwi3FG7Fq84mJXdvxbWuIiTxtwdaczdEU9bDYJQWvlfpn51L1O/upe6vGlsLT2IqtID8GWOSXjzG0d/a8BC9XjclBS5KSnK67Ge3x+gsclnd5ewWoobm6z5xiYfK1dv4uMvVtHY1EIo1PWqQrfbRV5OJrk5WeTmZPLl1xsoKylgVFkh5aWFkX7F+blZQ+bL0kg0HM9lqv/0dsFKqRGnfeW7VN91BMHtv8nyvW9g1bqtvPOvf7K+vp0FO63HU51B4WdW0ucQyM6BvFyrdTY3j0hS6/GMjKTjiKxirq5awedtTVxfMpkDM1MzrrCnvZbcxpXkNX6Ft826KLAlczQ1hftQXbQ3tYV7EHT17WLAoSJkDD5fW4dEuKHDvDVtbWunscnXZX1vhpvy0kJGldmPyHwRo8uLGFNeRFlpAR63tlWpkW0gbxesya5SqldH/eAa/vP0TakOo4PHn30TYwwNjS1U1TRQWVNPVU0DpZ5Wzm+/k6Z2B6d9cjhNQevCrGyP4Njex4oJVZzUMIF9s/MYU+CiLNfJPF/X0RhO3zqPx8qOBuCJhk2cmjfQIxr0XaLxHJdTSksoyOWVK/i6vYWbSndgurfnn/cHm7u9jpzmtWQ3ryOrZQNO48cg+DJHsbl8FrUFu9OQtxMhZ9cL6/ZddCGL9v1zCqIeWB26UER3o2hsob6xhYbGZuobWwgEgl3WzcnOJD8vi/zcbPbdawpj7ER4tP0YVVZEpnfgL0oc7obiuUzFNpDJrn41VEr16vNla1K271AoxKattaxdv4XV67ewZt0W1qzfwv8+W0lVbWOHMVwzXML9u79NVmY7f804j28dMyXSb3P5+qd5srGBYuNm3ATYLI1sBujauAbAKH9tZH5rsGt/zVRKNJ5/NVUC8M3MQqqC7fyyciXfzS3jxwXxX6A30PyeAmo9BdQW7gEmSKZvMznN68huWcekNU+wPY8TdLhpyNuJ2oLdqSvYnfq8nQg5M8hrWpmyuAdSPF0ojDH4Wtupb2iOJMH1Dc3U24lwTW0jf3/xnZh9irMyM8jPyyY/N8ueZpOfl8UJxxzIqNICRpUXpV23iVSey1TqaLKrlEq51rZ21m+sZO2GraxZv5V1G63p2g1bWLehkrZ2f6Su2+Vk/NhS8vNzmDxpDCVFeZQW5VFe4GXG5r8yZstmNo4+kmPzHICdFNXBW8F2tgbbOSyrKK0+3KN5HU6+m1POPxo380LjVr6ZVcQUzxDoNiBOfFlj8WWNpZIDcATbyPJVkNWygayWjUyqexLhCULioinHGid19KZXacidQnP2eJCROyKCiJCVmUFWZgajy4u6rdfe7qe+sYW6hmYrGW5ojkqMW1hfUUVzSysA/5i7ILKe1+uxEt+yIsrDo09ETcuKramOUayGM012lVK9Ki8t6Nf67f4Am7bUsH5jJesrqli/cSvrIvOVbK6s7VDf43ZRXJhLUWEuM/aZSnFhHiWFuRQX5VGQl43D0fGOWTmNK9l16a/JbllHZfF+NOTt1CWGT9sa8SBMizO5a3JkRuazh1gy1Z94sh1OTswt5++Nm7my8it+UTSJfTNT26Whs5Azg6acSZHENpz8Bp2Z5DV8iUHYefntAAScXhpzdqA5eyItWeNozhpHS9ZYWr1lIzoJ7swTx+gTgUCQhsYWZuwzlc1b69i8tYbNldZ0S2Udny5ZxZbKenytbV3WdTod5GRnkpPtJTc7k5xs66YdMw/cnWL7vVlcmBeZH6p9ivt7LlPDk/bZVUr1izGG2vomNm2poWJzDZu21LBhU1XksbGiis2VdUSfa0SEgrxsCgtyKCrIpbAgZ9sHZWEu2Vne+FqRTIgJ659j8qq/0e7OZ+m0S8ls3dylWk3Qz/crPmX3jFy+mdV961g6qQv6eaOlhnWBVo7OLua8gvHkOoZmgtKFMXjaa/G2biGzdSve1i24A024A02RKkGHG1/mGOvhHUWrtxyft5zWTGs63C+EGyzGGNraA/aIE9YQbA1N1njFTfbNO5qafDQ1WyNTxBp5AiAvN4si+/1dVGB9cS0syKWoIIfCfOv2z4X5297/hQW5QzZBVqmhfXaVUknR4mujsrqeO+9/gW8euDtbq+rYvLWOLZW1bKmspWJzDRVba2jt1F/Q4bCS2YL8HMaMKmbnqdtFPuCKCnIoyMvB6XR0s9f4ZLRWsvOy2yiq+4StpQeybOpFBNx5ZFa80qFeVbCdG6q+JgTskZEb9/a/0fA57+XtBsBCXx0HZBb0K96BNBDxFDjd3DdqZx6tr+CZxs0s8jVwcdF2fGMIvc7ulFR/QFXJ/rRnFNGQP80qNAZn0IenvQ5Pey0Z7bV4/LXkNqyguHoxDuPvsA2/Kwdf5mjr4R3VYdqWUYrp5XbMI5WI4M1w483ofYzicH/ipuZWmlt89rQ1Mm32WfNbKuto9rXS3NLWoY99Zx6Py+qy4bW6bWTa3Tem77EjBfk5FORlk5+X3WWak+3t8mtPd/5wz3Ncdv6JCf1N1PCnya5SacIYQ4uvjdq6JmrqGqmpbaSmvsma1jVSXdNAZXU9lTUNVFXXU1ldT4tv28+Zz7z4DmD9nBm+vWp+Xjb7jtmR/NxtHzz5ednk5WTG/eGTKAkFKat8h6kr7kZMgKU7/ZxNo46EGC3Bn7Y2cmP11/hMiGOySyhyxn+72gMbv4gku++31g+pZHeg4nm1uZpRrgy+nzuK11qq+VXVSnbyZDEzs4jv5Y3qf6CDpLT6Q6pK9u9YKELQlYXPlYUvq9PYvcbgDLbi9jfgDjTg9jfgdxeQ2bqZ3MavKa18D4fZloSFxEmrdxQtmWPwZY6mJSvcQjyaVm95zBEi0lF0f2KIryuM3x+g2ddGS0srLb42mu1p54evtY0tlXW0+NpY/MlXBKNvU9iJwyHk5Vjno7zIxXjWOSp8rsrLtcpuv/d5Dtx3Z+vW0DmZ5OZmkZudqXfCG+E02VVqmPD7A9t+QmxpjYzjGR7D07oVqjX/6dLV+Frb8fna8LW20+Jro7W1vdsPjPCHVq7dJ68wP4fxY0vJyfKSm5PJP+Yu4Ofnfoe83CyyMjNScqGK17eJMZv+w+hNr+Ftr8bnLWfj6KMQE2TMpn93qGuM4dmmLdxft4Exrgz+UDKVL9qautmyGuXK4P9yR/Nhaz0fttazzt9KAMNBmYWMd3tTHV7/iRB0ZRJ0ZdJKeaS4MXeyNWNCuALNePz1uNvr8fjr8fjryGleQ2HtJzg7tQq3eort7hBWS3Crt5y2jBJaM0poyygh6MpK5qsbVtxuFwVuFwV58XcjMcbg9wdosc9pLa1t+Hzt+Frbtp3n2tqtMvti169a22lta8fX2k57p9bkE8/5TZd9ZGVmkJeTRW5uZqRPcl5ultU3OWdbH+WcbG9UmZfsLLssy0t2dibeDLdeyDcEabKr1CAJt6SG+7pZP+9Z800trTQ3W0lrU3On+ZZtfeOam7eVtbb5e92nwyF4Mzxkej1kejPIzPRQkJ9DptdDlv08O8tLVmaGPfWSnZVBptfTY0vsP+Yu6PFK8MHiCLZTWvUuYyr+Q1HdJxgcVBftw4oxP8HdXgvSNeaWUJA/1KzhbV8tB2UWcEXRJLIdTk12e+EU4YDMAnZwZ/Gmr4YH6zfyYP1GJrq8HJRVyMGZhUx2Z47MD3JxEHDnEnDnQta4jsuiuke4/Q1WQuxvwN1eT3bzelyBJoSO/VYDzixavaW0eYpp9xTS7img3V2A35MfmQ+4cwg4swi4sjAObSnuiYjg8bjxeNwJJclhwWAokvj+/u5n+dFpR9NqJ8OtbX5aW9uZOKHc6ods3xmvqdnHpq01NDb6qK1v6jAiTE+cTkck8c3OyiA700tOtpesLCshjpx7I/MZUefjjt03rIdV7nY5R+Z7L0mGzQVqIjIeuAM4AhDgdeBiY8y6eNbXC9RUd8Inwla7BTT6Z7To580tbbT4tv30Zj2sn+OaIs+jHs2txPvu8rhdZGS4yfDYj4yOU2+GB2+G9dyb4YksC5d7Mzx4vZ5BOyFu2FTFuNElA77dzlz+RvIaviSv8UvyG76ksPZTnKE22t151OXvTH3eNCshicEYw/L2Zq6t+prakJ+DMgvYJyOvz3+P8vYatnisBH9LoI1yV0afX9dAS0Y8DaEAX7e3sNLfwsZAGwbIcziZlVXMnhm57JaRS74zNe0l3tat1mgLQ4EJ4vY34bIvkHP57WmgCSMuPP46PO11OENdRzgIC4mLgCuLoDOLgNNLyOkl6PASdGYQcmYQdFhTvyuHgCubQNTU78rG77aS6JBzBLTCD7K+nstCxtDe7qetzW8lyG3ttNnP29r9Xebb2wP21E+bPR8ub/db00Q4nQ4rEfZ6yMrMoLXNj8ftwm0/PC5nZH6PnSfh9VqNHtZnhIfMzG3zXq+bDI81zYyUWZ8rg/k5kqi0u0BNRLKAN4E2YA5ggJuAt0Rkd2NMcyrjU30TDIbwBwL4/QHa/UH8/vC89fD7rRNEeHl7u99aZp88wieMLicaexr+1t4W9e09fILy+bb9xBXvN/Ywh0PIzvKSneklEAySYbc4ZGS4yc3JpKQoz3oenbhGJ7OdklqP2zVo/VuHIgkF8bRXR66i97ZuJatlHfkNK8jybQTAIDRnjacxdzL1uVNpyRoXs09uuwnxaWsj77XWsdBXR2XQT6Y4+G5OGRPcmV3qq/jlOVzs5c1jL28eLaEgq/wtrGz38UpzFc83bQVge3cme2TksmdGLlM92WSIA4eAE0EAhwhOBOcQ+OAcNOLE78nH78nv7v4kVrWQH1egBWfQhyvYgiPYhjPUjiPy8OMIteMMteMMtOA2DUgogMP4rWnIjzPU1qUVOVrQkWG3HOfj9xTgd+fhd+Xid+cQcOXid+fid+UScGUTdGYSdHoj05DDE/M9piwOkUiyOBAD9YWMIeAPWJ9l4c82f9R8e8fPwg7z9uefP2B9bjY2teC3P0Pb/QH+99lK/IEgfW3MlMhrtRtb7OQ40vgSTow7f6ZluPF63GRENchYn3UuMjI8XRp1PB77udua93isz0qPxz3gfaiHRbIL/AjYHphqjFkJICKfAV8BPwZuT2FsKWOMIRQyBIJBgoEg/kCQQDBEIBCwp8HIo90fIBCuEwjY02DkzRM93+636ljPw/NB/IFAN/UDdqIatJPXTtu1E9rwGzO8TndD1vSV0+nA5XTicjlxu5w47ak7/I3XZS3Lz82mpCjf+lYc9W3Y7XZGvil77DecJzxvv0k9bheuIfKtN5n++OBL/P5XZ3ddYF/44wy24Aq04Aq24Ao04/bX4/bXW1fGR/pB1pHRVkVGWxUO07HvsN+Vjc87iq0l37CGh/KWEXJ2bLkMGkNlsJ1NgTYqAm0sbm1gUWs9PhPCKw728eZxpreAplAA7wBcSX9G5X+4dewpADzZuJmfF27X720OlGTHk+VwsmtGLrtm5BI0hs3BNjb429gQaGVuU2Uk+e1OscPNWHcGY1xexroy7IeXUqebHIcr4WR40tqnWTb1Z/15SUlnHG4rKe5PqmQMYvw4w4lysA1nqBVnsDWSSFvJtA9P02oQsVqdgy29bxpHJOkNOj2EHB5CjgxCDjdBRwbG4SIkrm1TcRFyWNNtZc4O5SGHm5DDRUjcGIfbei7W1DjchCJ1osvcHcoGeqzkbs9lSeaI6poBA//F3BgTaVBq9wcJhD+DAx0/q/2B6NzAevj9AfzBIAF/x/L2dj8tLa0d6waj5gPWfgbik93hGNjP2OGS7M4G3g8nugDGmNUi8i7wbeJIdiur6/nLI/8iFDKRJDEUChGy5409H1lmDKFgeHloW317WTAYijwPhsLzIYJBq962sm11g9H1g9Zza5lVp0O5vSwYDEa2GQgG7bJt88kggNPlxOlw4HQ6IklleH5buROXXeZwOCLDyETq2XWcDgdOh+B0het33FZ42y6X06rndOCO7NeBy+XA5XDgdtn7dFnrOESgw9usP2856TJvBCAEJhR+0mFfYrbNh/ctJmQvM5Hy8HxkGQY6lxkDhBBjEELWh5w93bZOyIosetuREEI4TBAxAcQEkVAQMUEcJrCtJSnYZrcWteMItdktS1brksNEzYf8PLz7RiYt+iiqFard+sAN+hAMVrQQsqIkZLc/hRACTi/tzix8Li+1riJaMifgc+XQ4srG58zC58yiDQdNJkhzyHo0NVXTbII0hIJsCbSxOdDGlmA70f/js8XJDu4sJnuyGO/KwCUOApgBSXRV95wijHV5GevyMoP8SPJbGfATsv8vmPDUQBBDQyhAVdDPynYfLSbYZZs54iTP6SLP4SLP4STb4cKF4BLZNrVbiYMY8twTqaxZSwBDwBiCGBwITrtV2Rm1nkccZImTLEd46iRbnGQ4HFZCAIQwBI2x/+92f94Qu8VaAAeCCDjseWe4FdvebziWbfXs1m57OxA+S0T/a3EgkRby8Lwj/O4WNwGX2/4Lb1uvx7OdCeIKtuMKtuIMteEMtkXe4xJ531utyM5QACcB6/wRCuAMNFutzCZonZ/sqRCKlIUfEF42cELiwIjbSsId4SQ8nAx77GTbiRHntoTb4Yo+GxJ9Pr9pyip2XloF4rDqiGBwWFPptB1xWNu3y7c9HFZCjwMjTmtbkfrWFKTL9umxXR77GoRt64Xnre06tu2H8HO7Lg4Q6wvLtlccsv9ThBCnAYdBMsKfJeHPj6jPERz2+m7r7yXh/6Oy7W/V4fXQad/bXl8oBH47TwkEjD0fIhA0VmIcChEI2GUBu14oRCgQtBvvQgRCIR74pF//dToYLsnuLsCLMcqXAN+LZwMbfJu54ovfDWhQMTnsh0pM0H6olOj4FSH6Q8KNwU2ofEckZJ3IjP1/3LiIJLRxf60IYXVGamsH2oHabqs6EbziIM/hJM/hYrzbaydDLvLtR7q1sA9F0clvPNpNiLpggLqQn+ZQEJ8J0WpCtIaC+EJBaoJ+2k3ITkCt5DN63oGQ6SrD31IdSXAd9v/BcPJq7HVCBgKRtFB1T7CSnG1D84WTeIdgNyRs0/F8YTBm21d8E/X37vDFIDK/rR5YpwTTaZvb9tbxyEVvQ0IGCRHVyBAWPsnEZrYbjTRoz8chy2k/BtiwuEBNRNqB240xV3Uqvwm4yhgTM2kXkXOBc+2nU4EvBzXQ1CoBqlIdhOoXPYbDnx7DkUGP4/Cnx3D4m2qMif9OQD0YLi27fWKMuR+4P9VxJIOILB6oqxZVaugxHP70GI4MehyHPz2Gw5+IDNgQWsPlB/daoDBGeRE9/Q6qlFJKKaXS2nBJdpdg9dvtbGdgaZJjUUoppZRSw8RwSXZfAvYXke3DBSIyETjQXqbSpLvGCKfHcPjTYzgy6HEc/vQYDn8DdgyHywVq2cCngA+4BusSzRuBXGB3Y4zeB1QppZRSSnUxLFp27TukHQasAB4DngBWA4dpoquUUkoppbozLFp2lVJKKaWU6oth0bKruhIRh4hcLSJrRKRVRD4VkRP7sJ3tRaRFRIyI7DAYsarY+noMRSRPRK4VkfdEpFpE6uz57yQh7LQkIuNF5FkRqReRBhH5p4hMiHNdr4jcKiKbRMQnIgtF5JDBjll11NdjKCLTReR+EVlunyvXicgTIjIpGXGrbfrzPuy0navsz7wFgxGn6l5/j6GITBORf4hIlX0+/VJELuptPU12h68bgeuBPwPfAt4H/iEixyS4nb8A9QMbmopTX4/hBOAnwNvAacD3sbr4PC8iFwxatGlKRLKAN4GdgDnA6cCOwFv29QS9+SvwI+Ba4DhgE/AfEdlzUAJWXfTzGP4AazSgP2K9T68C9gYWi8j4QQtadTAA78PwdrbHuvZn62DEqbrX32MoItOBD4AM4IfAMcBtxHPPNWOMPobZAyjDuh/irzuVvwF8lsB2/g/YAlyMddHfDql+beny6M8xBLKBrBjlbwDrUv3aRtoDuAjrZtY7RJVNAgLAJb2su4f93jorqsyFdTfHl1L92tLl0c9jWBqjbDus+9LekOrXli6P/hzDTtv5D3AfMB9YkOrXlU6Pfr4PHVhDzT7fl31ry+7wdBTgAR7vVP44sFs8P6+JSCFwO3AZUDfQAape9fkYGmOajTEtMRYtBsYMXIjKNht43xizMlxgjFkNvAt8O451/cAzUesGgKeBo0QkY+DDVTH0+RgaYypjlK0FKoGxAxyn6l5/3ocAiMj/YbXKXz0oEare9OcYzgSmYeUtCdNkd3jaBatVcGWn8iX2dOc4tvF7YLkx5rGBDEzFbSCOYWeHAMv7E5SKaRfgixjlS+j9OO0CrI7x5WQJ1pcd7SefHP05hl2IyDSsX2eW9TMuFb9+HUO7gecO4ApjTM0Ax6bi059jeJA99YrI+yLiF5GtIvJHEcnsbceuBANVQ0MRUGfstv0oNVHLuyUiBwNnAHsNQmwqPv06hp2JyLnA/lh9eNXA6u625DXEvo15vOuGl6vB159j2IGIuIB7sVp2/9r/0FSc+nsMb8W6tuGRAYxJJaY/xzD8q+UzWNe5XAVMB24AxgMn9LSyJrtDgIgcDrwWR9W3jTEz+7kvD1Z/pTuMMXqr5QGSzGMYY98zsS6eedQY88RAblsp1cWfgW8AxxpjYn1wqyEmqoFn7xgNDGp4CPdEeNwYc609P19EnMAtIjLNGNPtLy2a7A4N72H1RelN+KfQWqBARKTTGzfcStTTTzQXY32D+qOIFNhlWfY0V0RyjTGNcUWtoiXzGEaIyL5Yt8x+E+vqVDXwaond6tBdK0XndbfrZl2I8zirfuvPMYwQkVuAc4E5xphXByg2FZ/+HMP7sFrhN0R97rkAp/3cZ4xpG6A4Vff6cwyr7WnnRqVXgVuwfqnWZHcos/vzJdLXcgnW0BuT6djnM9znpacW252BUcDGGMv+h3Vb5j0TiEWR9GMIgIjshnVl8SfAicYYfwL7V/FbgtXXrLOd6f04LQFOEJGsTv12dwba6dpnWw2O/hxDAETkl8CVwE/1WoeU6M8xnGY/zouxrBb4OXBnf4JTcenvubQnoZ4W6gVqw9M8rCu8T+1UfhrwhX11Y3duAb7Z6fG7qPW1dTA5+nMMEZEdsb7hrgKOM8b4BiVKBVbL+f72+JwAiMhE4EB7WU/mAm7ge1HrurDGRn5VW5OSpj/HEBH5GXAT8EtjzJ8HK0jVo/4cw86fed/Eatj5wp5/dhDiVV315xj+G+ui7qM6lR9tTxf3uHaqx13TR5/Hq7sFaAUuwRqS4x6sbzbHdar3BrCyl22diY6zO2yOIdZV4GuwfgI/FuvCtOhHRqpf20h6YI1rvBL4HGt4nNlYH5SrgJyoetthjRd5baf1n8ZqPfohMAvrg7UVq/9gyl9fOjz6cwyxbioRsj9sO7/Xdk71a0uXR3/fhzG2Nx8dZ3dYHUPgOrv8t8DhWBep+YBHetu3dmMYvn4JNGEN0jwKa5D6k40x/+pUz4l2Vxmq+noMd2ZbP9DOdcEapHvNgEaaxowxzSJyGNawRY8BgvUF5GJjTFNUVcE6Vp1/MTsL+A1Wy2AB1sn9aGPM/wY5dGXr5zE82i4/mm2tSGFvY31RVYNsAN6HKsUG4BjeADRi3UH0Mqy7Ud6KdTfSHomdLSullFJKKTXi6DcfpZRSSik1Ymmyq5RSSimlRixNdpVSSiml1Iilya5SSimllBqxNNlVSimllFIjlia7SimllFJqxNJkVymlBoCITBQRIyKPpGDf19v7npnsfQ+WvrymVB4DpdTQpcmuUiql7OQkkceZqY5ZpY79f2B+quNQSg0femctpVSq/TpG2cVAPnAXUNdp2SeDG06fbQSmAfWpDkQppdQ2muwqpVLKGHN95zK79TYfuNMYsybJIfWJMcYPLE91HEoppTrSbgxKqWFFRGaIyLMisllE2kVkvYjcJyJjYtSdb//s7RGRa0XkSxFpC/fpjFrutpd/LSKtdr0fRW3nPBH5XER8IrJBRH4tIo5O+4rZX1REHrHLJ4rIj+3ttIrIFhG5X0TyY8T9TXvZUhFpsPf7hYhcJyLefv79brbjOaJT+a/t8q9jrLNZRNbFKD9KRF4RkSr77/q1iNwqIgUD/ZpE5EwRCd/f/tBOXVuuj1F/oog8bcfWKiKLReS43vajlBp5tGVXKTVsiMjZwP1AG/ASsB7YEfghcLyI7G+M6ZKUAc8B+wL/Bl4AtnZa/jQwA3gF8AMnAfeLiB/YHZgD/At4A5gNXAu0AL9LIPzfA0cBc4FXgW8CPwJ2AA7rVPdKYCfgPeBlwAscCFwPzBSRw40xwQT2He0N4CpgFvBaVPkse7q9iEwMt6iLyK5AOfBI9EZE5Do7nhqsv81WrL/VZcAxInKAMaZhAF/TJ1hdXq4D1naKZ36nutsBHwKrgMeAIuD7wIv2ft7qYT9KqZHGGKMPfehDH0PqAawBDDAxqmwK0A6sBMZ2qj8LCALPdyqfb2/nM6Akxn7CyxcBBVHl29v7qgVWR+8PKACqgErAFVU+0d7WI5328Yhdvg6YEFXuAt6xl+3XaZ3tAYkR7412/e93Kr/eLp8Zx982E2gFFkWV5div91V7O+dELbvILjs9quybdtl70X83e9mZ9rI7BuM12WXzu3lt4WNggOs6LTvKLn8l1f+/9aEPfST3od0YlFLDxfmAG7jIGLMxeoEx5g2slt7jRSQ3xrq/MsZU9bDtq4wxdVHbWwUswEpsb4zen11vLlACjE0g/htMVKuzMSYAPGw/3a/T61lljDF0dYc9PSqB/XZgjPEBC4G9RaTQLj4E6297J1YSPytqlfD8G1FlP7OnP4r+u9nbfwSrFfbUTuWD9ppiWAvc1Gn//8H6wrFfzDWUUiOWdmNQSg0XB9jTQ0Vk3xjLywAnVgvwR52WfdjLthfHKKuwp523BdbICwDjsBKreMTax3p7WhhdKCLZWC2qJ2C9nlxAoqokkmTH8iYw0348j9WNwg+8DbxlP0dEnFiJ8JfGmIqo9Q+w639PRL4XY/seoFREio0x1Ul6TdE+MbG7RKxn2/8jpVSa0GRXKTVcFNvTy3uplxOjbHNPKxhjYg0XFrCnPS1z9xJLtLoetuMMF4iIGysZ3Q/4AngGq7XVb1e5DshIYL+xvAHcgNVq+7w9/cAY0ywibwAn2311s7FGxXii0/rFWJ8f1/WynxygOkmvKVpdN+UB9MJspdKOJrtKqeEinHTmm44XPvWqm5/Ph6pvYyWFjxhjzopeICKj6T3BjMeHQBNwuIgUA3tgJb9gJaUAhwNZncrC6gGHMaYozv0l4zUppVRM+g1XKTVcvG9PD05pFINvB3v6zxjLDh2IHdj9hd8BpgKnY3UneMNethKrb+ssrO4MIayuDdHeBwpFZJc4dzmQrylEVEu4Ukr1RpNdpdRw8Wesn73vEJEpnRfaY+mOhER4jT2dGV0oItuT2FBnvQm31l4NNLPty0R42aFYQ4N9aoyp6bRu+KKyB7oZ3zhbRPaPKlpjT2d2qteX11QNjE9wHaVUGtNuDEqpYcEYs9weZ/chYImIzANWYPWbnYDV4luJNZbrcDYXa3i1S0RkN+BjrNd3HNb4tBMGaD/h0RXKgHnGugNc9LIzO9WLMMa8ISJXATcDX4nIK1hDtOVgjXF7KNZoFkcPwmt6A/iBiMwF/of1BegdY8w7CWxDKZVGNNlVSg0bxpjHReRT4FKssV6PxGqVrACexbrwaVizLxI7DLgFqyX0YKybI9wI3I51c4SB8CnWeMEldO2T+2Y389Fx/k5E3sUahuwgrH659VgjVdwPPBlVdyBfU3jc31nAMVi/UP4aq1uGUkp1IcPrug2llFJKKaXip312lVJKKaXUiKXJrlJKKaWUGrE02VVKKaWUUiOWJrtKKaWUUmrE0mRXKaWUUkqNWJrsKqWUUkqpEUuTXaWUUkopNWJpsquUUkoppUYsTXaVUkoppdSI9f/UbdIYPFyrXgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot rewards instead of costs\n",
    "rewards_total = -1 * np.sum(costs, axis=1)\n",
    "\n",
    "# set a grid for the histogram\n",
    "grid = np.linspace(-0.5, 0.6, 100)\n",
    "\n",
    "##############################################################################\n",
    "###########################     PLOT: TERMINAL COST     ######################\n",
    "\n",
    "plt.figure(figsize=(10,4))\n",
    "for idx_method, method in enumerate(rm_list):\n",
    "    # plot the histogram for each method\n",
    "    reward = rewards_total[:,idx_method]\n",
    "    plt.hist(x=rewards_total[:,idx_method],\n",
    "            alpha=0.4,\n",
    "            bins=grid,\n",
    "            color=utils.colors[idx_method],\n",
    "            density=True)\n",
    "\n",
    "plt.legend(rm_labels)\n",
    "plt.xlabel(\"Terminal wealth\")\n",
    "plt.ylabel(\"Density\")\n",
    "plt.title(\"Distribution of the terminal wealth\")\n",
    "\n",
    "for idx_method, method in enumerate(rm_list):\n",
    "    # plot gaussian KDEs\n",
    "    kde = gaussian_kde(rewards_total[:,idx_method], bw_method='silverman')\n",
    "    plt.plot(grid,\n",
    "            kde(grid),\n",
    "            color=utils.colors[idx_method],\n",
    "            linewidth=1.5)\n",
    "    # plot quantiles of the distributions\n",
    "    plt.axvline(x=np.quantile(rewards_total[:,idx_method],0.1),\n",
    "                linestyle='dashed',\n",
    "                color=utils.colors[idx_method],\n",
    "                linewidth=1.0)\n",
    "    plt.axvline(x=np.quantile(rewards_total[:,idx_method],0.9),\n",
    "                linestyle='dashed',\n",
    "                color=utils.colors[idx_method],\n",
    "                linewidth=1.0)\n",
    "\n",
    "plt.xlim(-0.5, 0.6)\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "###########################     PLOT: TERMINAL COST     ######################\n",
    "##############################################################################"
   ]
  }
 ],
 "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}