Project-G2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import os\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from scipy import stats\n",
    "from statsmodels.tsa.stattools import adfuller\n",
    "from sklearn import preprocessing\n",
    "from tensorflow.keras.preprocessing.sequence import TimeseriesGenerator\n",
    "import warnings\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "import math\n",
    "warnings.filterwarnings('ignore')\n",
    "from statsmodels.tsa.stattools import pacf\n",
    "import datetime\n",
    "%matplotlib inline\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras.optimizers import RMSprop, Adam,SGD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_series(df,format='-',start = 0, end = None):\n",
    "    plt.plot(df['Umsatz'],format)\n",
    "    plt.xlabel(\"Time\")\n",
    "    plt.ylabel(\"sales\")\n",
    "    plt.grid(True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "dataframe = pd.read_csv('meteolytix_Corona_bakery_turnover_updated.csv')\n",
    "\n",
    "dataframe['Datum']=pd.to_datetime(dataframe['Datum'])\n",
    "\n",
    "dataframe['Quarter'] = dataframe['Datum'].apply(lambda x: x.quarter)\n",
    "dataframe['DayofTheWeek'] = dataframe['Datum'].apply(lambda x: x.weekday()+1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "cdf = pd.read_csv('covid-data.csv')\n",
    "cdf['date'] = pd.to_datetime(cdf['date'])\n",
    "\n",
    "cdf.rename(columns={'date':'Datum'},inplace=True)\n",
    "\n",
    "cdf = cdf[['Datum','total_cases','new_cases','total_cases_per_million','new_cases_per_million']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataframe = dataframe.join(cdf.set_index('Datum'),on='Datum')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataframe[['total_cases','new_cases','total_cases_per_million','new_cases_per_million']]= dataframe[['total_cases','new_cases','total_cases_per_million','new_cases_per_million']].fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "prodG2 = dataframe[dataframe['Warengruppe'] == 2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "prodG2.drop('Warengruppe',axis=1,inplace=True) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "prodG2.set_index('Datum',inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1a346960dc8>]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1224x1152 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(17,16))\n",
    "fig.subplots_adjust(hspace=.4)\n",
    "ax1 = fig.add_subplot(5,1,1)\n",
    "ax1.plot(prodG2['Umsatz'].rolling(window=1).mean(),linewidth=1)\n",
    "ax1.set_title('Mean Global sales resampled over day')\n",
    "ax1.tick_params(axis='both', which='major')\n",
    "ax1.plot(prodG2['new_cases_per_million'])\n",
    "\n",
    "ax2 = fig.add_subplot(5,1,2, sharex=ax1)\n",
    "ax2.plot(prodG2['Umsatz'].rolling(window=7).mean(),linewidth=1)\n",
    "ax2.set_title('Mean Global Sales resampled over week')\n",
    "ax2.tick_params(axis='both', which='major')\n",
    "ax2.plot(prodG2['new_cases_per_million'].resample('W').mean())\n",
    "\n",
    "ax3 = fig.add_subplot(5,1,3, sharex=ax1)\n",
    "ax3.plot(prodG2['Umsatz'].rolling(window=30).mean(),linewidth=1)\n",
    "ax3.set_title('Mean Global Sales resampled over Month')\n",
    "ax3.tick_params(axis='both', which='major')\n",
    "ax3.plot(prodG2['new_cases_per_million'].resample('M').mean())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>Datum</th>\n",
       "      <th>Umsatz</th>\n",
       "      <th>Quarter</th>\n",
       "      <th>DayofTheWeek</th>\n",
       "      <th>total_cases</th>\n",
       "      <th>new_cases</th>\n",
       "      <th>total_cases_per_million</th>\n",
       "      <th>new_cases_per_million</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-02</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-03</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-05</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-06</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Datum  Umsatz  Quarter  DayofTheWeek  total_cases  new_cases  \\\n",
       "0 2015-01-02     0.0        1             5          0.0        0.0   \n",
       "1 2015-01-03     0.0        1             6          0.0        0.0   \n",
       "2 2015-01-05     0.0        1             1          0.0        0.0   \n",
       "3 2015-01-06     0.0        1             2          0.0        0.0   \n",
       "4 2015-01-07     0.0        1             3          0.0        0.0   \n",
       "\n",
       "   total_cases_per_million  new_cases_per_million  \n",
       "0                      0.0                    0.0  \n",
       "1                      0.0                    0.0  \n",
       "2                      0.0                    0.0  \n",
       "3                      0.0                    0.0  \n",
       "4                      0.0                    0.0  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "meanprodG2 = prodG2\n",
    "meanprodG2['Umsatz'] = meanprodG2['Umsatz'].rolling(window=14).mean()\n",
    "meanprodG2['Umsatz'] = meanprodG2['Umsatz'].fillna(0)\n",
    "meanprodG2.reset_index(inplace=True)\n",
    "meanprodG2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "    dataset = meanprodG2.Umsatz.values \n",
    "    dataset = dataset.astype('float32')\n",
    "    dataset = np.reshape(dataset, (-1, 1))\n",
    "    scaler = MinMaxScaler(feature_range=(0, 1))\n",
    "    dataset = scaler.fit_transform(dataset)\n",
    "    train_size = 1300\n",
    "    test_size = len(dataset) - train_size\n",
    "    train, test = dataset[0:train_size,:], dataset[train_size:len(dataset),:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_dataset(dataset, window_size):\n",
    "    X, Y = [], []\n",
    "    for i in range(len(dataset)-window_size-1):\n",
    "        a = dataset[i:(i+window_size), 0]\n",
    "        X.append(a)\n",
    "        Y.append(dataset[i + window_size, 0])\n",
    "    return np.array(X), np.array(Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "window_size = 14\n",
    "X_train, Y_train = create_dataset(train, window_size)\n",
    "X_test, Y_test = create_dataset(test, window_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(347, 14)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1285,)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = np.reshape(X_train, (X_train.shape[0], 1, X_train.shape[1]))\n",
    "X_test = np.reshape(X_test, (X_test.shape[0], 1, X_test.shape[1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv1d (Conv1D)              (None, 1, 32)             2272      \n",
      "_________________________________________________________________\n",
      "lstm (LSTM)                  (None, 1, 50)             16600     \n",
      "_________________________________________________________________\n",
      "lstm_1 (LSTM)                (None, 50)                20200     \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 128)               6528      \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 1)                 129       \n",
      "_________________________________________________________________\n",
      "lambda (Lambda)              (None, 1)                 0         \n",
      "=================================================================\n",
      "Total params: 45,729\n",
      "Trainable params: 45,729\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 1285 samples\n",
      "Epoch 1/100\n",
      "1285/1285 [==============================] - 4s 3ms/sample - loss: 0.4614 - mae: 0.6255\n",
      "Epoch 2/100\n",
      "1285/1285 [==============================] - 0s 332us/sample - loss: 0.4588 - mae: 0.6248\n",
      "Epoch 3/100\n",
      "1285/1285 [==============================] - 0s 238us/sample - loss: 0.4644 - mae: 0.6273\n",
      "Epoch 4/100\n",
      "1285/1285 [==============================] - 0s 215us/sample - loss: 0.4490 - mae: 0.6181\n",
      "Epoch 5/100\n",
      "1285/1285 [==============================] - 0s 194us/sample - loss: 0.4595 - mae: 0.6266\n",
      "Epoch 6/100\n",
      "1285/1285 [==============================] - 0s 195us/sample - loss: 0.4345 - mae: 0.6090\n",
      "Epoch 7/100\n",
      "1285/1285 [==============================] - 0s 205us/sample - loss: 0.4398 - mae: 0.6097\n",
      "Epoch 8/100\n",
      "1285/1285 [==============================] - 0s 194us/sample - loss: 0.4551 - mae: 0.6255\n",
      "Epoch 9/100\n",
      "1285/1285 [==============================] - 0s 194us/sample - loss: 0.4413 - mae: 0.6142\n",
      "Epoch 10/100\n",
      "1285/1285 [==============================] - 0s 195us/sample - loss: 0.4215 - mae: 0.5955\n",
      "Epoch 11/100\n",
      "1285/1285 [==============================] - 0s 201us/sample - loss: 0.4340 - mae: 0.6062\n",
      "Epoch 12/100\n",
      "1285/1285 [==============================] - 0s 187us/sample - loss: 0.4179 - mae: 0.5969\n",
      "Epoch 13/100\n",
      "1285/1285 [==============================] - 0s 208us/sample - loss: 0.4054 - mae: 0.5839\n",
      "Epoch 14/100\n",
      "1285/1285 [==============================] - 0s 191us/sample - loss: 0.4098 - mae: 0.5849\n",
      "Epoch 15/100\n",
      "1285/1285 [==============================] - 0s 199us/sample - loss: 0.3970 - mae: 0.5791\n",
      "Epoch 16/100\n",
      "1285/1285 [==============================] - 0s 196us/sample - loss: 0.3700 - mae: 0.5537\n",
      "Epoch 17/100\n",
      "1285/1285 [==============================] - 0s 191us/sample - loss: 0.3832 - mae: 0.5648\n",
      "Epoch 18/100\n",
      "1285/1285 [==============================] - 0s 191us/sample - loss: 0.3516 - mae: 0.5331\n",
      "Epoch 19/100\n",
      "1285/1285 [==============================] - 0s 207us/sample - loss: 0.3395 - mae: 0.5279\n",
      "Epoch 20/100\n",
      "1285/1285 [==============================] - 0s 195us/sample - loss: 0.3238 - mae: 0.5098\n",
      "Epoch 21/100\n",
      "1285/1285 [==============================] - 0s 208us/sample - loss: 0.3073 - mae: 0.4954\n",
      "Epoch 22/100\n",
      "1285/1285 [==============================] - 0s 201us/sample - loss: 0.2900 - mae: 0.4804\n",
      "Epoch 23/100\n",
      "1285/1285 [==============================] - 0s 202us/sample - loss: 0.2869 - mae: 0.4797\n",
      "Epoch 24/100\n",
      "1285/1285 [==============================] - 0s 194us/sample - loss: 0.2559 - mae: 0.4459\n",
      "Epoch 25/100\n",
      "1285/1285 [==============================] - 0s 195us/sample - loss: 0.2519 - mae: 0.4455\n",
      "Epoch 26/100\n",
      "1285/1285 [==============================] - 0s 204us/sample - loss: 0.2278 - mae: 0.4123\n",
      "Epoch 27/100\n",
      "1285/1285 [==============================] - 0s 209us/sample - loss: 0.2111 - mae: 0.3978\n",
      "Epoch 28/100\n",
      "1285/1285 [==============================] - 0s 199us/sample - loss: 0.1901 - mae: 0.3714\n",
      "Epoch 29/100\n",
      "1285/1285 [==============================] - 0s 207us/sample - loss: 0.1587 - mae: 0.3371\n",
      "Epoch 30/100\n",
      "1285/1285 [==============================] - 0s 213us/sample - loss: 0.1496 - mae: 0.3247\n",
      "Epoch 31/100\n",
      "1285/1285 [==============================] - 0s 211us/sample - loss: 0.1368 - mae: 0.3071\n",
      "Epoch 32/100\n",
      "1285/1285 [==============================] - 0s 206us/sample - loss: 0.1184 - mae: 0.2775\n",
      "Epoch 33/100\n",
      "1285/1285 [==============================] - 0s 202us/sample - loss: 0.1016 - mae: 0.2533\n",
      "Epoch 34/100\n",
      "1285/1285 [==============================] - 0s 205us/sample - loss: 0.0934 - mae: 0.2412\n",
      "Epoch 35/100\n",
      "1285/1285 [==============================] - 0s 199us/sample - loss: 0.0919 - mae: 0.2350\n",
      "Epoch 36/100\n",
      "1285/1285 [==============================] - 0s 201us/sample - loss: 0.0750 - mae: 0.2131\n",
      "Epoch 37/100\n",
      "1285/1285 [==============================] - 0s 196us/sample - loss: 0.0733 - mae: 0.2122\n",
      "Epoch 38/100\n",
      "1285/1285 [==============================] - 0s 208us/sample - loss: 0.0728 - mae: 0.2109\n",
      "Epoch 39/100\n",
      "1285/1285 [==============================] - 0s 200us/sample - loss: 0.0696 - mae: 0.2052\n",
      "Epoch 40/100\n",
      "1285/1285 [==============================] - 0s 205us/sample - loss: 0.0671 - mae: 0.1995\n",
      "Epoch 41/100\n",
      "1285/1285 [==============================] - 0s 202us/sample - loss: 0.0660 - mae: 0.1987\n",
      "Epoch 42/100\n",
      "1285/1285 [==============================] - 0s 204us/sample - loss: 0.0696 - mae: 0.2051\n",
      "Epoch 43/100\n",
      "1285/1285 [==============================] - 0s 198us/sample - loss: 0.0701 - mae: 0.2045\n",
      "Epoch 44/100\n",
      "1285/1285 [==============================] - 0s 194us/sample - loss: 0.0702 - mae: 0.2039\n",
      "Epoch 45/100\n",
      "1285/1285 [==============================] - 0s 202us/sample - loss: 0.0668 - mae: 0.1989\n",
      "Epoch 46/100\n",
      "1285/1285 [==============================] - 0s 198us/sample - loss: 0.0646 - mae: 0.1978\n",
      "Epoch 47/100\n",
      "1285/1285 [==============================] - 0s 208us/sample - loss: 0.0685 - mae: 0.2043\n",
      "Epoch 48/100\n",
      "1285/1285 [==============================] - 0s 200us/sample - loss: 0.0684 - mae: 0.2009\n",
      "Epoch 49/100\n",
      "1285/1285 [==============================] - 0s 196us/sample - loss: 0.0616 - mae: 0.1953\n",
      "Epoch 50/100\n",
      "1285/1285 [==============================] - 0s 206us/sample - loss: 0.0608 - mae: 0.1934\n",
      "Epoch 51/100\n",
      "1285/1285 [==============================] - 0s 198us/sample - loss: 0.0588 - mae: 0.1895\n",
      "Epoch 52/100\n",
      "1285/1285 [==============================] - 0s 205us/sample - loss: 0.0604 - mae: 0.1910\n",
      "Epoch 53/100\n",
      "1285/1285 [==============================] - 0s 205us/sample - loss: 0.0542 - mae: 0.1800\n",
      "Epoch 54/100\n",
      "1285/1285 [==============================] - 0s 200us/sample - loss: 0.0568 - mae: 0.1829\n",
      "Epoch 55/100\n",
      "1285/1285 [==============================] - 0s 210us/sample - loss: 0.0539 - mae: 0.1796\n",
      "Epoch 56/100\n",
      "1285/1285 [==============================] - 0s 207us/sample - loss: 0.0515 - mae: 0.1788\n",
      "Epoch 57/100\n",
      "1285/1285 [==============================] - 0s 247us/sample - loss: 0.0458 - mae: 0.1680\n",
      "Epoch 58/100\n",
      "1285/1285 [==============================] - 0s 237us/sample - loss: 0.0484 - mae: 0.1707\n",
      "Epoch 59/100\n",
      "1285/1285 [==============================] - 0s 209us/sample - loss: 0.0437 - mae: 0.1605\n",
      "Epoch 60/100\n",
      "1285/1285 [==============================] - 0s 208us/sample - loss: 0.0443 - mae: 0.1611\n",
      "Epoch 61/100\n",
      "1285/1285 [==============================] - 0s 205us/sample - loss: 0.0400 - mae: 0.1560\n",
      "Epoch 62/100\n",
      "1285/1285 [==============================] - 0s 199us/sample - loss: 0.0353 - mae: 0.1467\n",
      "Epoch 63/100\n",
      "1285/1285 [==============================] - 0s 198us/sample - loss: 0.0336 - mae: 0.1422\n",
      "Epoch 64/100\n",
      "1285/1285 [==============================] - 0s 204us/sample - loss: 0.0345 - mae: 0.1439\n",
      "Epoch 65/100\n",
      "1285/1285 [==============================] - 0s 197us/sample - loss: 0.0308 - mae: 0.1354\n",
      "Epoch 66/100\n",
      "1285/1285 [==============================] - 0s 197us/sample - loss: 0.0281 - mae: 0.1307\n",
      "Epoch 67/100\n",
      "1285/1285 [==============================] - 0s 204us/sample - loss: 0.0243 - mae: 0.1209\n",
      "Epoch 68/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1285/1285 [==============================] - 0s 204us/sample - loss: 0.0233 - mae: 0.1184\n",
      "Epoch 69/100\n",
      "1285/1285 [==============================] - 0s 199us/sample - loss: 0.0199 - mae: 0.1092\n",
      "Epoch 70/100\n",
      "1285/1285 [==============================] - 0s 197us/sample - loss: 0.0173 - mae: 0.1021\n",
      "Epoch 71/100\n",
      "1285/1285 [==============================] - 0s 198us/sample - loss: 0.0154 - mae: 0.0970\n",
      "Epoch 72/100\n",
      "1285/1285 [==============================] - 0s 192us/sample - loss: 0.0130 - mae: 0.0891\n",
      "Epoch 73/100\n",
      "1285/1285 [==============================] - 0s 200us/sample - loss: 0.0114 - mae: 0.0831\n",
      "Epoch 74/100\n",
      "1285/1285 [==============================] - 0s 194us/sample - loss: 0.0095 - mae: 0.0757\n",
      "Epoch 75/100\n",
      "1285/1285 [==============================] - 0s 198us/sample - loss: 0.0088 - mae: 0.0729\n",
      "Epoch 76/100\n",
      "1285/1285 [==============================] - 0s 193us/sample - loss: 0.0062 - mae: 0.0614\n",
      "Epoch 77/100\n",
      "1285/1285 [==============================] - 0s 198us/sample - loss: 0.0056 - mae: 0.0588\n",
      "Epoch 78/100\n",
      "1285/1285 [==============================] - 0s 188us/sample - loss: 0.0048 - mae: 0.0543\n",
      "Epoch 79/100\n",
      "1285/1285 [==============================] - 0s 204us/sample - loss: 0.0044 - mae: 0.0523\n",
      "Epoch 80/100\n",
      "1285/1285 [==============================] - 0s 192us/sample - loss: 0.0036 - mae: 0.0464\n",
      "Epoch 81/100\n",
      "1285/1285 [==============================] - 0s 197us/sample - loss: 0.0033 - mae: 0.0446\n",
      "Epoch 82/100\n",
      "1285/1285 [==============================] - 0s 195us/sample - loss: 0.0030 - mae: 0.0422\n",
      "Epoch 83/100\n",
      "1285/1285 [==============================] - 0s 195us/sample - loss: 0.0025 - mae: 0.0386\n",
      "Epoch 84/100\n",
      "1285/1285 [==============================] - 0s 195us/sample - loss: 0.0031 - mae: 0.0427\n",
      "Epoch 85/100\n",
      "1285/1285 [==============================] - 0s 198us/sample - loss: 0.0023 - mae: 0.0361\n",
      "Epoch 86/100\n",
      "1285/1285 [==============================] - 0s 198us/sample - loss: 0.0017 - mae: 0.0324\n",
      "Epoch 87/100\n",
      "1285/1285 [==============================] - 0s 199us/sample - loss: 0.0018 - mae: 0.0321\n",
      "Epoch 88/100\n",
      "1285/1285 [==============================] - 0s 197us/sample - loss: 0.0021 - mae: 0.0353\n",
      "Epoch 89/100\n",
      "1285/1285 [==============================] - 0s 196us/sample - loss: 0.0017 - mae: 0.0298\n",
      "Epoch 90/100\n",
      "1285/1285 [==============================] - 0s 196us/sample - loss: 0.0013 - mae: 0.0271\n",
      "Epoch 91/100\n",
      "1285/1285 [==============================] - 0s 199us/sample - loss: 0.0012 - mae: 0.0263\n",
      "Epoch 92/100\n",
      "1285/1285 [==============================] - 0s 196us/sample - loss: 0.0013 - mae: 0.0269\n",
      "Epoch 93/100\n",
      "1285/1285 [==============================] - 0s 189us/sample - loss: 9.9329e-04 - mae: 0.0232\n",
      "Epoch 94/100\n",
      "1285/1285 [==============================] - 0s 198us/sample - loss: 9.5483e-04 - mae: 0.0232\n",
      "Epoch 95/100\n",
      "1285/1285 [==============================] - 0s 195us/sample - loss: 7.4812e-04 - mae: 0.0208\n",
      "Epoch 96/100\n",
      "1285/1285 [==============================] - 0s 196us/sample - loss: 8.9607e-04 - mae: 0.0232\n",
      "Epoch 97/100\n",
      "1285/1285 [==============================] - 0s 187us/sample - loss: 6.7313e-04 - mae: 0.0192\n",
      "Epoch 98/100\n",
      "1285/1285 [==============================] - 0s 194us/sample - loss: 5.7475e-04 - mae: 0.0180\n",
      "Epoch 99/100\n",
      "1285/1285 [==============================] - 0s 191us/sample - loss: 5.6903e-04 - mae: 0.0180\n",
      "Epoch 100/100\n",
      "1285/1285 [==============================] - 0s 191us/sample - loss: 5.6094e-04 - mae: 0.0178\n"
     ]
    }
   ],
   "source": [
    "model = tf.keras.models.Sequential([\n",
    "  tf.keras.layers.Conv1D(filters=32, \n",
    "                         kernel_size=5,\n",
    "                         strides=1,\n",
    "                         padding=\"causal\",\n",
    "                         activation='relu',\n",
    "                         input_shape=(X_train.shape[1], X_train.shape[2])),\n",
    "  tf.keras.layers.LSTM(50,return_sequences=True),\n",
    "  tf.keras.layers.LSTM(50),\n",
    "  tf.keras.layers.Dense(128,activation='relu'),\n",
    "  tf.keras.layers.Dropout(0.1),\n",
    "  tf.keras.layers.Dense(1),\n",
    "  tf.keras.layers.Lambda(lambda x: x * 400)\n",
    "])\n",
    "\n",
    "model.summary()\n",
    "\n",
    "lr_schedule = tf.keras.callbacks.LearningRateScheduler(\n",
    "    lambda epoch: 1e-8 * 10**(epoch / 20))\n",
    "optimizer = tf.keras.optimizers.Adam(lr=1e-8)\n",
    "model.compile(loss='mse',\n",
    "              optimizer=optimizer,\n",
    "              metrics=[\"mae\"])\n",
    "\n",
    "history = model.fit(X_train, Y_train, epochs=100,callbacks=[lr_schedule])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1e-08, 0.1, 0.0001, 0.002)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.semilogx(history.history[\"lr\"], history.history[\"loss\"])\n",
    "plt.axis([1e-8, 1e-1, 0.0001, 0.002])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1285 samples, validate on 347 samples\n",
      "Epoch 1/300\n",
      "1285/1285 - 5s - loss: 0.1865 - mae: 0.3399 - val_loss: 0.0158 - val_mae: 0.1157\n",
      "Epoch 2/300\n",
      "1285/1285 - 0s - loss: 0.1194 - mae: 0.2714 - val_loss: 0.0071 - val_mae: 0.0682\n",
      "Epoch 3/300\n",
      "1285/1285 - 0s - loss: 0.1126 - mae: 0.2570 - val_loss: 0.0066 - val_mae: 0.0627\n",
      "Epoch 4/300\n",
      "1285/1285 - 0s - loss: 0.0997 - mae: 0.2437 - val_loss: 0.0072 - val_mae: 0.0710\n",
      "Epoch 5/300\n",
      "1285/1285 - 0s - loss: 0.0853 - mae: 0.2290 - val_loss: 0.0053 - val_mae: 0.0513\n",
      "Epoch 6/300\n",
      "1285/1285 - 0s - loss: 0.0772 - mae: 0.2147 - val_loss: 0.0050 - val_mae: 0.0490\n",
      "Epoch 7/300\n",
      "1285/1285 - 0s - loss: 0.0616 - mae: 0.1935 - val_loss: 0.0050 - val_mae: 0.0502\n",
      "Epoch 8/300\n",
      "1285/1285 - 0s - loss: 0.0522 - mae: 0.1780 - val_loss: 0.0048 - val_mae: 0.0458\n",
      "Epoch 9/300\n",
      "1285/1285 - 0s - loss: 0.0445 - mae: 0.1659 - val_loss: 0.0038 - val_mae: 0.0423\n",
      "Epoch 10/300\n",
      "1285/1285 - 0s - loss: 0.0386 - mae: 0.1503 - val_loss: 0.0058 - val_mae: 0.0686\n",
      "Epoch 11/300\n",
      "1285/1285 - 0s - loss: 0.0330 - mae: 0.1435 - val_loss: 0.0039 - val_mae: 0.0510\n",
      "Epoch 12/300\n",
      "1285/1285 - 0s - loss: 0.0271 - mae: 0.1266 - val_loss: 0.0033 - val_mae: 0.0379\n",
      "Epoch 13/300\n",
      "1285/1285 - 0s - loss: 0.0228 - mae: 0.1180 - val_loss: 0.0034 - val_mae: 0.0449\n",
      "Epoch 14/300\n",
      "1285/1285 - 0s - loss: 0.0203 - mae: 0.1131 - val_loss: 0.0037 - val_mae: 0.0402\n",
      "Epoch 15/300\n",
      "1285/1285 - 0s - loss: 0.0179 - mae: 0.1074 - val_loss: 0.0024 - val_mae: 0.0328\n",
      "Epoch 16/300\n",
      "1285/1285 - 0s - loss: 0.0146 - mae: 0.0957 - val_loss: 0.0023 - val_mae: 0.0321\n",
      "Epoch 17/300\n",
      "1285/1285 - 0s - loss: 0.0120 - mae: 0.0857 - val_loss: 0.0019 - val_mae: 0.0299\n",
      "Epoch 18/300\n",
      "1285/1285 - 0s - loss: 0.0121 - mae: 0.0858 - val_loss: 0.0020 - val_mae: 0.0346\n",
      "Epoch 19/300\n",
      "1285/1285 - 0s - loss: 0.0108 - mae: 0.0826 - val_loss: 0.0014 - val_mae: 0.0261\n",
      "Epoch 20/300\n",
      "1285/1285 - 0s - loss: 0.0097 - mae: 0.0775 - val_loss: 0.0013 - val_mae: 0.0267\n",
      "Epoch 21/300\n",
      "1285/1285 - 0s - loss: 0.0087 - mae: 0.0723 - val_loss: 0.0015 - val_mae: 0.0291\n",
      "Epoch 22/300\n",
      "1285/1285 - 0s - loss: 0.0081 - mae: 0.0685 - val_loss: 0.0011 - val_mae: 0.0242\n",
      "Epoch 23/300\n",
      "1285/1285 - 0s - loss: 0.0071 - mae: 0.0647 - val_loss: 0.0013 - val_mae: 0.0287\n",
      "Epoch 24/300\n",
      "1285/1285 - 0s - loss: 0.0070 - mae: 0.0662 - val_loss: 0.0012 - val_mae: 0.0263\n",
      "Epoch 25/300\n",
      "1285/1285 - 0s - loss: 0.0057 - mae: 0.0594 - val_loss: 0.0011 - val_mae: 0.0250\n",
      "Epoch 26/300\n",
      "1285/1285 - 0s - loss: 0.0056 - mae: 0.0591 - val_loss: 0.0018 - val_mae: 0.0354\n",
      "Epoch 27/300\n",
      "1285/1285 - 0s - loss: 0.0057 - mae: 0.0586 - val_loss: 8.5812e-04 - val_mae: 0.0232\n",
      "Epoch 28/300\n",
      "1285/1285 - 0s - loss: 0.0048 - mae: 0.0551 - val_loss: 8.1754e-04 - val_mae: 0.0215\n",
      "Epoch 29/300\n",
      "1285/1285 - 0s - loss: 0.0051 - mae: 0.0559 - val_loss: 8.7615e-04 - val_mae: 0.0232\n",
      "Epoch 30/300\n",
      "1285/1285 - 0s - loss: 0.0043 - mae: 0.0516 - val_loss: 8.7221e-04 - val_mae: 0.0217\n",
      "Epoch 31/300\n",
      "1285/1285 - 0s - loss: 0.0044 - mae: 0.0505 - val_loss: 8.6109e-04 - val_mae: 0.0202\n",
      "Epoch 32/300\n",
      "1285/1285 - 0s - loss: 0.0041 - mae: 0.0496 - val_loss: 9.2392e-04 - val_mae: 0.0205\n",
      "Epoch 33/300\n",
      "1285/1285 - 0s - loss: 0.0042 - mae: 0.0497 - val_loss: 8.7905e-04 - val_mae: 0.0204\n",
      "Epoch 34/300\n",
      "1285/1285 - 0s - loss: 0.0034 - mae: 0.0454 - val_loss: 9.3385e-04 - val_mae: 0.0212\n",
      "Epoch 35/300\n",
      "1285/1285 - 0s - loss: 0.0034 - mae: 0.0457 - val_loss: 0.0011 - val_mae: 0.0225\n",
      "Epoch 36/300\n",
      "1285/1285 - 0s - loss: 0.0036 - mae: 0.0465 - val_loss: 0.0010 - val_mae: 0.0226\n",
      "Epoch 37/300\n",
      "1285/1285 - 0s - loss: 0.0034 - mae: 0.0457 - val_loss: 0.0012 - val_mae: 0.0253\n",
      "Epoch 38/300\n",
      "1285/1285 - 0s - loss: 0.0034 - mae: 0.0438 - val_loss: 0.0012 - val_mae: 0.0256\n",
      "Epoch 39/300\n",
      "1285/1285 - 0s - loss: 0.0030 - mae: 0.0417 - val_loss: 0.0011 - val_mae: 0.0235\n",
      "Epoch 40/300\n",
      "1285/1285 - 0s - loss: 0.0028 - mae: 0.0410 - val_loss: 9.9869e-04 - val_mae: 0.0226\n",
      "Epoch 41/300\n",
      "1285/1285 - 0s - loss: 0.0026 - mae: 0.0403 - val_loss: 0.0010 - val_mae: 0.0227\n",
      "Epoch 42/300\n",
      "1285/1285 - 0s - loss: 0.0025 - mae: 0.0396 - val_loss: 0.0012 - val_mae: 0.0246\n",
      "Epoch 43/300\n",
      "1285/1285 - 0s - loss: 0.0025 - mae: 0.0388 - val_loss: 0.0013 - val_mae: 0.0247\n",
      "Epoch 44/300\n",
      "1285/1285 - 0s - loss: 0.0027 - mae: 0.0401 - val_loss: 0.0014 - val_mae: 0.0280\n",
      "Epoch 45/300\n",
      "1285/1285 - 0s - loss: 0.0026 - mae: 0.0388 - val_loss: 0.0013 - val_mae: 0.0248\n",
      "Epoch 46/300\n",
      "1285/1285 - 0s - loss: 0.0023 - mae: 0.0370 - val_loss: 0.0016 - val_mae: 0.0262\n",
      "Epoch 47/300\n",
      "1285/1285 - 0s - loss: 0.0025 - mae: 0.0378 - val_loss: 0.0016 - val_mae: 0.0272\n",
      "Epoch 48/300\n",
      "1285/1285 - 0s - loss: 0.0021 - mae: 0.0354 - val_loss: 0.0017 - val_mae: 0.0291\n",
      "Epoch 49/300\n",
      "1285/1285 - 0s - loss: 0.0021 - mae: 0.0351 - val_loss: 0.0016 - val_mae: 0.0273\n",
      "Epoch 50/300\n",
      "1285/1285 - 0s - loss: 0.0020 - mae: 0.0348 - val_loss: 0.0017 - val_mae: 0.0280\n",
      "Epoch 51/300\n",
      "1285/1285 - 0s - loss: 0.0019 - mae: 0.0331 - val_loss: 0.0019 - val_mae: 0.0290\n",
      "Epoch 52/300\n",
      "1285/1285 - 0s - loss: 0.0018 - mae: 0.0333 - val_loss: 0.0018 - val_mae: 0.0282\n",
      "Epoch 53/300\n",
      "1285/1285 - 0s - loss: 0.0018 - mae: 0.0325 - val_loss: 0.0019 - val_mae: 0.0297\n",
      "Epoch 54/300\n",
      "1285/1285 - 0s - loss: 0.0019 - mae: 0.0338 - val_loss: 0.0019 - val_mae: 0.0293\n",
      "Epoch 55/300\n",
      "1285/1285 - 0s - loss: 0.0019 - mae: 0.0333 - val_loss: 0.0020 - val_mae: 0.0298\n",
      "Epoch 56/300\n",
      "1285/1285 - 0s - loss: 0.0017 - mae: 0.0315 - val_loss: 0.0021 - val_mae: 0.0304\n",
      "Epoch 57/300\n",
      "1285/1285 - 0s - loss: 0.0018 - mae: 0.0319 - val_loss: 0.0021 - val_mae: 0.0309\n",
      "Epoch 58/300\n",
      "1285/1285 - 0s - loss: 0.0018 - mae: 0.0311 - val_loss: 0.0024 - val_mae: 0.0325\n",
      "Epoch 59/300\n",
      "1285/1285 - 0s - loss: 0.0017 - mae: 0.0314 - val_loss: 0.0030 - val_mae: 0.0344\n",
      "Epoch 60/300\n",
      "1285/1285 - 0s - loss: 0.0015 - mae: 0.0291 - val_loss: 0.0025 - val_mae: 0.0331\n",
      "Epoch 61/300\n",
      "1285/1285 - 0s - loss: 0.0014 - mae: 0.0285 - val_loss: 0.0028 - val_mae: 0.0338\n",
      "Epoch 62/300\n",
      "1285/1285 - 0s - loss: 0.0014 - mae: 0.0277 - val_loss: 0.0029 - val_mae: 0.0341\n",
      "Epoch 63/300\n",
      "1285/1285 - 0s - loss: 0.0012 - mae: 0.0267 - val_loss: 0.0030 - val_mae: 0.0344\n",
      "Epoch 64/300\n",
      "1285/1285 - 0s - loss: 0.0013 - mae: 0.0259 - val_loss: 0.0024 - val_mae: 0.0317\n",
      "Epoch 65/300\n",
      "1285/1285 - 0s - loss: 0.0012 - mae: 0.0266 - val_loss: 0.0024 - val_mae: 0.0321\n",
      "Epoch 66/300\n",
      "1285/1285 - 0s - loss: 0.0013 - mae: 0.0261 - val_loss: 0.0028 - val_mae: 0.0339\n",
      "Epoch 67/300\n",
      "1285/1285 - 0s - loss: 0.0013 - mae: 0.0260 - val_loss: 0.0031 - val_mae: 0.0343\n",
      "Epoch 68/300\n",
      "1285/1285 - 0s - loss: 0.0012 - mae: 0.0250 - val_loss: 0.0030 - val_mae: 0.0339\n",
      "Epoch 69/300\n",
      "1285/1285 - 0s - loss: 0.0011 - mae: 0.0244 - val_loss: 0.0028 - val_mae: 0.0341\n",
      "Epoch 70/300\n",
      "1285/1285 - 0s - loss: 0.0011 - mae: 0.0242 - val_loss: 0.0035 - val_mae: 0.0355\n",
      "Epoch 71/300\n",
      "1285/1285 - 0s - loss: 9.9652e-04 - mae: 0.0238 - val_loss: 0.0033 - val_mae: 0.0352\n",
      "Epoch 72/300\n",
      "1285/1285 - 0s - loss: 9.0138e-04 - mae: 0.0228 - val_loss: 0.0035 - val_mae: 0.0354\n",
      "Epoch 73/300\n",
      "1285/1285 - 0s - loss: 0.0013 - mae: 0.0243 - val_loss: 0.0036 - val_mae: 0.0356\n",
      "Epoch 74/300\n",
      "1285/1285 - 0s - loss: 0.0012 - mae: 0.0235 - val_loss: 0.0031 - val_mae: 0.0357\n",
      "Epoch 75/300\n",
      "1285/1285 - 0s - loss: 0.0010 - mae: 0.0232 - val_loss: 0.0034 - val_mae: 0.0357\n",
      "Epoch 76/300\n",
      "1285/1285 - 0s - loss: 9.4086e-04 - mae: 0.0222 - val_loss: 0.0032 - val_mae: 0.0339\n",
      "Epoch 77/300\n",
      "1285/1285 - 0s - loss: 8.3705e-04 - mae: 0.0215 - val_loss: 0.0036 - val_mae: 0.0351\n",
      "Epoch 78/300\n",
      "1285/1285 - 0s - loss: 8.2421e-04 - mae: 0.0215 - val_loss: 0.0034 - val_mae: 0.0347\n",
      "Epoch 79/300\n",
      "1285/1285 - 0s - loss: 8.0762e-04 - mae: 0.0212 - val_loss: 0.0038 - val_mae: 0.0350\n",
      "Epoch 80/300\n",
      "1285/1285 - 0s - loss: 8.3804e-04 - mae: 0.0211 - val_loss: 0.0031 - val_mae: 0.0332\n",
      "Epoch 81/300\n",
      "1285/1285 - 0s - loss: 9.3035e-04 - mae: 0.0207 - val_loss: 0.0037 - val_mae: 0.0358\n",
      "Epoch 82/300\n",
      "1285/1285 - 0s - loss: 7.0748e-04 - mae: 0.0201 - val_loss: 0.0039 - val_mae: 0.0351\n",
      "Epoch 83/300\n",
      "1285/1285 - 0s - loss: 7.6047e-04 - mae: 0.0201 - val_loss: 0.0040 - val_mae: 0.0356\n",
      "Epoch 84/300\n",
      "1285/1285 - 0s - loss: 8.4885e-04 - mae: 0.0201 - val_loss: 0.0033 - val_mae: 0.0341\n",
      "Epoch 85/300\n",
      "1285/1285 - 0s - loss: 6.5497e-04 - mae: 0.0193 - val_loss: 0.0035 - val_mae: 0.0352\n",
      "Epoch 86/300\n",
      "1285/1285 - 0s - loss: 8.3356e-04 - mae: 0.0200 - val_loss: 0.0038 - val_mae: 0.0351\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 87/300\n",
      "1285/1285 - 0s - loss: 8.7072e-04 - mae: 0.0202 - val_loss: 0.0038 - val_mae: 0.0356\n",
      "Epoch 88/300\n",
      "1285/1285 - 0s - loss: 7.1743e-04 - mae: 0.0196 - val_loss: 0.0035 - val_mae: 0.0346\n",
      "Epoch 89/300\n",
      "1285/1285 - 0s - loss: 6.4950e-04 - mae: 0.0189 - val_loss: 0.0039 - val_mae: 0.0352\n",
      "Epoch 90/300\n",
      "1285/1285 - 0s - loss: 7.1828e-04 - mae: 0.0199 - val_loss: 0.0043 - val_mae: 0.0363\n",
      "Epoch 91/300\n",
      "1285/1285 - 0s - loss: 8.2716e-04 - mae: 0.0202 - val_loss: 0.0038 - val_mae: 0.0352\n",
      "Epoch 92/300\n",
      "1285/1285 - 0s - loss: 7.0926e-04 - mae: 0.0191 - val_loss: 0.0041 - val_mae: 0.0359\n",
      "Epoch 93/300\n",
      "1285/1285 - 0s - loss: 5.9778e-04 - mae: 0.0185 - val_loss: 0.0041 - val_mae: 0.0366\n",
      "Epoch 94/300\n",
      "1285/1285 - 0s - loss: 6.4681e-04 - mae: 0.0187 - val_loss: 0.0038 - val_mae: 0.0352\n",
      "Epoch 95/300\n",
      "1285/1285 - 0s - loss: 6.4368e-04 - mae: 0.0187 - val_loss: 0.0034 - val_mae: 0.0341\n",
      "Epoch 96/300\n",
      "1285/1285 - 0s - loss: 7.3057e-04 - mae: 0.0193 - val_loss: 0.0043 - val_mae: 0.0370\n",
      "Epoch 97/300\n",
      "1285/1285 - 0s - loss: 6.4130e-04 - mae: 0.0182 - val_loss: 0.0039 - val_mae: 0.0361\n",
      "Epoch 98/300\n",
      "1285/1285 - 0s - loss: 7.7123e-04 - mae: 0.0194 - val_loss: 0.0041 - val_mae: 0.0367\n",
      "Epoch 99/300\n",
      "1285/1285 - 0s - loss: 6.6332e-04 - mae: 0.0186 - val_loss: 0.0038 - val_mae: 0.0353\n",
      "Epoch 100/300\n",
      "1285/1285 - 0s - loss: 6.5168e-04 - mae: 0.0183 - val_loss: 0.0035 - val_mae: 0.0346\n",
      "Epoch 101/300\n",
      "1285/1285 - 0s - loss: 6.3184e-04 - mae: 0.0184 - val_loss: 0.0039 - val_mae: 0.0360\n",
      "Epoch 102/300\n",
      "1285/1285 - 0s - loss: 6.3177e-04 - mae: 0.0182 - val_loss: 0.0041 - val_mae: 0.0369\n",
      "Epoch 103/300\n",
      "1285/1285 - 0s - loss: 5.9177e-04 - mae: 0.0176 - val_loss: 0.0038 - val_mae: 0.0349\n",
      "Epoch 104/300\n",
      "1285/1285 - 0s - loss: 5.7251e-04 - mae: 0.0179 - val_loss: 0.0038 - val_mae: 0.0358\n",
      "Epoch 105/300\n",
      "1285/1285 - 0s - loss: 5.9861e-04 - mae: 0.0177 - val_loss: 0.0040 - val_mae: 0.0359\n",
      "Epoch 106/300\n",
      "1285/1285 - 0s - loss: 5.9592e-04 - mae: 0.0177 - val_loss: 0.0038 - val_mae: 0.0367\n",
      "Epoch 107/300\n",
      "1285/1285 - 0s - loss: 5.7414e-04 - mae: 0.0175 - val_loss: 0.0038 - val_mae: 0.0354\n",
      "Epoch 108/300\n",
      "1285/1285 - 0s - loss: 5.8243e-04 - mae: 0.0174 - val_loss: 0.0033 - val_mae: 0.0342\n",
      "Epoch 109/300\n",
      "1285/1285 - 0s - loss: 5.8170e-04 - mae: 0.0173 - val_loss: 0.0038 - val_mae: 0.0359\n",
      "Epoch 110/300\n",
      "1285/1285 - 0s - loss: 5.9710e-04 - mae: 0.0180 - val_loss: 0.0040 - val_mae: 0.0365\n",
      "Epoch 111/300\n",
      "1285/1285 - 0s - loss: 7.2968e-04 - mae: 0.0183 - val_loss: 0.0042 - val_mae: 0.0371\n",
      "Epoch 112/300\n",
      "1285/1285 - 0s - loss: 5.2594e-04 - mae: 0.0168 - val_loss: 0.0039 - val_mae: 0.0370\n",
      "Epoch 113/300\n",
      "1285/1285 - 0s - loss: 5.2669e-04 - mae: 0.0167 - val_loss: 0.0040 - val_mae: 0.0364\n",
      "Epoch 114/300\n",
      "1285/1285 - 0s - loss: 5.6992e-04 - mae: 0.0175 - val_loss: 0.0039 - val_mae: 0.0360\n",
      "Epoch 115/300\n",
      "1285/1285 - 0s - loss: 5.6834e-04 - mae: 0.0175 - val_loss: 0.0040 - val_mae: 0.0372\n",
      "Epoch 116/300\n",
      "1285/1285 - 0s - loss: 4.9807e-04 - mae: 0.0168 - val_loss: 0.0041 - val_mae: 0.0368\n",
      "Epoch 117/300\n",
      "1285/1285 - 0s - loss: 6.1524e-04 - mae: 0.0172 - val_loss: 0.0036 - val_mae: 0.0350\n",
      "Epoch 118/300\n",
      "1285/1285 - 0s - loss: 5.4965e-04 - mae: 0.0171 - val_loss: 0.0039 - val_mae: 0.0354\n",
      "Epoch 119/300\n",
      "1285/1285 - 0s - loss: 5.4675e-04 - mae: 0.0169 - val_loss: 0.0037 - val_mae: 0.0355\n",
      "Epoch 120/300\n",
      "1285/1285 - 0s - loss: 6.0314e-04 - mae: 0.0175 - val_loss: 0.0040 - val_mae: 0.0363\n",
      "Epoch 121/300\n",
      "1285/1285 - 0s - loss: 5.6768e-04 - mae: 0.0175 - val_loss: 0.0037 - val_mae: 0.0362\n",
      "Epoch 122/300\n",
      "1285/1285 - 0s - loss: 6.1520e-04 - mae: 0.0170 - val_loss: 0.0038 - val_mae: 0.0357\n",
      "Epoch 123/300\n",
      "1285/1285 - 0s - loss: 5.7741e-04 - mae: 0.0175 - val_loss: 0.0038 - val_mae: 0.0365\n",
      "Epoch 124/300\n",
      "1285/1285 - 0s - loss: 5.5544e-04 - mae: 0.0176 - val_loss: 0.0037 - val_mae: 0.0376\n",
      "Epoch 125/300\n",
      "1285/1285 - 0s - loss: 5.9138e-04 - mae: 0.0175 - val_loss: 0.0038 - val_mae: 0.0360\n",
      "Epoch 126/300\n",
      "1285/1285 - 0s - loss: 5.9819e-04 - mae: 0.0173 - val_loss: 0.0038 - val_mae: 0.0361\n",
      "Epoch 127/300\n",
      "1285/1285 - 0s - loss: 5.7343e-04 - mae: 0.0171 - val_loss: 0.0038 - val_mae: 0.0355\n",
      "Epoch 128/300\n",
      "1285/1285 - 0s - loss: 5.4385e-04 - mae: 0.0168 - val_loss: 0.0038 - val_mae: 0.0362\n",
      "Epoch 129/300\n",
      "1285/1285 - 0s - loss: 5.1509e-04 - mae: 0.0168 - val_loss: 0.0039 - val_mae: 0.0365\n",
      "Epoch 130/300\n",
      "1285/1285 - 0s - loss: 5.7620e-04 - mae: 0.0169 - val_loss: 0.0037 - val_mae: 0.0366\n",
      "Epoch 131/300\n",
      "1285/1285 - 0s - loss: 4.7919e-04 - mae: 0.0164 - val_loss: 0.0037 - val_mae: 0.0354\n",
      "Epoch 132/300\n",
      "1285/1285 - 0s - loss: 5.7699e-04 - mae: 0.0172 - val_loss: 0.0036 - val_mae: 0.0364\n",
      "Epoch 133/300\n",
      "1285/1285 - 0s - loss: 5.1976e-04 - mae: 0.0169 - val_loss: 0.0036 - val_mae: 0.0348\n",
      "Epoch 134/300\n",
      "1285/1285 - 0s - loss: 5.1767e-04 - mae: 0.0167 - val_loss: 0.0034 - val_mae: 0.0341\n",
      "Epoch 135/300\n",
      "1285/1285 - 0s - loss: 5.6063e-04 - mae: 0.0172 - val_loss: 0.0035 - val_mae: 0.0369\n",
      "Epoch 136/300\n",
      "1285/1285 - 0s - loss: 5.3884e-04 - mae: 0.0169 - val_loss: 0.0034 - val_mae: 0.0343\n",
      "Epoch 137/300\n",
      "1285/1285 - 0s - loss: 5.2522e-04 - mae: 0.0169 - val_loss: 0.0038 - val_mae: 0.0355\n",
      "Epoch 138/300\n",
      "1285/1285 - 0s - loss: 5.3149e-04 - mae: 0.0167 - val_loss: 0.0035 - val_mae: 0.0350\n",
      "Epoch 139/300\n",
      "1285/1285 - 0s - loss: 5.0437e-04 - mae: 0.0169 - val_loss: 0.0037 - val_mae: 0.0364\n",
      "Epoch 140/300\n",
      "1285/1285 - 0s - loss: 5.2511e-04 - mae: 0.0167 - val_loss: 0.0035 - val_mae: 0.0353\n",
      "Epoch 141/300\n",
      "1285/1285 - 0s - loss: 5.4357e-04 - mae: 0.0165 - val_loss: 0.0037 - val_mae: 0.0357\n",
      "Epoch 142/300\n",
      "1285/1285 - 0s - loss: 5.2741e-04 - mae: 0.0168 - val_loss: 0.0037 - val_mae: 0.0363\n",
      "Epoch 143/300\n",
      "1285/1285 - 0s - loss: 5.1767e-04 - mae: 0.0165 - val_loss: 0.0035 - val_mae: 0.0348\n",
      "Epoch 144/300\n",
      "1285/1285 - 0s - loss: 5.6723e-04 - mae: 0.0175 - val_loss: 0.0038 - val_mae: 0.0366\n",
      "Epoch 145/300\n",
      "1285/1285 - 0s - loss: 5.6331e-04 - mae: 0.0168 - val_loss: 0.0030 - val_mae: 0.0334\n",
      "Epoch 146/300\n",
      "1285/1285 - 0s - loss: 5.5476e-04 - mae: 0.0169 - val_loss: 0.0035 - val_mae: 0.0346\n",
      "Epoch 147/300\n",
      "1285/1285 - 0s - loss: 5.1576e-04 - mae: 0.0165 - val_loss: 0.0032 - val_mae: 0.0349\n",
      "Epoch 148/300\n",
      "1285/1285 - 0s - loss: 4.9406e-04 - mae: 0.0165 - val_loss: 0.0034 - val_mae: 0.0361\n",
      "Epoch 149/300\n",
      "1285/1285 - 0s - loss: 5.6514e-04 - mae: 0.0171 - val_loss: 0.0034 - val_mae: 0.0349\n",
      "Epoch 150/300\n",
      "1285/1285 - 0s - loss: 5.3950e-04 - mae: 0.0169 - val_loss: 0.0035 - val_mae: 0.0345\n",
      "Epoch 151/300\n",
      "1285/1285 - 0s - loss: 5.9746e-04 - mae: 0.0171 - val_loss: 0.0035 - val_mae: 0.0350\n",
      "Epoch 152/300\n",
      "1285/1285 - 0s - loss: 5.5117e-04 - mae: 0.0172 - val_loss: 0.0035 - val_mae: 0.0348\n",
      "Epoch 153/300\n",
      "1285/1285 - 0s - loss: 4.8364e-04 - mae: 0.0161 - val_loss: 0.0031 - val_mae: 0.0322\n",
      "Epoch 154/300\n",
      "1285/1285 - 0s - loss: 5.5925e-04 - mae: 0.0173 - val_loss: 0.0032 - val_mae: 0.0332\n",
      "Epoch 155/300\n",
      "1285/1285 - 0s - loss: 4.8090e-04 - mae: 0.0166 - val_loss: 0.0030 - val_mae: 0.0326\n",
      "Epoch 156/300\n",
      "1285/1285 - 0s - loss: 4.7933e-04 - mae: 0.0162 - val_loss: 0.0032 - val_mae: 0.0339\n",
      "Epoch 157/300\n",
      "1285/1285 - 0s - loss: 4.7083e-04 - mae: 0.0159 - val_loss: 0.0031 - val_mae: 0.0329\n",
      "Epoch 158/300\n",
      "1285/1285 - 0s - loss: 4.9693e-04 - mae: 0.0167 - val_loss: 0.0032 - val_mae: 0.0333\n",
      "Epoch 159/300\n",
      "1285/1285 - 0s - loss: 5.1486e-04 - mae: 0.0167 - val_loss: 0.0034 - val_mae: 0.0353\n",
      "Epoch 160/300\n",
      "1285/1285 - 0s - loss: 5.3261e-04 - mae: 0.0166 - val_loss: 0.0032 - val_mae: 0.0335\n",
      "Epoch 161/300\n",
      "1285/1285 - 0s - loss: 5.1896e-04 - mae: 0.0168 - val_loss: 0.0032 - val_mae: 0.0340\n",
      "Epoch 162/300\n",
      "1285/1285 - 0s - loss: 4.5801e-04 - mae: 0.0155 - val_loss: 0.0034 - val_mae: 0.0336\n",
      "Epoch 163/300\n",
      "1285/1285 - 0s - loss: 5.2457e-04 - mae: 0.0172 - val_loss: 0.0031 - val_mae: 0.0345\n",
      "Epoch 164/300\n",
      "1285/1285 - 0s - loss: 5.1697e-04 - mae: 0.0164 - val_loss: 0.0031 - val_mae: 0.0342\n",
      "Epoch 165/300\n",
      "1285/1285 - 0s - loss: 5.1322e-04 - mae: 0.0160 - val_loss: 0.0030 - val_mae: 0.0328\n",
      "Epoch 166/300\n",
      "1285/1285 - 0s - loss: 5.5458e-04 - mae: 0.0166 - val_loss: 0.0031 - val_mae: 0.0339\n",
      "Epoch 167/300\n",
      "1285/1285 - 0s - loss: 5.0500e-04 - mae: 0.0167 - val_loss: 0.0032 - val_mae: 0.0350\n",
      "Epoch 168/300\n",
      "1285/1285 - 0s - loss: 5.1045e-04 - mae: 0.0158 - val_loss: 0.0030 - val_mae: 0.0330\n",
      "Epoch 169/300\n",
      "1285/1285 - 0s - loss: 4.6869e-04 - mae: 0.0159 - val_loss: 0.0033 - val_mae: 0.0342\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 170/300\n",
      "1285/1285 - 0s - loss: 4.9814e-04 - mae: 0.0166 - val_loss: 0.0031 - val_mae: 0.0338\n",
      "Epoch 171/300\n",
      "1285/1285 - 0s - loss: 5.6429e-04 - mae: 0.0178 - val_loss: 0.0033 - val_mae: 0.0349\n",
      "Epoch 172/300\n",
      "1285/1285 - 0s - loss: 5.0890e-04 - mae: 0.0168 - val_loss: 0.0033 - val_mae: 0.0343\n",
      "Epoch 173/300\n",
      "1285/1285 - 0s - loss: 5.4747e-04 - mae: 0.0167 - val_loss: 0.0032 - val_mae: 0.0353\n",
      "Epoch 174/300\n",
      "1285/1285 - 0s - loss: 5.1520e-04 - mae: 0.0161 - val_loss: 0.0035 - val_mae: 0.0347\n",
      "Epoch 175/300\n",
      "1285/1285 - 0s - loss: 5.4360e-04 - mae: 0.0174 - val_loss: 0.0032 - val_mae: 0.0354\n",
      "Epoch 176/300\n",
      "1285/1285 - 0s - loss: 4.3318e-04 - mae: 0.0156 - val_loss: 0.0031 - val_mae: 0.0347\n",
      "Epoch 177/300\n",
      "1285/1285 - 0s - loss: 4.5590e-04 - mae: 0.0154 - val_loss: 0.0031 - val_mae: 0.0326\n",
      "Epoch 178/300\n",
      "1285/1285 - 0s - loss: 5.1317e-04 - mae: 0.0169 - val_loss: 0.0032 - val_mae: 0.0340\n",
      "Epoch 179/300\n",
      "1285/1285 - 0s - loss: 5.0136e-04 - mae: 0.0168 - val_loss: 0.0032 - val_mae: 0.0340\n",
      "Epoch 180/300\n",
      "1285/1285 - 0s - loss: 5.0383e-04 - mae: 0.0159 - val_loss: 0.0031 - val_mae: 0.0349\n",
      "Epoch 181/300\n",
      "1285/1285 - 0s - loss: 4.3568e-04 - mae: 0.0157 - val_loss: 0.0030 - val_mae: 0.0327\n",
      "Epoch 182/300\n",
      "1285/1285 - 0s - loss: 5.3123e-04 - mae: 0.0173 - val_loss: 0.0032 - val_mae: 0.0353\n",
      "Epoch 183/300\n",
      "1285/1285 - 0s - loss: 4.7160e-04 - mae: 0.0156 - val_loss: 0.0030 - val_mae: 0.0324\n",
      "Epoch 184/300\n",
      "1285/1285 - 0s - loss: 4.7101e-04 - mae: 0.0160 - val_loss: 0.0030 - val_mae: 0.0327\n",
      "Epoch 185/300\n",
      "1285/1285 - 0s - loss: 4.5402e-04 - mae: 0.0159 - val_loss: 0.0030 - val_mae: 0.0325\n",
      "Epoch 186/300\n",
      "1285/1285 - 0s - loss: 4.9646e-04 - mae: 0.0163 - val_loss: 0.0031 - val_mae: 0.0338\n",
      "Epoch 187/300\n",
      "1285/1285 - 0s - loss: 4.2217e-04 - mae: 0.0149 - val_loss: 0.0032 - val_mae: 0.0331\n",
      "Epoch 188/300\n",
      "1285/1285 - 0s - loss: 4.7470e-04 - mae: 0.0156 - val_loss: 0.0030 - val_mae: 0.0322\n",
      "Epoch 189/300\n",
      "1285/1285 - 0s - loss: 4.8707e-04 - mae: 0.0170 - val_loss: 0.0029 - val_mae: 0.0326\n",
      "Epoch 190/300\n",
      "1285/1285 - 0s - loss: 5.2052e-04 - mae: 0.0168 - val_loss: 0.0027 - val_mae: 0.0332\n",
      "Epoch 191/300\n",
      "1285/1285 - 0s - loss: 5.3752e-04 - mae: 0.0164 - val_loss: 0.0030 - val_mae: 0.0331\n",
      "Epoch 192/300\n",
      "1285/1285 - 0s - loss: 4.9449e-04 - mae: 0.0158 - val_loss: 0.0028 - val_mae: 0.0320\n",
      "Epoch 193/300\n",
      "1285/1285 - 0s - loss: 4.6346e-04 - mae: 0.0160 - val_loss: 0.0030 - val_mae: 0.0332\n",
      "Epoch 194/300\n",
      "1285/1285 - 0s - loss: 4.5272e-04 - mae: 0.0157 - val_loss: 0.0029 - val_mae: 0.0322\n",
      "Epoch 195/300\n",
      "1285/1285 - 0s - loss: 4.8488e-04 - mae: 0.0166 - val_loss: 0.0032 - val_mae: 0.0339\n",
      "Epoch 196/300\n",
      "1285/1285 - 0s - loss: 4.5009e-04 - mae: 0.0155 - val_loss: 0.0030 - val_mae: 0.0330\n",
      "Epoch 197/300\n",
      "1285/1285 - 0s - loss: 5.0532e-04 - mae: 0.0163 - val_loss: 0.0031 - val_mae: 0.0335\n",
      "Epoch 198/300\n",
      "1285/1285 - 0s - loss: 4.4901e-04 - mae: 0.0155 - val_loss: 0.0029 - val_mae: 0.0324\n",
      "Epoch 199/300\n",
      "1285/1285 - 0s - loss: 5.0359e-04 - mae: 0.0164 - val_loss: 0.0030 - val_mae: 0.0329\n",
      "Epoch 200/300\n",
      "1285/1285 - 0s - loss: 4.3025e-04 - mae: 0.0155 - val_loss: 0.0029 - val_mae: 0.0321\n",
      "Epoch 201/300\n",
      "1285/1285 - 0s - loss: 4.9793e-04 - mae: 0.0169 - val_loss: 0.0030 - val_mae: 0.0336\n",
      "Epoch 202/300\n",
      "1285/1285 - 0s - loss: 4.7404e-04 - mae: 0.0157 - val_loss: 0.0029 - val_mae: 0.0328\n",
      "Epoch 203/300\n",
      "1285/1285 - 0s - loss: 4.8069e-04 - mae: 0.0161 - val_loss: 0.0030 - val_mae: 0.0329\n",
      "Epoch 204/300\n",
      "1285/1285 - 0s - loss: 4.7224e-04 - mae: 0.0163 - val_loss: 0.0031 - val_mae: 0.0339\n",
      "Epoch 205/300\n",
      "1285/1285 - 0s - loss: 4.6144e-04 - mae: 0.0161 - val_loss: 0.0030 - val_mae: 0.0329\n",
      "Epoch 206/300\n",
      "1285/1285 - 0s - loss: 5.1099e-04 - mae: 0.0158 - val_loss: 0.0031 - val_mae: 0.0335\n",
      "Epoch 207/300\n",
      "1285/1285 - 0s - loss: 5.6600e-04 - mae: 0.0180 - val_loss: 0.0031 - val_mae: 0.0344\n",
      "Epoch 208/300\n",
      "1285/1285 - 0s - loss: 4.6262e-04 - mae: 0.0155 - val_loss: 0.0029 - val_mae: 0.0321\n",
      "Epoch 209/300\n",
      "1285/1285 - 0s - loss: 4.4472e-04 - mae: 0.0159 - val_loss: 0.0029 - val_mae: 0.0336\n",
      "Epoch 210/300\n",
      "1285/1285 - 0s - loss: 5.0095e-04 - mae: 0.0158 - val_loss: 0.0030 - val_mae: 0.0323\n",
      "Epoch 211/300\n",
      "1285/1285 - 0s - loss: 5.0255e-04 - mae: 0.0171 - val_loss: 0.0029 - val_mae: 0.0319\n",
      "Epoch 212/300\n",
      "1285/1285 - 0s - loss: 4.8863e-04 - mae: 0.0168 - val_loss: 0.0030 - val_mae: 0.0334\n",
      "Epoch 213/300\n",
      "1285/1285 - 0s - loss: 4.5614e-04 - mae: 0.0154 - val_loss: 0.0029 - val_mae: 0.0334\n",
      "Epoch 214/300\n",
      "1285/1285 - 0s - loss: 4.4072e-04 - mae: 0.0157 - val_loss: 0.0029 - val_mae: 0.0317\n",
      "Epoch 215/300\n",
      "1285/1285 - 0s - loss: 4.5898e-04 - mae: 0.0162 - val_loss: 0.0028 - val_mae: 0.0316\n",
      "Epoch 216/300\n",
      "1285/1285 - 0s - loss: 4.3063e-04 - mae: 0.0154 - val_loss: 0.0028 - val_mae: 0.0317\n",
      "Epoch 217/300\n",
      "1285/1285 - 0s - loss: 4.2177e-04 - mae: 0.0151 - val_loss: 0.0028 - val_mae: 0.0313\n",
      "Epoch 218/300\n",
      "1285/1285 - 0s - loss: 4.6356e-04 - mae: 0.0157 - val_loss: 0.0027 - val_mae: 0.0308\n",
      "Epoch 219/300\n",
      "1285/1285 - 0s - loss: 4.4639e-04 - mae: 0.0159 - val_loss: 0.0028 - val_mae: 0.0312\n",
      "Epoch 220/300\n",
      "1285/1285 - 0s - loss: 4.9244e-04 - mae: 0.0163 - val_loss: 0.0027 - val_mae: 0.0310\n",
      "Epoch 221/300\n",
      "1285/1285 - 0s - loss: 4.8063e-04 - mae: 0.0160 - val_loss: 0.0027 - val_mae: 0.0314\n",
      "Epoch 222/300\n",
      "1285/1285 - 0s - loss: 4.8377e-04 - mae: 0.0161 - val_loss: 0.0028 - val_mae: 0.0317\n",
      "Epoch 223/300\n",
      "1285/1285 - 0s - loss: 4.5101e-04 - mae: 0.0155 - val_loss: 0.0028 - val_mae: 0.0311\n",
      "Epoch 224/300\n",
      "1285/1285 - 0s - loss: 4.3940e-04 - mae: 0.0159 - val_loss: 0.0027 - val_mae: 0.0314\n",
      "Epoch 225/300\n",
      "1285/1285 - 0s - loss: 4.5386e-04 - mae: 0.0159 - val_loss: 0.0028 - val_mae: 0.0320\n",
      "Epoch 226/300\n",
      "1285/1285 - 0s - loss: 4.1782e-04 - mae: 0.0153 - val_loss: 0.0027 - val_mae: 0.0309\n",
      "Epoch 227/300\n",
      "1285/1285 - 0s - loss: 4.8049e-04 - mae: 0.0159 - val_loss: 0.0027 - val_mae: 0.0315\n",
      "Epoch 228/300\n",
      "1285/1285 - 0s - loss: 4.4078e-04 - mae: 0.0155 - val_loss: 0.0025 - val_mae: 0.0303\n",
      "Epoch 229/300\n",
      "1285/1285 - 0s - loss: 4.7510e-04 - mae: 0.0161 - val_loss: 0.0028 - val_mae: 0.0317\n",
      "Epoch 230/300\n",
      "1285/1285 - 0s - loss: 4.5235e-04 - mae: 0.0157 - val_loss: 0.0028 - val_mae: 0.0320\n",
      "Epoch 231/300\n",
      "1285/1285 - 0s - loss: 4.6278e-04 - mae: 0.0157 - val_loss: 0.0028 - val_mae: 0.0319\n",
      "Epoch 232/300\n",
      "1285/1285 - 0s - loss: 4.5666e-04 - mae: 0.0158 - val_loss: 0.0030 - val_mae: 0.0318\n",
      "Epoch 233/300\n",
      "1285/1285 - 0s - loss: 4.9063e-04 - mae: 0.0161 - val_loss: 0.0028 - val_mae: 0.0319\n",
      "Epoch 234/300\n",
      "1285/1285 - 0s - loss: 4.2947e-04 - mae: 0.0155 - val_loss: 0.0027 - val_mae: 0.0315\n",
      "Epoch 235/300\n",
      "1285/1285 - 0s - loss: 4.7367e-04 - mae: 0.0155 - val_loss: 0.0026 - val_mae: 0.0312\n",
      "Epoch 236/300\n",
      "1285/1285 - 0s - loss: 4.4246e-04 - mae: 0.0156 - val_loss: 0.0028 - val_mae: 0.0313\n",
      "Epoch 237/300\n",
      "1285/1285 - 0s - loss: 4.9135e-04 - mae: 0.0165 - val_loss: 0.0028 - val_mae: 0.0318\n",
      "Epoch 238/300\n",
      "1285/1285 - 0s - loss: 4.6208e-04 - mae: 0.0163 - val_loss: 0.0026 - val_mae: 0.0315\n",
      "Epoch 239/300\n",
      "1285/1285 - 0s - loss: 4.2100e-04 - mae: 0.0152 - val_loss: 0.0026 - val_mae: 0.0313\n",
      "Epoch 240/300\n",
      "1285/1285 - 0s - loss: 4.0583e-04 - mae: 0.0148 - val_loss: 0.0027 - val_mae: 0.0312\n",
      "Epoch 241/300\n",
      "1285/1285 - 0s - loss: 4.5832e-04 - mae: 0.0161 - val_loss: 0.0028 - val_mae: 0.0316\n",
      "Epoch 242/300\n",
      "1285/1285 - 0s - loss: 4.6430e-04 - mae: 0.0159 - val_loss: 0.0027 - val_mae: 0.0312\n",
      "Epoch 243/300\n",
      "1285/1285 - 0s - loss: 4.4669e-04 - mae: 0.0154 - val_loss: 0.0026 - val_mae: 0.0305\n",
      "Epoch 244/300\n",
      "1285/1285 - 0s - loss: 5.6788e-04 - mae: 0.0165 - val_loss: 0.0030 - val_mae: 0.0327\n",
      "Epoch 245/300\n",
      "1285/1285 - 0s - loss: 4.7892e-04 - mae: 0.0163 - val_loss: 0.0027 - val_mae: 0.0314\n",
      "Epoch 246/300\n",
      "1285/1285 - 0s - loss: 4.5066e-04 - mae: 0.0156 - val_loss: 0.0027 - val_mae: 0.0311\n",
      "Epoch 247/300\n",
      "1285/1285 - 0s - loss: 4.4519e-04 - mae: 0.0153 - val_loss: 0.0025 - val_mae: 0.0311\n",
      "Epoch 248/300\n",
      "1285/1285 - 0s - loss: 4.5036e-04 - mae: 0.0160 - val_loss: 0.0025 - val_mae: 0.0317\n",
      "Epoch 249/300\n",
      "1285/1285 - 0s - loss: 4.7587e-04 - mae: 0.0159 - val_loss: 0.0028 - val_mae: 0.0324\n",
      "Epoch 250/300\n",
      "1285/1285 - 0s - loss: 4.5955e-04 - mae: 0.0158 - val_loss: 0.0027 - val_mae: 0.0320\n",
      "Epoch 251/300\n",
      "1285/1285 - 0s - loss: 4.1573e-04 - mae: 0.0153 - val_loss: 0.0027 - val_mae: 0.0316\n",
      "Epoch 252/300\n",
      "1285/1285 - 0s - loss: 4.4980e-04 - mae: 0.0158 - val_loss: 0.0028 - val_mae: 0.0325\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 253/300\n",
      "1285/1285 - 0s - loss: 4.6950e-04 - mae: 0.0156 - val_loss: 0.0027 - val_mae: 0.0305\n",
      "Epoch 254/300\n",
      "1285/1285 - 0s - loss: 4.0956e-04 - mae: 0.0154 - val_loss: 0.0027 - val_mae: 0.0308\n",
      "Epoch 255/300\n",
      "1285/1285 - 0s - loss: 4.2567e-04 - mae: 0.0153 - val_loss: 0.0028 - val_mae: 0.0309\n",
      "Epoch 256/300\n",
      "1285/1285 - 0s - loss: 4.3506e-04 - mae: 0.0157 - val_loss: 0.0025 - val_mae: 0.0300\n",
      "Epoch 257/300\n",
      "1285/1285 - 0s - loss: 4.3442e-04 - mae: 0.0154 - val_loss: 0.0026 - val_mae: 0.0307\n",
      "Epoch 258/300\n",
      "1285/1285 - 0s - loss: 4.3498e-04 - mae: 0.0156 - val_loss: 0.0026 - val_mae: 0.0311\n",
      "Epoch 259/300\n",
      "1285/1285 - 0s - loss: 4.5637e-04 - mae: 0.0156 - val_loss: 0.0026 - val_mae: 0.0318\n",
      "Epoch 260/300\n",
      "1285/1285 - 0s - loss: 4.4970e-04 - mae: 0.0156 - val_loss: 0.0027 - val_mae: 0.0315\n",
      "Epoch 261/300\n",
      "1285/1285 - 0s - loss: 4.9366e-04 - mae: 0.0167 - val_loss: 0.0025 - val_mae: 0.0306\n",
      "Epoch 262/300\n",
      "1285/1285 - 0s - loss: 4.2229e-04 - mae: 0.0153 - val_loss: 0.0027 - val_mae: 0.0315\n",
      "Epoch 263/300\n",
      "1285/1285 - 0s - loss: 4.3033e-04 - mae: 0.0154 - val_loss: 0.0027 - val_mae: 0.0322\n",
      "Epoch 264/300\n",
      "1285/1285 - 0s - loss: 3.9789e-04 - mae: 0.0148 - val_loss: 0.0025 - val_mae: 0.0300\n",
      "Epoch 265/300\n",
      "1285/1285 - 0s - loss: 4.9023e-04 - mae: 0.0160 - val_loss: 0.0027 - val_mae: 0.0316\n",
      "Epoch 266/300\n",
      "1285/1285 - 0s - loss: 4.7124e-04 - mae: 0.0158 - val_loss: 0.0026 - val_mae: 0.0307\n",
      "Epoch 267/300\n",
      "1285/1285 - 0s - loss: 4.3595e-04 - mae: 0.0156 - val_loss: 0.0026 - val_mae: 0.0309\n",
      "Epoch 268/300\n",
      "1285/1285 - 0s - loss: 4.2335e-04 - mae: 0.0150 - val_loss: 0.0026 - val_mae: 0.0307\n",
      "Epoch 269/300\n",
      "1285/1285 - 0s - loss: 4.3031e-04 - mae: 0.0154 - val_loss: 0.0025 - val_mae: 0.0301\n",
      "Epoch 270/300\n",
      "1285/1285 - 0s - loss: 3.9977e-04 - mae: 0.0148 - val_loss: 0.0025 - val_mae: 0.0299\n",
      "Epoch 271/300\n",
      "1285/1285 - 0s - loss: 4.3756e-04 - mae: 0.0154 - val_loss: 0.0027 - val_mae: 0.0312\n",
      "Epoch 272/300\n",
      "1285/1285 - 0s - loss: 4.9697e-04 - mae: 0.0167 - val_loss: 0.0026 - val_mae: 0.0307\n",
      "Epoch 273/300\n",
      "1285/1285 - 0s - loss: 4.4275e-04 - mae: 0.0156 - val_loss: 0.0025 - val_mae: 0.0301\n",
      "Epoch 274/300\n",
      "1285/1285 - 0s - loss: 4.0525e-04 - mae: 0.0149 - val_loss: 0.0025 - val_mae: 0.0306\n",
      "Epoch 275/300\n",
      "1285/1285 - 0s - loss: 4.4599e-04 - mae: 0.0155 - val_loss: 0.0027 - val_mae: 0.0310\n",
      "Epoch 276/300\n",
      "1285/1285 - 0s - loss: 4.1871e-04 - mae: 0.0151 - val_loss: 0.0026 - val_mae: 0.0306\n",
      "Epoch 277/300\n",
      "1285/1285 - 0s - loss: 4.2370e-04 - mae: 0.0153 - val_loss: 0.0025 - val_mae: 0.0302\n",
      "Epoch 278/300\n",
      "1285/1285 - 0s - loss: 4.4673e-04 - mae: 0.0158 - val_loss: 0.0029 - val_mae: 0.0322\n",
      "Epoch 279/300\n",
      "1285/1285 - 0s - loss: 4.1091e-04 - mae: 0.0148 - val_loss: 0.0026 - val_mae: 0.0302\n",
      "Epoch 280/300\n",
      "1285/1285 - 0s - loss: 4.2121e-04 - mae: 0.0154 - val_loss: 0.0025 - val_mae: 0.0309\n",
      "Epoch 281/300\n",
      "1285/1285 - 0s - loss: 4.0929e-04 - mae: 0.0153 - val_loss: 0.0026 - val_mae: 0.0310\n",
      "Epoch 282/300\n",
      "1285/1285 - 0s - loss: 4.1191e-04 - mae: 0.0149 - val_loss: 0.0026 - val_mae: 0.0305\n",
      "Epoch 283/300\n",
      "1285/1285 - 0s - loss: 4.1276e-04 - mae: 0.0148 - val_loss: 0.0025 - val_mae: 0.0307\n",
      "Epoch 284/300\n",
      "1285/1285 - 0s - loss: 4.0545e-04 - mae: 0.0149 - val_loss: 0.0025 - val_mae: 0.0301\n",
      "Epoch 285/300\n",
      "1285/1285 - 0s - loss: 4.4697e-04 - mae: 0.0155 - val_loss: 0.0025 - val_mae: 0.0300\n",
      "Epoch 286/300\n",
      "1285/1285 - 0s - loss: 4.9919e-04 - mae: 0.0160 - val_loss: 0.0025 - val_mae: 0.0308\n",
      "Epoch 287/300\n",
      "1285/1285 - 0s - loss: 4.3907e-04 - mae: 0.0152 - val_loss: 0.0026 - val_mae: 0.0301\n",
      "Epoch 288/300\n",
      "1285/1285 - 0s - loss: 4.2561e-04 - mae: 0.0152 - val_loss: 0.0026 - val_mae: 0.0301\n",
      "Epoch 289/300\n",
      "1285/1285 - 0s - loss: 4.8013e-04 - mae: 0.0160 - val_loss: 0.0027 - val_mae: 0.0309\n",
      "Epoch 290/300\n",
      "1285/1285 - 0s - loss: 4.1164e-04 - mae: 0.0152 - val_loss: 0.0026 - val_mae: 0.0300\n",
      "Epoch 291/300\n",
      "1285/1285 - 0s - loss: 4.5301e-04 - mae: 0.0155 - val_loss: 0.0025 - val_mae: 0.0299\n",
      "Epoch 292/300\n",
      "1285/1285 - 0s - loss: 4.8114e-04 - mae: 0.0160 - val_loss: 0.0026 - val_mae: 0.0306\n",
      "Epoch 293/300\n",
      "1285/1285 - 0s - loss: 4.2978e-04 - mae: 0.0151 - val_loss: 0.0025 - val_mae: 0.0301\n",
      "Epoch 294/300\n",
      "1285/1285 - 0s - loss: 4.1636e-04 - mae: 0.0153 - val_loss: 0.0024 - val_mae: 0.0297\n",
      "Epoch 295/300\n",
      "1285/1285 - 0s - loss: 4.0533e-04 - mae: 0.0148 - val_loss: 0.0026 - val_mae: 0.0303\n",
      "Epoch 296/300\n",
      "1285/1285 - 0s - loss: 4.3973e-04 - mae: 0.0157 - val_loss: 0.0026 - val_mae: 0.0312\n",
      "Epoch 297/300\n",
      "1285/1285 - 0s - loss: 4.2550e-04 - mae: 0.0153 - val_loss: 0.0027 - val_mae: 0.0311\n",
      "Epoch 298/300\n",
      "1285/1285 - 0s - loss: 4.8418e-04 - mae: 0.0164 - val_loss: 0.0025 - val_mae: 0.0303\n",
      "Epoch 299/300\n",
      "1285/1285 - 0s - loss: 4.3954e-04 - mae: 0.0156 - val_loss: 0.0025 - val_mae: 0.0306\n",
      "Epoch 300/300\n",
      "1285/1285 - 0s - loss: 3.9520e-04 - mae: 0.0150 - val_loss: 0.0024 - val_mae: 0.0292\n"
     ]
    }
   ],
   "source": [
    "model = tf.keras.models.Sequential([\n",
    "  tf.keras.layers.Conv1D(filters=32, \n",
    "                         kernel_size=5,\n",
    "                         strides=1,\n",
    "                         padding=\"causal\",\n",
    "                         activation='relu',\n",
    "                         input_shape=(X_train.shape[1], X_train.shape[2])),\n",
    "  tf.keras.layers.LSTM(50,return_sequences=True),\n",
    "  tf.keras.layers.LSTM(50),\n",
    "  tf.keras.layers.Dense(128,activation='relu'),\n",
    "  tf.keras.layers.Dropout(0.1),\n",
    "  tf.keras.layers.Dense(1),\n",
    "  tf.keras.layers.Lambda(lambda x: x * 400)\n",
    "])\n",
    "\n",
    "#model.summary()\n",
    "\n",
    "model.compile(loss='mse',\n",
    "                  optimizer=Adam(lr = 1e-5),\n",
    "                  metrics=[\"mae\"]\n",
    "                  )\n",
    "\n",
    "\n",
    "history = model.fit(X_train, Y_train, epochs=300, batch_size=30, validation_data=(X_test, Y_test), \n",
    "                     verbose=2, shuffle=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "# make predictions\n",
    "train_predict = model.predict(X_train)\n",
    "test_predict = model.predict(X_test)\n",
    "# invert predictions\n",
    "train_predict = scaler.inverse_transform(train_predict)\n",
    "Y_train_Sca = scaler.inverse_transform([Y_train])\n",
    "test_predict = scaler.inverse_transform(test_predict)\n",
    "Y_test_Sca = scaler.inverse_transform([Y_test])\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Mean Absolute Error: 4.232238028328137\n",
      "Train Root Mean Squared Error: 5.468953295691056\n",
      "Test Mean Absolute Error: 8.26552727653987\n",
      "Test Root Mean Squared Error: 13.832681516463017\n"
     ]
    }
   ],
   "source": [
    "print('Train Mean Absolute Error:', mean_absolute_error(Y_train_Sca[0], train_predict[:,0]))\n",
    "print('Train Root Mean Squared Error:',np.sqrt(mean_squared_error(Y_train_Sca[0], train_predict[:,0])))\n",
    "print('Test Mean Absolute Error:', mean_absolute_error(Y_test_Sca[0], test_predict[:,0]))\n",
    "print('Test Root Mean Squared Error:',np.sqrt(mean_squared_error(Y_test_Sca[0], test_predict[:,0])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1a350a4de48>]"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(Y_train_Sca[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1a361b1d348>]"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD6CAYAAABamQdMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOx9ebgcVZn+e2rp7rtm30lIAgkQdoUIIsOiAgMqjuOCjo7LOLgwKjP6cwDHXUZcxn1mlHF3EAYFFfeNIIsQDCEsIQmEJGTfl7v2UlXn90fVV3Xq1Nr3dt++t3Pe58mTvtXV3aerT33nPe+3Mc45FBQUFBTaC1qrB6CgoKCg0Hgo466goKDQhlDGXUFBQaENoYy7goKCQhtCGXcFBQWFNoQy7goKCgptiEzjzhibzxhbwRhbxxhbyxh7n3f8DMbYQ4yxNYyxVYyx5cJrrmeMbWSMbWCMXdrML6CgoKCgEAXLinNnjM0BMIdzvpox1gPgEQCvBPAlAF/knP+aMXY5gA9yzi9kjC0DcCuA5QDmAvgDgKWcczvpM6ZPn84XLlzYkC+koKCgcLTgkUce2c85nxH3nJH1Ys75LgC7vMf9jLF1AOYB4AB6vdMmAdjpPb4SwG2c8wqAzYyxjXAN/YNJn7Fw4UKsWrUq59dRUFBQUAAAxthzSc9lGnfpjRYCOBPASgDXAvgtY+zzcOWdF3qnzQPwkPCy7d4xBQUFBYUxQm6HKmOsG8AdAK7lnPcBeBeAf+aczwfwzwC+RafGvDyi/TDGrva0+lX79u2rf+QKCgoKConIZdwZYyZcw34L5/xO7/CbAdDjH8GVXgCXqc8XXn4MAsnGB+f8Zs75WZzzs2bMiJWMFBQUFBRGiDzRMgwuK1/HOf+C8NROABd4jy8G8Iz3+C4AVzHGioyxRQCWAHi4cUNWUFBQUMhCHs39PABvAvAEY2yNd+wGAP8I4MuMMQNAGcDVAMA5X8sYux3AUwAsANekRcooKCgoKDQeeaJl7ke8jg4Az094zY0AbhzFuBQUFBQURgGVoaqgoKDQhlDGXUFBQaEBODJUw8/W7Gj1MHzUFeeuoKCgoBCPD/z4Mfz+qT04Zd4kHDeju9XDUcxdQUFBoRHYeXgYADBUGR/xI8q4KygoKDQAGnPjTuxx0pdaGXcFBQWFUWLFhr14YscRAEDVclo8GhfKuCsoKCiMEj9etd1/PFS1WjiSAMq4KygoKIwSkzpN//FQVWnuCgoKCm2BSR2Bca/ZSpZRUFBQaAuIxl1p7goKCgptiJqtomUUFBQU2gI1ga0rWUZBQUGhTVBzAraujLuCgoJCm8ASDHpVGXcFBQWF9kDNdlA0XHNas5TmrqCgoNAWqNkcJVOHrjElyygoKCi0C2q2A1NnMHVl3BUUFBTaBsNVGx0FHaauKc1dQUFBoV0wVLXRaRoo6Jpi7goKCgrtgsGqhc6iy9yVQ1VBQUGhTTBUtdFZ0GEaSnNXUFBQaBsMVix0mAZMTWnuCgoKCm0Bzjl2HBrG3MklV5ZRxl1BQUFh4mPnkTL6KxaOn9ntyTJKc1dQUFCY8Pjd2t0AgHMWT1PMXUFBQaFdsGX/IHpLBpbO6nHj3FU9dwUFBYWJj8Gqje6iAQAqzl1BQUGhXTBUtdDpGXdTZ6hMFObOGJvPGFvBGFvHGFvLGHuf8Nx7GGMbvOOfFY5fzxjb6D13abMGr6CgoNBqDFZsdBV0AMDMnhL29JVbPCIXRo5zLADv55yvZoz1AHiEMfZ7ALMAXAngNM55hTE2EwAYY8sAXAXgZABzAfyBMbaUcz4+WoIrKCgoNBD7+it+D9UF0zqxf6CKcs1GydRbOq5M5s4538U5X+097gewDsA8AO8CcBPnvOI9t9d7yZUAbuOcVzjnmwFsBLC8GYNXUFBQaCX+suUgntrVh7U7jwCAb9ArtUCa4Zzjjd9ciT88tWdMx1aX5s4YWwjgTAArASwFcD5jbCVj7E+MsbO90+YB2Ca8bLt3TEFBQaGtsHaHa9T7yhYAoKAzAOFuTDWb4/6N+/H2768a07HlkWUAAIyxbgB3ALiWc97HGDMATAFwDoCzAdzOGFsMgMW8PBLVzxi7GsDVALBgwYIRDF1BQUGhtZjWXQQAvPjEmQAAU/e6MQnG3eGtSWrKxdwZYyZcw34L5/xO7/B2AHdyFw8DcABM947PF15+DICd8ntyzm/mnJ/FOT9rxowZo/kOCgoKCi0BGe7rLz8JQLxxt51xatwZYwzAtwCs45x/QXjqpwAu9s5ZCqAAYD+AuwBcxRgrMsYWAVgC4OFGD1xBQUGh1SBtvWS6ptQ0osbdapFxzyPLnAfgTQCeYIyt8Y7dAODbAL7NGHsSQBXAmznnHMBaxtjtAJ6CG2lzjYqUUVBQaEdULNe0kSPV19yFmu7OeDXunPP7Ea+jA8AbE15zI4AbRzEuBQUFhXGPssfcix5jj5NlWsXcVYaqgoKCwggRYe4xssy4dqgqKCgoKERRrjnQGGBorrhBzL2qmLuCgoLCxEXFcjNR3bgTUZYJDLrdovruyrgrKCgojBDlmuPr7YBbFRIAakLxMFvJMgoKCgoTCxXLRtEIasiYhsvgw3HurakSqYy7goKCwghRsRw/xh2Iau4128FQtTWR4LnLDygoKCgohFGuhZl7QdLcr/jKfXh6z0BLxqaYu4KCgsIIwDnHH9ft9aUYIBznXrHslhl2QBl3BQUFhRFhzbbDsByOJ3f0+cdMPdDcb/zlulYNDYAy7goKCgojwnDN1dL/6aLj/WNUW6ZqOVi/q78l4yIo466goKAwAgx7jtKXLJvlHxM1dx6tdD6mUMZdQUFBYQSgKJjOghAKGVNbhnD8zO6xGZgHZdwVFBQy8fSe/pbVJR+vIFmmQ+iVqmsMGnONu5y7RH1WxwrKuCsoKKRi494BXPLFe/GlPzzd6qGMKwzHMHfAZe/VGObOxzhTVRl3BQWFVOw+UgYAPLr1cItHMr4gV4QkFHQNNavVirsy7goKChmg2igsqavDUQpKVCKdnWAamifLhM37WBt7ZdwVFBRSQZ2EdE1ZdxFVrzgYxbYTTJ25xl06f6zrhynjrqCgkApypGqKuodQsx2YOvPL/RJIc5eNuWLuCgoK4wrkHFTEPQzXuEdNaEHXQvXcfSiHqoKCwngCRYUo5h5G1Yo37qauoWbFyDJjMywfyrgrKCikguK5lXEPo2rzeONusNhQyLGGMu4KCgqpKHvGXTlUw6jZDgp69JqYuhstI8swyqGqoKAwrkCyjCLuYdRsxy8UJsLUNT+SRsRYR74r496m+PSv1mHJh37V6mEotAFIllHlB8JId6iGNfeT5vQq5q7QGHzj3k3xHnsFhTpBBbIqMWz0aEbV4n4VSBFunDv3jfmi6V2YN7mkjLtCY0Ep0goKIwVp7mouhZEmy9RsB45nzd0FgKloGYXGYqBstXoI4wrPHRjEw5sPtnoYEwoky8TpyEczEh2qhpvERDudotkaM6uMe5tjoKKMu4gP/2wtXvuNB3FwsNrqoUwYDCtZJhZZmjtdt4KugTFVFVKhwehXzD2EP2/cDwDYenCoxSOZOCDmXqkp4y4iMc5dZ6hZHENV994rGBpaEWiUadwZY/MZYysYY+sYY2sZY++Tnv8AY4wzxqYLx65njG1kjG1gjF3ajIErpIPC1pRxD2P2pBIAYNUWJc3kwWDFwn3PuAui0tzDqCVkqBa8qpAnzu4FALz7wuM95j6248vD3C0A7+ecnwTgHADXMMaWAa7hB/BSAFvpZO+5qwCcDOAyAP/FGNMj79pm+NPT+/CVPz7T6mH4IC++kmXCWDS9C4DbgEIhGyI5UJp7GFXbQcGIT2Kq2g56SgZOnN2DFy2ZDgY2/uLcOee7OOervcf9ANYBmOc9/UUAH0S4bMKVAG7jnFc455sBbASwvKGjHkdwHI63fOdhvPnbD+MLvx8/nWoC415r8Ujqxw9XbsW//fSJprw36cb9atHLBbEXqNLcw8jS3B3O/azeViSA1aW5M8YWAjgTwErG2CsA7OCcPyadNg/ANuHv7QgWA/G9rmaMrWKMrdq3b19dgx5PGKrZuGdDMH7S2VoNmkwDlYm3lb7hJ0/gfx/amn3iCEDGSkUR5QNdr+6ioYy7hCRZxvDi3C2Hh0o2jEdZBgDAGOsGcAeAa+FKNR8C8JG4U2OORb4W5/xmzvlZnPOzZsyYkXcY4w7kESccGBgfURiWl01YUzdkCCQtKLkqH6gA1vTugtLcJSQ5VA1Ng+1w2E6YuY/LOHfGmAnXsN/COb8TwHEAFgF4jDG2BcAxAFYzxmbDZerzhZcfA2BnIwc9niAbd4osELFi/V5sG+PoDGJclqOMuwjfuCvmngs1yzVJPSUTNZv7XZkUgOGqFWmODQSdmco1G7q3hWZg4y8UkrltRr4FYB3n/AsAwDl/gnM+k3O+kHO+EK5Bfx7nfDeAuwBcxRgrMsYWAVgC4OGmfYMWQzbmceFib/3uX3D+Z1c0dRycc/zgwS04MuRq7FR6QJUgCIMWvQ17+vHZ36xv8WjGP6q2O7+7i4b39/ghC5v2DWDhdb/EQ5sOjPlnOw7HYNVGV4xx1zXXrFYsJ5BlGLD90DBWemO97eGt+MGDW5o6xjzM/TwAbwJwMWNsjffv8qSTOedrAdwO4CkAvwFwDed8XO7nqpYDa5STVdbYy9LWday2so9tP4IP/2wt3nXLIzj303/0j6sIhzDE6/Ff9zzbwpFMDFQ95t5dco37eIp1//OzrqG867GxFwaGPFLX5S16Ioi5V2oODJ2Yu2vsX3fzQwCA6+58Ah/+2dqmjjE6Mgmc8/sRr6OL5yyU/r4RwI2jGtkY4KSP/AYLpnZixQcuHPF7ZDH3sdr+0yJFE55QG0dMKw/EratlOzBiNM3RYDwxz4kAul49nhFzyYrZwhEFoLnSijLzQ57PJs64G96AKpbtNzgR+6yOVY7FUZ2hajscm/cPjuo9LEn2KEvGflCIVlm3q29Un5WGOCNIIVkTCbevCgKtmmGI1U6mPpBDvoeY+zi6fiT/sxbkf1IobXeMcdf1QJYxYlaeT/7iqeYOzsNRbdwJ9z498lBMuca1PPn3DZT9x3/7338e8efUi6KhoaOgTzjN/fdP7fUfN8MQV20Hr1/u+vtPnN3T8PdvN9AC2z0OjXsrmfthz7c1qTO6izF95h5o7uIQq2N0TyrjDuDvvx319z61sw9/ybF9kpmxzNwf3XrYfzxUbZ7+LvsObr36HD9TbiJBvFGbYtwtB9O6irhk2ayGv3c7guZ3T8k1YuMpHNJn7i3IEDo85IY8T+ksRJ6jXXSlZscmMck2ollQxt3DzsPDob8v/8p9eM3XH8x8XRZzF0MlF8/oGsUI0yEbwuctmIKCziZ0nHujWaLtcDjcTQ/v7TDxzN4BFdqXAfoNuovjj7lTvfRWZH9SVdEpMczdyGDuYhBGM8Mjj1rjLt/UIy0BaznpmvtwzYahMVx5xtymGpI4hl4q6Nh1pBxz9viF+D0aveugBbBgaCgYbqLJ+/5vTUM/o90QMPfxEy1zz4a9uOnX6/2Mz1Zo7pv3D8LQGOZM6og8RxEylsNjHapHhoOSILL9aCSOWuMuhyyKF7weZDH3cs1Bh6mjw9SbKsuQtn7O4qn4yMuWAQBOmzdpwhXIEo1Ho53BonGnRfjnLQijm0ioScx9PMh8b/nOX/D1Pz3rF+Jqhea+60gZsyeVUIjpxGRowbE45l4W5ngz+9Ietcad5JLXL18AAOgboXEnA/TF150OIJ65F00dHQU9ks3aSNA4Pv6KU/C2Fy0CAHQWjQmXoTqrt+g/brTmToapoDNcfOJMAK3Z0k8kVGXNfYz04jwgu6g10bpzzvH49sOR4+WaHZudCiAUIZM1smZGsx21xp1Y9Ixu1yEy0lojtPKes3gaCroWw9xtdBQ0dBZ0DNXspmlsZAhNoe2XGwo5sTRlMW64acbd0PCy0+Zi0fQunDl/ckM/o91A82c8au507zVzff787zbgFV97AGt3HgkdL9dslMwE4y7cgyTLJA1SMfcmgBj2ZM/bPdJJS5qZrjEUTS3K3Ks2SoaOzoIB2+FNM7ZkuMRCRobGJlycuzjZG23caedEN+Wi6V3jQmYYz6D7oquoh/4eDyAptZhgZBuB3z+1B0B8DamSkcTcBbPq2/Z4695M8nXUGnfKLJ3S5W43RxqeRMbI0DQUDT0SKtZfqaG3w/QNSrOkmZrASgmmoUWSrMY7RAdTpcGGl37jDu+3KOiaSmrKgNsEWvPn73i6XhThpjFg495+7DoynPGK+kHMu1yL+tKSGl/HMfck+U8x9yaAZJnRMncyqrrG0F3UI/XT+4Yt9JYMX58bqjWnHAE5vgoCczc1hqrtjHk1utGgmcydjDsZqoKhjHsWyjUbRUND0aCsy/Gjuf/i8V0AgGf2DOAlX7gX53767oZ/BjlE5TIjabKMKI0y6X8ZSnNvAp7a6ZYCmNFdBGPJjqKs8EUyRqbOMKnDjETd9JVd5u4b96Yxd28cInP3DH0z2UGjITL3Rk98ukE7Csq458VgxUJ3yfB3hONJliH88oldTXtvI8G4Vywn0bjrgiyjmHsL8AmvvsP8KZ0oGlFHKOGhzQfQX06OpBE1994OMxJ1M1C20FU00FlwHVKDTWoSEWjuwSyiTLmJ5FS1HcfffTRLc+8QmbvS3FMxULHQXTT832Q8xLnXi837B0e8e6VInHI1hrnHhEECUrRMhre3mdFsR6VxF9n4pE4TRUNP1Nzf8D8r8e5bVie+l6i593aY6JMWAsvhMDWGGT1uiN++/spohx8LP1pGE5m7O7NqEygcsmZzn1k33Lj7sox7jeKimxTC6C+7zN3QNRgaGxeyzMJpnbnPXbnpAC76/D2hgnT1QGbulu3gZV+9D7uOlP15KkMMamBCs444qCSmBoMSmK776xMBuDd72k2++rlDic+5ZT1dp06HqUeYjeNw6Jrmx2/v7mtOxmjNdivQiTG/NMkmUgkC2+F+A4RGs2r6bWg7XVSyTCb6hmt+GGTaDncsEdfaLglPeZVY1+7MV5F1b18Zh4RsdZJVyLgfHq7hyR3ueyXLMlHmnsTgmxnwcFQad3l7LjN3eQtnp2zphqsOOgsGGGMoxEx+t0kuMKnDjcrpb1J995rtRLLl6CZoJjvIg/W7+/wOUVmwnOYxd2KdRSMsy0wkh/NYY/uhYcyb7KbYG7o26uY2jUA9i/6wFCGVhD19Zdz12E4s//c/4hyh2Q0Z95t+vR6f+c36UJPrJFkm1qGaZNwVc28syp7RoB9cZu6ykyNtdR2u2b4xcplNeNtqcw5NY35MbDMdqjKjoZCsVrPTy750H153c3YRNsDV3Mk/0WiWWPaZeyDLcN76xW+84shwDQcGq1g03S14Z+psXFyreuZzWdqtJeHt31uF9976KIBg3m3aNxAKkPjve54NaeRJ8fVibwU/iSlJlmniYpnZiakd4SezFOKZu+yATJvQ5Zod2gHIBslxuC+XdJg6hqvNYe4Vy4kYd1MoYNQqECtev7s/1/mWzf3r2ehomTjmDrjGop6t/tGCLV4jm4Wecdc1Ni7yJqqWExuZFgc5/DUJcoE9zjku/o8/Rc6jhuFp7xnnUBX/F9l/M2/No3JGy8ksMnOPls9NTlEfqlrh6AsrvM23HO53QO8o6JGQqkahYtl+LDLB19xbuJWud2GxHQ7TYDA01oQ4dwcaCxa98RzeNx6weqvrazpl3iQAbtDAeGHus3tLuc4lIpe1dst1YuSkJf+z7eD+LSUlMQnGXZP0mL9aMiP0t6NK/jYWsg4nM3dZ04vrkxi8lyPsALTQ6x0/TFLzP69ZskzFimbMURp0K417vZ9d8xzQlsMb3sDaXQB1P4JBZO4KUew6UkbJ1ATNnY2LQnQV28GsSVHj7rLisLGkez3rN5aNe1L4c1Vg7nEt9oD4lpdk4nWpyFkz3T1HtSzTUXB/hJKp4dBQcqnZtNW1XLXRaYaNe8VyUDR0IQYe3uclh1yOFpWYWhcFwwuFbOFWWtzG5oHtOH6bskajXHNCbKtZ8fQTHdsPDeFFn1mB6d1F9JaCZhTGOJBlOOdeN61oByTO3Z0f+Zqu/M8H8Ng2t6Jj1m8s6+dySDNBtA3kG5IRx9yJwMtTWzXraDCGJR2us2CEqkLKxj0ti0x2qAJByB0tCsTc6705frt2N3YczlcvI425tzLCoSJsYzfs7sevM7IJLZtD1xjOXTwN07uLqefWCzll3Gfudutjt8cSFcvGz9bsSIxZ/96ftwAA9g9U0NsRGHdT11rO3ImoUPMQGeKumww7kF2nSPaFyWVEgs8P3oeKqcnoLOiY4+0sgqKQ7gONMbzurPk+61eae4MhO1lm9hSxt6/ir6KydJL0A/zqiV04PFz15R3DDz10J4AtMXeNsdwam+1wvOMHj+C1OVr90XeSmTtp7q3MwqSbUdcYLv3SvXhXSkIYEDCvY6d1Zuqk9cLdUQVvWjxKNfcv/v4ZvO+2NXhg4/7Y5//nvs3+Y9GIjgeHKs1lcUchImmnmMXcZedsEiES76WuBObOGMNsz7hrkkNV1xg+8+rT8N23ng1Aae4NhxznPrO3iOGa7bP35w4Mhs6Pqy+zctMBvPuW1dh2cNhfJGg7RjeAJWnuusZy15KgMgZ5M1rjmLsfLdPCG5JuKllrTILtae5xOQOjhczcKWrmaJNlth8aApAv50KMDzf01jtU6bfq7Uhn7vI9mzWXBqRr0S/s5LsEPV6cK2m+OLIFQYaqC7ntXjOv5tFp3CWHKnWZIeO+b8DNUHvguovxouOnx66ufcJk6CxIoXWyQ5X0No0hr5095HVXT9p+ykhj7uPBoSrqkGnF2GqOm2nbjHK8MnM/Wh2qtNAm1YnpEYyWeL1MbXQO1cGKhVM/9lvcvX7PiF7vOBw/XPkcgGTDSvfekOTbyvqNZb/U1gPuAnj2wil43rFTYt8nSZYBAklUpjSUQU63g2LuDQaFOZFWLldspC1Zp6lD01isLCMSUXqfQOOWmLtnZHWWXWWSQNvEzpQJJCJWc9db71D1mbsQEpbGomxPcy9mlIQYCcpey0OCvBgfLfjZGrdv7FBMzkW5ZocMTlEgDLrGRjWXdveV0V+28OGfrh3R63/8yHZ8/ndPAwiXtqaxAYHkSt9tRk8Rx07rTDXunPNI/aXnPON+w+UnhYiJ6JtLZe66xNxZ2KgTg3/rd/6Cvf3NKUlyVBp3Yu7ESojBk1zjs02dQWPxq6sYv+rLMr4xlRyq3rn1yDI0GUNdXVJQqTnRaBnJB9AK+N9XWAzTIoYsh8PUGQq6DtvhDS2JWpbKtB6N0TLitR+MCcvddnAodFwkDKaujer3oPtgpPWVaDcLIFJq47gZbqLVs15DeFqEPnDJ0swaQrbDIyGJWzxp1tA0dApGXNTmO1MSowJZJnxcZ9Hj92zYl/g+o8FRadwpq5RWUwppGvKNO9Vo16AnOUFF5u79yHJGaJxDNa1OjQhik3mjAsuWHcPcWy/LxF27tEQuV3Nn/ndppOGt1OyjXpYRMzHjuoJt8/R4gni9DH10iWU090e6QIiEqmhouOXtL8CJs3sAAAumupUi9w24PirafQf+m3RCIWOzl51r6CzE3KlUuPtcsvmk52SHasDgg/dslk8s07gzxuYzxlYwxtYxxtYyxt7nHf8cY2w9Y+xxxthPGGOThddczxjbyBjbwBi7tCkjHwWGq3aoXCc9HvS2cpZg3BljiCO+orzSGZFl5GiZwKGaV5Yhg8zkpT8BlVq0eYA5DmSZOOOexdwNTQvqhzewxKzcYOFolGVE5hn3vR/adBAAsNhjwqIsM39KJzbvH8w9h2WM9HUE8VYoGBrOO346jpvZDUBo4O1JrmSw3V1get1++bnJnabvdDZ1hiXeZwBB0tHxwrE40L1HIZD0P60H4nexm7SzzsPcLQDv55yfBOAcANcwxpYB+D2AUzjnpwF4GsD1AOA9dxWAkwFcBuC/GGPN62A7AgwL9WCAwDFCTMZyHDDmGuMkWUZkwwFzDxsLmbnrWh3M3crP3DnnLnMfh+UH4u7ntHucmHszWHWEufsLyNFj3A8OBtFXcdd2/0AFs3tLOG6Ga7zE63XMlA4MVCy/ZHa9GG2kjch2C3o4t4T0bxobETTqbZw2j2TmPK2r4BMiQ9PwzguOw7+8dGnonB+949zUsRKhk+9fLYa5N4t8ZRp3zvkuzvlq73E/gHUA5nHOf8c5J+/CQwCO8R5fCeA2znmFc74ZwEYAyxs/9JFjuGaHMhU7zagsY4psO8YgiwaByg/IoYdBdyT6ofMz94pv3LOte812NcMocw87eFuBuC14WlYe1aVvRgy6q7lH49yPJlnmbd9d5T9+dGu0T8Fw1UZPyfBLVItp+aNdcEfrP9Ek5g7EGHePuYt+s66inhr2STvtT77yFKz/5GUoSE5kQ9dw9sKpodeYRrrpNCU9xs9QjdHim9Vqry7NnTG2EMCZAFZKT70NwK+9x/MAiG1PtnvHxg3KCbIMZanVbMc31BqLd4KKE7wzIYnJbw7hTZZ6mDut5nmMe9kKO4gJsoM3Ca5DqTkTzF8Yc1bCk5l7I427y9yjxupoYu4iHtt+BL9duzt0bLBqo7No+DozNZAHRp8UN1ojJkqU9NvR7svUvX4K1DFJkGVm9BSxf6CKJND3KeoaSlLlVvrOcp6GWLM9DnLIYxDnTv8LzL2FsgwAgDHWDeAOANdyzvuE4x+CK93cQodiXh75VRljVzPGVjHGVu3b1xxvcRJkWSYuFNJ3iGgstriP6BQMQiGpfrr7AjK6xKjdhSLfGH1ZJscvRIuIXB+j4Msy6TfVcTf8Cu/4wSP5BlYn4uYtT0jd4Jy7mruuNSXBSGbuR6NDVcYze8KlmIcqFroKeqijF2HUzH2UBEIkKTS3aUy65kbEfOPeTQACNm5oGqZ3F3FwsJK4uPgSjmewtxwInMp07KQ5PaHXmBk3ph8tgzBT11nY6ANu+G8zkMu4M8ZMuIb9Fs75ncLxNwN4GYC/4wH12w5gvvDyYwDslN+Tc34z5/wszvlZM2bMkJ9uKoalTEUy9KLyxgcAACAASURBVL4s44XjAUjU3A8KrbiCGjXu/+QwDMoc0ATM71Sq2fllmXItgblr+Zg7APzuqZEllmSBrp2Y8ZdEVOjShGWZxjhUa7YD2+GhcNGjMRQSQMhBKE+Nw8M1dBUNXHHqHPfcWcG5eclCEkbrUN0jhFCKRh0IJ8kBwRgNnaG3ZMLh0bj+l3/1frz9e6v8+yOupj8Z8Z6Sie+9LVCX4xY/EbL8Ise7sxBzb120DAPwLQDrOOdfEI5fBuBfAbyCcy7GT90F4CrGWJExtgjAEgAPN3bYo0O55oSYu6YxlEzN//Et2/EjX+LCFx/YuB9/ejrYbdB7UZElqignd4GpR5ahmN7Htx/Bqi0HU88lWUHW3HWNgbH0wmHNbjEX932TsvLo5h2u2f7Nu3rr4dhz6wVdIzFc1NA1aKy9C4f95sndeNpj55xzMAZcdsps/3nRng1WLGzcO4CT5/bivOOnY81HXorzhfrjo2Xuo3Woig1fikZAmACXhC31FqIndxzxpVFT13w9Xq4Z9cSOI/jDuj1C6HPUYJtGOPwyL2b2uEXvdhxyC//JJX/FT2pWYb88oz0PwJsAXMwYW+P9uxzA1wD0APi9d+zrAMA5XwvgdgBPAfgNgGs45+Pq7pFrjAAuK6EfuWYHZUO1mFDIv/vmSjwqGB1i7FQqgJw3MnPP61DlnONLf3jG//t/7tuU+X2A6ORjzJWUnt0/GPcyAM3Xm+MWj6T15LteNcLv/3mLnwn4SSGueDRI6shDDVbaFe/830dwyRfvBeAaV87D2Z3ib3HYC5OkRhii3g6MPvpqtMxdLPkhJzEBwOdefToAl3wF0TLMj4YbrMQ7VeUqsSJEEpjVzUnEkpmujNNfCRcki9Pcm1WvJ0+0zP2cc8Y5P41zfob371ec8+M55/OFY+8UXnMj5/w4zvkJnPNfp71/KzBctSMNcwtCi7ya7fg3QJIsI6JDaNdXNDS/6FdgdOtj7nv6wsXCsqSZJOZO+OXjyWV24xJZ4vCLx3di4XW/TLxBkhBnB5KuJzWFePv5i3HO4mkAgAuWNkay85m7ZBSaUcNmvIK+p2gYxflIDSrEMr8iiNmOlBCIRuzrf6q/EUu/EKMvlx8AgNPnT8bs3hLW7+4PSS1UvXFQKOO7XUjWkgsJihDlk6TOS3Gg4cl6flwoZMuSmNoRQ1UrFC0DIJSibMnMPcMgd0qRN77mLkkBekLkjYzDw2HPflbn9koCcxeRJL/IBZaS8LW7NwII0rLzIu7aJV1Pmu9vOvdYTOowccyUjtimDCNBEnMvmvpRk8QkLnDvufh4AGE2TTvOpGJ1tCiMlLmLCwnVjK8HWcwdcH0Ez+4bECqyMr8+06Cgub/oMyv8xyTHJjXfIMjlPdJAvgCa6/TeNC6RrzWrPMhRZ9wdh6O/YqFXmsAFI8hisxxBc5cKh4nZlfd98CJ8723L/aqSgMcEvZW4IhkULWeG6m6pWW+pkGHcM5g7APQNxzPuoZxMvNMPF61PYYv7vkmXgJgVMa3OQn1tCQcqFt5z66PYGdPg5DVeXfw45t4MaWrH4WG88weP1L3TaSYC5q7j/ZecENlJ0o6zJ6FW+mgd0OJcyBMoIENsfUcSEZOC83pKBoaqdigUMivfw68S681xqrUuox5Zxmfu3ji6pGAL8eu3LIlpImJvXxn3PRMfXjlQtcB5dOtp6gw1X5aRomWc6A0AAPOndkZkA1PY5vtskWSZnLVl3vKdvwAAPvbyZQCAH67cGuoqI6OSEOcOAJ+88mT3OyWwg7zG088FqLNNYJwhT9pFDFYsaCzY/nYWjBDbysKNv3wKP39sJ7734JbIcxTdJIeLlkwNd67egcNDyXHQI8EXfvc0frN2d6okNtYQE3sAkhyD54m5y8SHMFrmLsoyOevh+ahaTmgRlp2fNKWKho6KZYdCIWkhSbr36B4gAjM1YbdYnyzjnks7dSo+RvePuLi1svzAhMPVP3gEb/rWw7ElTY8MebqixE5E5u4mMcVnqFJtjq+8/szYzy4Ymh92R+VJ/YVCyx/nDgDHzwxia//1jscTz5OjckToWjpryW3czfiIgyzER8vEnztYtdBVMHyds8PM33P2wEAFtz7s5s6RQ5BgxcRHE57d58pMH7trZGVok0DShtzhZywh75poHlPYoOzgJ2acxNzNBjJ3mXFngRzsLz99Lt55wXGJNZeKhoZKzQlp3RShkrRrpkYdFFVz7LQuTO8uhEIfgTqZOy0oZNwl5q409xFi1xF3W/7s3qg+TAxuirQ6i441WXMXdfJHnnNTticlOJ3cqBvHD4ECAqdM0TP8eWEI7GT97n4MVix84EeP4aFNB0LnpTH3rCzVuAUwDtRsu94G33EsPUlzH6rYoRrZhp6/friYXi5X6xMZ4yGJodM9Vu+ilYXuhPC7sYS8WxN1aPpfnNt9OTX3kfooRhMVQgvPhUtn4Lq/PtE/Ltv4kkcIakIopGxoZezuK6PD1P0dy6QOE6v+7aWRXXk9oZB0jekz5bLiIVlmPJQfmCigSfjDh5+LPLfdM7rHTOkIHRfllJoTMHcmbV1/8ugOAPAb4MowDbcs6mPbozJKMaP0aOS9pK3nHau348ePbMfffTNc/YGYe9zkk8sQyxCNT1rMO/kg6jXu8bVl4s8dqFqh5iT1NGQWv19FGqO4sF14wszQczO8JtxGRjp5vSCm2ej3rQcyI3Rk4y7JhH3lGgqGlshQszT3Xz6+C/+5YmPieMRFXSZXWaDFuzujM1nRa88ohkKSBJRUcnh3XxmzJ5UyK7DmrdAKBL87fRZd0+EYzb2Vce4TDmQPaJsu4oBXFW+6d1MTQg5Vm/tbV9nALJvbCwBYOiucjkwwdQ1/XL8X77n10chzNPHyJg7pkjC5w3MUyk0CKlZ8JAgQGOUkXU90kKY5Fmmy1utQradwmJv6PrKGzOJvJH8PGsNHXrbMZ9SEj77c9UnMkObDaDBctf2Y/awaJM2EvKCTcfNT4DVZlokGGogINPf43+SaH67G5367IfH19DucOLsnsgBnISuSh0ZE91jgX9AiLFr24+w+Usas3ny//z9ddDz+9bITM8+jz7Qk405ETJRlmpVH2J7GXbhasnEJ6rCEv7oYCilq7mJyE+Aa/iSHCxBNYb7pVacGn2Hq4Dy/d1xOqd552I2ikaNn8jD3pM8UJ3qahEBp2MMJfTeTEDdxk3ah+wYqob6Ups5yb+XFRSDK3JMzEK84bQ5KppYYWjcSPLM3yKSsV1tuJGRGSNdILGolXt6+4VrEFyUi0NxHV/K3aNTfaPuep/cCAHqK4fHJV5cc5jSXTaHZhm/chQimzoKOPX3liJ8mCR+49AS868LjMs+jzyRb1GEma+5fe0O8/260aEvjLhp0eZUmdi471pIcqtRajN6zajmpbEw2sK87e37kubzSjLxQ7PASL+S494plw/BKk8qox6GaFglDBqHeaJm85Qcqlo0nd/Th9Pl+zxcYmpZ7yyoai3ICc5d3QoTOglH390qDmITWrOzDPJA/W3aoyqGQv3h8V2qFwizmngXHN+56XRUiLdvBN/7kZmlnNYyne4wMuBgtQ9//gFch8rgZXRiq2jgyXEv0oY0U9Jl038nRZizm3EajLY27uC2X6zjTc7JxFzV3tzKhe8Hl8C/R8MdBfO7cxdNCOl2eGuXiNlnWa33mLu06Dg/VErVIw9fcs2WZ4RTnKsk6+wcqiefEIW8S02Evimn+lE7/mKHlZ+6i7JSkuSfp3x1mffH0WRAXimZ2t8+CfO3oby0mWoYW0aR8CCDY+cQ5VPNIjbSQFM36AgvE1oARIywZRmLuvs9DYxFZ5mVfvR8AMNfLiO4vW5m5JPWCasucv2Q6ANfHd87iqfjs354GIGzQm2Tbkb4MTlCEmgBLSSTEvOWqbhTlArirre5r7kHKdcnLZkzbwousXja4VIYgzbiLzEkuK0qRPvJKv3HvgN85JzIeYu4JRlLc2QxXs7vVPLH9SOI5cYgLP4uzAxTFIkpehp5fcxfZZLkWz9xlmYvQUdDr9iWkQVwkm9WIIQ/kXY/jxDB37xjtdt5xweLE90tzqH7/wSB4gWryy7AF5l5P+J+4mMiO2DPmTwIAnHrMJO+93TEOeDkTmsaCOHfpt5g7KQiqyMoCrxcze0t44LqLMcsz8qau4barg+5NTLi163HU1oO2M+57+8sh41mRbvSq5cTWpRALSHHOfadTMYa5x70+eJ9gkshbSNL505xJ4qSXqxXSJB+QFqwjwzUcO60TccgKhRzOKcvUBFmqHsS32YsepPwDkZkZen5tVrxxZdmLdi1JDY07C3pDZRlxF9BK4y7/Vr48JdQ3ITZNhKgrJQWfMQZTZ7HMXaySWrWcSHkP8fOLZv4oKCC4J94do3VffOIsPHT9izHbi16j+/XRrYf9uSczd8JsIeKt0cYdCGolxaFZUkzoM5r+CWOMjXsHAACXneyWNZVv9KptxzJv0bg7PLj4cuKGy/zzMXfZOZVHlhGN+7HTuvzH4lyQjTsl/6SNJ4/mnhbzTtm79cY4xxm3OHsXV9jLlWXyfZ64eMnM3cpi7g2WZcaLcU+KGiJjp2nBLkquYJqEgq75c0GEyNSTwmUD5q7VpdvTbyv6Y0SIRpp2x1sPBoXBfOMukYqZQoRM3GLUTIyFm73tjDtN6OcdOzn0N6FqxcsqYod0h3M/NjaqufNUh6rI6uWwu3pkmU9ceXJoERGLFslS02DFDsWHi5DToGXs7iv7N/SHfvIkFl73y9jzaJGsn7nn09zjKha6DtUGMHc7w7g3XJaxwZi7sI62+1C9qFqOHzIrzjPOeRAKKca5kyyTkuUswhQCD0ToAvtImt9hWaYO5i7UicmCHAUHBERNlginCiWN6ykK1ggo5j4CkORB2/sIc08y7t6k5ZzD4YEOJtewrmY4VEXIIZM+c88hyxiS3i46A92G2EJEUCWZudNiE3fD3bl6Ox7efBC6t90moxDHvCojZO5xhjzO3vlRTKJx1/Mzd7FAUyJzTzAOboGyxhX4Gqra6DT1uhqiNwrX3fk4zrvpbpRrdmjur93ZF8lQ1bSoLFPMMHKib0qE+FmJzJ0HzL2eKCKxTkwWRCN9w+VuPHqSLDNXkE0aGQqbB2Ng29vPuNON7Rt36Uav2Txec9c1cE4NDbhfVF9uolyJafQhQjSiLzlpVug5X3NPY+5SZMffPu8YANGoH9rW2g5HJUHjBIJtdlz45a+ecJsjD1ZtLJoeSEB9MfVQKoIsVQ/yFg7zmbsuyTJ5k5i88zqLRgxzTzcOHaYRWRBGCs45vv3AZgxW7Uh6/1jgF4+5hcpWrN/rRyABboRIXIYqHQsS4dJNgplQRVN0dCYyd1vU3OuRZdIXZxEic6e5FMgy7nFTZ3jXhceFigeOdSZxPaUMRoo2NO7uJO31mXt4olUsJ+T0JIgtxBzO/W2T3DdyqGqH6rfLEA3L/Klhh0pBz5ZlbGkL+vnXnIbNn748cp6/k/C16iTjnlyqV2Thzz92iv/4sGDcByoWhqqW/73qrQiYV3OPlWU8h2qeMDti+L1eydfwc1myjNYw5i6GEup1hHI2CvSbvuuW1Xj3LatDz8mau7j4VGKufxyS9PLJHYFxz2bubpx73kxtsWVeForS/AFE5u5g495+1Gy3l654H+fZFTQSzYqQEdF20TJlSZbpl0MhE0IZRflFdKjKfSOHa9EuTiJop/CNNz0/8gMWU1g0wY/s0Ki2TfwkIKYanB9/XknKjBNBZQyWL5qKE4RyCiLjO+WjvwUAPxqnHuZ+/zP78f0Ht0SOf+IXa/E/923CT685zz9WiZNlhO10FrMigzOzp4RN+wdCzwUVApOTmPI4VHccHkbVckK7HBn7hDwAuaJoqyEbd8aCXgV0jbIYpZsPEr1WZeFYmubOGPDwZrfw3a+e2I0rTpuTOe4sn4mIYkwD9KBwGPCeW9cAAJ7e0x+STVtRJmLlDS/O9Z1GirZj7jSx5k/txKzeIu57OlzXvWrZKCaEQrrPu8ydbKqsuZdrdmrCA7HeGT3RWhV04zyxI4gVf/DZAzjrU3/wq96lpcqHvocQkw8kbytpmy1nbQJAb4e7tn/nLWdjUmewRY2rbf7cATf6wHJ4bh35jd9aGWkZCLgZnGu2HQ4tFP4ORBfYVEbRMxHUBHrelA4cHKyGWCEtgHGx14AbLVOxnMzvdd5Nd+Oiz9+Teo547fJ23krDE9uP4DsPbB7VexB2ew3IydjpWuATEdvSpaGQwNxF+TMtWkZnzM/X2LC7L9e4acGsl7lTc2si5Q4P5q6cjJi08DcTs3pLmNbAmkYy2s64k3baVTBwzuJpWLszPIHSHKoA/Js8CIVk/uuAwFmWhH+74iS85vnH4PRjomFbc7ykiT1Cxt1Nv16H/QMVP4QzyaEqljkFhOidjBhucjAlOUmPndaJrqIRCtsUmXscRlLyNS69W2R7cbJMVgKWiJvvddPTT5zdg5rN8ZiQbGVlLJh5GpHklRBo/p27eJrXxWt0xv3lX7sfH//5yJqEL184NfSY3idI0NNC+Rt0LA2mzmJ3b7mYO3eTm25+01mJ7/+dBzbj54/tDB374I/dXgZ5dHHRH2ZKzN2yOZbMcpP9KHPUP7eJDLpVaDvjvvXgEEyvQH9PKVozJEmWoRW/6ssy7nGxhjXnHMMZDtUzF0zB515zeixLLBgals7qDjnvSA4gA0PGWpcm8sUnhkvV1iTmnjQ5NY2hYGixDsOq5fjfWzS+cg/XyOtGYNwf/fBLseIDF4aOic7SwLgE38OvrJfxeet2BQs4+Q4ODETruyQxdwpZlZ3WIp7eM5D4nAhaRK+//MS6HMJZyLNbkncJNcfB8xZMxhnzJ4cykSlFX2wxWM25YywkhEJWarYvgyQydy/z+9hpnZjUYeKRrYci53z850/FVlQFgmqvaRAdqmLDHY25c2yxl8n9xnOOBRC0v2sFc2822uobPbnjCO5Yvd3fNnaY0fjlpAzVsObO/fIEBeG46wQaXdhUhxnOhiTjThM3MNbhz6DyAos9vVc27kmGCwBKhpbI3Atxxj2DucclsWRB01hEXxSNdtVyoLHwTZZV0RJw/Rd//eX7/L+pEbHIHrMcctSEm+SCOIj9O9Mg9rMVM0BHCzFa6cxP/A4/9foKiNjmJe7Qgj1cdRP2ioYW2sGWvOepaxgQ/KbZzF2LZe4Vy0kMYiDYXuY3YwzPWzA5tY5NHHJFywj3Jt27jDGUTLf9XsVyrwn5sqhESCvr7jcLbWXcn90XZlclU0fZskNb6rQkJnqeJzhUs+Kl86AkGXdqwxYY63AoJEHXGDZ/+nJ80KslXZMcqmk3JXWnkSEudL0h5h4YsimdUTllpJ14ZN+w2IEmriAbGfo03Vo2NIG8Fs0STVoAp+Qw7nnLEwTx4hoMvXGhkPS+h4eqODRUw4d+8kToec453vY9t/fu5ae6Tsqhqo2CoUfKYJQE5i6WuQayjXtSN7FyzcYkz4eTqrl787qjjpIP1BEpqYeCiIIeZe407nLNcXerYnKgdy1kMtUOaKtvJEeWlLz66VWJIaZp7kEopHtcZPR+ptwoJkJHIWxoqZQAGWkyeHHbY8aY3+6O2FGeBSfJuFcs248uEJn7nau34+Z7nwXgXg/ZOZw3YkZm6vLvE2LuMXIZGeO08EuZ1QeJYk7knKTfjYzfQCWZnefNYBWZu9gjYLQgQ0jRX4PSeB7ffgSbvH6wZ3hp+kNVG0VDw8Jp4eieosDcg+5jNO9GxtzLNcefQ0nG3XKCJjilmF11UhSZwznOXBBfekCGOMfEeyhg7k5sLHwb2vb2Mu4yMaNJLOrNWaGQkTh379x9/RUMeTdWmgSSBVEqElld1fKYeEbCzbQu19Du9SIfahnnA27ETJLmTt+vZOr4+CtO9jM8//1X691zbCeUpi1+ZhZkziobe9Ewx8llWS0C5bH8v0tPiK3fYyf4MYLPoR1C4seMiLnLEtxoQO8bl2AGhDOBTUH7Lhha5HvTjqhgaHhm7wD+7adP+LJMWlE8eu+kDNXJ3jz51C/Xxb52qGL5slnJ1CO9FpIW0ErNGVHSj2mEmXvF8pi7EC5Jhr5ZTapbibYy7rK8GZfAU0mpCgkEDlViABRF8vnfPY3XfuNBAKOLiRVv+J2HgybaviyTwcSp9yuVCsiKBAECeUpGxQrfNG9+4UKcMDvY+l7wuRVuwocU+vm9Pz+HLIgNTghyPQ2xtEAtlrmT0U22uiKLLBqa7ywUWWCw44q/RkFLtOTPEdlomoOXkqFKpo5SAwuS+cxdcPpuEmTIUFif932GqhaKQoNoGfTz/O9DWwNZxsh2qMb5QMo1J7ORxqCQALhy0wEcHqrhKcEXkLTLqdhOZlmE2LGGZBl39yo326FospFKjeMZbWXcZUcOJZus3RmExVWteBYgMz45WgYAtnix3kkdffKgJNQ+2SZUrotEvyQwqB5vsSGjkRUJArgTOI4VuVvU8E0jfl+Kbe+QUtJ/8FC2cY9jd1GHapi5y9/Z9GWZFM3dloy7lHQmfk7SNTL8qJzkzxGdfxQvHocdh8uY1lVAydQjEtxocNeanfjW/ZtD73fFV+73H4sLKQUDOJ7zPymjWtwF0PXKytRMKj9QsezM4ltDVQtdXmTSs56EtPXgoPAeCca9Zo+MuYe09aC3qjjH332RW0b4+IR+CBMZbWXcid385trzAQBnLpgMxoA12w4DcJ1OSbJMwQi2skB61bbROFQ7BP17a8i4uzcnRWUkZcGaOgNjQfExK4cjrGgm3JAxN00cQzpn8TQsm9OLq4SWgXGNID732/W+XBRbOVCXZZmwLi7vPkg+SDO6ohHXNc3tds/CMkqwG4q/RkmFpUTsEHZZtOjFYW9fGbO8fpxytNZv1+7Gwut+iecODEZe97M1O/xcBxG08Hzj3k345C+eCl1X8TuKx8VFtGBoeNuLFvl/f+etZ/uPxd3NQS/5KmtXmuxQdTLr0gxWbN+4f/mqMwAEZEX+DiKqlhOSWPJC/C5FQ0el5kQc9xefOAtbbroi0gSkHZB5xRhj8xljKxhj6xhjaxlj7/OOT2WM/Z4x9oz3/xThNdczxjYyxjYwxi5t5hcg2A73t3jHeK3aOgsGTps3CSs2uM11LQplTAmFlJk7EI0YGU3KMMkynHPJuLuf++y+ARQNLVSxTgRjzNcPgeCGyGr9Fyc5uN2lwq+LW/gmd5j41fvOx6WnzA69VsSTO4/gP1c8i2v/z03vjq35LS2YtsPxy8d3YdWWg96iG15YjBxyiWhoDN0Ns+suGhisRGWUpN/NyJEstbc/YOtpbFy8prLm/rM1bvjiBZ+7J/K69922BpcLIZ0EebdBjmJKUPIdot7///43p4ZeU9A1X+cGwk5lcSHf11/xiEP63E5MYvLyP97ywoUAomWp6RjFlZO8GPY7JRh324nNKs9CKFrG1FC2bFRtnruq60RHnm9pAXg/5/wkAOcAuIYxtgzAdQD+yDlfAuCP3t/wnrsKwMkALgPwX4yxphdL/s8VG3Hrw1sBBHG8gNvwYsDTKeOyIAl+CB01sBUm+alStuloEh5Kpttwu2bzUNMNMgJHhmuY2lVIlVnKNQff8DIySS6gUgJxSEqmqVhRLTNu+0tG96ITZuL1yxf4rxVB4yU/QpyUIn+nTfsHcc0PV+PVX3/Qc6jKzD3boSoaBDLePSUzpE1n+THoeJq2318ODFNa4TexJHRJkmUmd8azw6q0UIfGJl0zOmf5Ite47/aynak2z4lzesLGXfo9xWtwmteaDiDjnj2vCzHMnXPu+2/6vJ3nPRvCZT/W7+7DM3sH/IUmrs+A+Ft+8MePhY6PJLdEfI3P3BN8bu2IzG/JOd/FOV/tPe4HsA7APABXAvied9r3ALzSe3wlgNs45xXO+WYAGwEsb/TAZTwqZLuJxlcM90o17n6EATH34CaQU89Hw9x9J2/NhuVwn+V9/R439HComl6YTAbFyYtV+WSkRTjIxjxt4QOA0z2DILPXV/+362ymRSqP5k5p5XS+/Nl+GGqaMRWNu3d+T8kIJR1lRRQFO4TkRWSgYvl1QNKYu/g9ZFmG5tiFJ8wIvSaO5RLkBZG+72RvN/lXn1vhfq4Q7ZJm3EWp4n0vXoJrPM05r3E3vUqdYsbsXV65gEe3HcbbX7QYQHS3ddmX3F1Jt9dURrzmKzbsxQ9Xbg0tmrev2h6qRDoS4y4z94plu7JMhtO4XVDXFWOMLQRwJoCVAGZxzncB7gIAgPLj5wHYJrxsu3esqUiamKau+anVcQ0hCFHNPfmzRiXLFIJaL7bN/djgTfsHsaevjHLNzt3yy3G4b9zjarf449WjpWdp9xBl7snlkIFgcRJvxMGK5Rt1P6Q0TnNPuW5xDlXfyZ0SySDKHvS7dBeN0K6oajnQNZb4+Xk094GyhWnd7gKaVY+fjDj1ZiVyQMZKjqChseaZV2TcxaQzy3b8nVLB0CKauwhxgTN0DX9zptsvoB7mDoR/3y37XXmxr2xhuneN+hJKOXR6mnvQtNrBW7/zF9zwkycii+Zer+hcVmvLJMhRMeUYzb2dkftbMsa6AdwB4FrOeVo5t7gZGrlrGGNXM8ZWMcZW7du3L+Yl9SHJ4eImkoRbxCU16wBEzb15DlXADc98YseRkHzRN1zLLCksYrBqYbhqgbH0Jgtx7er8KoxmvEE9XdiyF0KOqfAiCAD7hTouYqavjDQ9N+6mC6SyOo17KWzcs5qaZ2nunHMcGqpiusfc0zpp1axA0y2ZOhwhiY5+azlyaVAInxSx+0g5YiRpgRALvfWXLb+Zuikxd3mxniz5j+jv/oqVKksRCjGLN+nnN77yFD+dfyDBuAe1XMi4B8999K61oXNpjiUFQWRBnE/rd/dhx+FhbDkwpIy7CMaYCdew38I5v9M7vIcxNsd7fg6AnxAc+QAAIABJREFUvd7x7QDmCy8/BkC4zBsAzvnNnPOzOOdnzZgxQ366biTdvGKho7SGBDJzT/MrjaawPxnu367djad29eHgYBWfvPJkAG7a/3A1m7nTzTRYsVFzOExNSzWcps4iMgkZiSRZ5myhomAWcxdr0cglkvOiavPI71KMYYkyRBZMBkPW3LM02yzmvvNIGfsHqn5RskzNXZBlAKBcdULvH+2BS8Y9PMZ/uX1N5P0HPEexGFNecxzUrIC5z0tpHzddKjE7OaHsRBL8nsLCNaDrMb27iA5Th66xxGxfipaJyy3YvD8cRUS7ntoInaDia6i2zpHhmtLcCcy1Gt8CsI5z/gXhqbsAvNl7/GYAPxOOX8UYKzLGFgFYAuDhxg05HklbWtewebJMCnOnbTsl+4Q19/C5J83pHfE4KSHo079e7x87zXPYHh6qYbjmZDYp/n+XngDAZXyW7WTuJIxY4+4x9wSHajhFW488LzJ30ShQJ/q6jbtlR34X+ty0FP5yiLm7r+8uGmHjnsH8aO4kjfmIt3jRoppq3IUkGbmUML3/ziPDIc2aDLb8W8Rp8f1l1ziJc8SyuS9dmTrD/Kmd/nPElF/zfFd+6ZKatov+qTw1zuJkt6pAFChaKYm503cMGmgkf+hQ1fY/p54491ed6arA4g7mb84MlOFWNOZoBfJcsfMAvAnAxYyxNd6/ywHcBOCljLFnALzU+xuc87UAbgfwFIDfALiGc9641vIJSJJlCrruZ0umae7uuZrgUE3+rLhGHHkRJ7l0FQMjMFy1MmUZaoY9WLFQs3mmVhsny5DUId80tKiJiVqheOEY5v6sEJ9NQ6FyCnkRF+cud8GKw3AMc++VHKpJWcn+mDU3dyDJ0FDWaW/JhKGx1E5aNTtIkusww8ad3r9cc7DtkKtTf/eBzXjzt13uIzP3OP/HLSu3omBooe9j2TzkUBV/U4pO+czfnoanP/XXsWM+7/hpid9Hhl8lVfh9K5LE1100Ih3QCBzhhDIxqEDGcM3O3bxGxGdffRqe/Hg4AvtDV5zkPz5aZJnMNnuc8/sRr6MDwIsTXnMjgBtHMa66Idc/IYgGIi1aBnAnkK+5CwaTR10GI0ac4S4KDTXyaO7EvgYqVq5IAlNnfp14gi/LSDcWfVPxBw+HlEWZ+/ZDw+gq6Fg6u8fXrT/32/WoB3HSSfDbJRvTkCzjLUgdhaCzkqaxxKxkEUZKv1Mq0tVVdIuBpfkARN+BXP5C9K88s2cAx07rwseERhyyMU/akQ1ULEztDuZ71Xb8XYFYzhYIatVrGkMhgQRc/9cn4WVfvT/2ORmmL5VFm3OQ4XejldKZO303x8s9icNw1U7dbSfB0DV0S+dP7jDBmLs7GUlC1ERE23zLpNWfUucHq1bmRCkYuqC5J8syo0Gcni5G0OTR3OmGHay4oV15UsYtm2PbwSH8eeN+AMmyDH1ZUZbK0tyHazY6CgZMLQi5XL3VzQou6Bp6itmtetMcqmmae0iW8QyGPMY8cdKGpiUyd5JHuooGil5LvuTvwUOLDAAM19zX246Dmd6ub09/tISBPIfT8inmTe7ATa86FYCrW9Pclq9hZzHbOZ/W8F1G4DAXmbsNXWP+eGVZRgwlvvKMuQDCzD3p/qpa4qI1ulQZQ9f8iLKjRXNvmwbZxLr+8C8XhI7P8ZxLuw6XfbaRdKO7NZ+joZBk55bN6Q0V1hoJ4lg5HXts2xGUa06mcScZZ7BiwbKzm0cbXmzy+Z91Y6K33HQFvnr3MwCisgz1WhVjgYsh4x5l7pWajZLpRmnI8c3festZOH9JtsO8r1yL/C6UjDZczelQ9X40cXfRUdBzRVukdU3yjXvBCM2ROLhJTJ7m7jN3r26QwzGrt4S9/RXs74/Wjpd9LVmt36jBs2Vz1GwHjEV9T2JUTRLyht4C8EtOy2W0Q00ypFLHdG9+4JKlkdZ3tsMTWxFWbXHRGr1OTgvOeGpa3ky0jXF3vM7qx88MFwCiv1duPuA7xBI1d0PzQ89E5nrTq07D1+7eiE/9zSmj1uvinKV07I7V2wEk15UhTOsqgjE3Nr7m8OwyrdINzznHb9fuARBNyNnX74Y1zhCiKsJND6LMvWy5qeeGzjBcc2+cU+dNwhM7juC848K9KpNQi4mWMXQNXQU9tQuSGApJkSAyc88T26zrLDEUkBaQzoLu1QVPXmws4bNkzd2yOTpMHZM7zVD4KGHj3gE3RV+KKEmCGJlU8T5XjpqiOZ+GrIJfIsjJLTqfK9LOSJckrrgoNTH8NMnU1mwnNcKtXtCY5KY+7Yq22Z9YXmd1GcfN6Mas3iI27O4XpIhkzf0xr8iYuEjMn9qJz7z6tIY4YuLkI/kmzirANKnTxLFTO7Fp3wBqVp5omfD7VSzHTzZ5kdQo+C0vXIjJnSYuEnq2FuKYu8CYyzUHHaYeqWFz0QkzQr6LLHSaUa7R22H6Ke1xID372pcs8evxyLuLpNaKItI09wFRljG05KYSDofDA3moo+B+JjlkLcdNpprWVYg17nv7K3iv0D80Sxumz6nZHDWLh+qvkNSSFXnljjO/cZcbxgPReuvyPKC8AFEC1AXNPUkOq4myTAPuvb8/91h/fEcD2uZb2g5PZDpzJnVg15Gy7+TpLsZvVUUjllS4a7RIutmmCVXp8kzkjoKBcs2B5eTR3MPXpb9soaBrePXzjwkVlQKA0+dPxpqPXBKKhy6EjIan9wuNFoargiwjVLfsySEJiOiK0Yd7SkZqr83hmo1T503CtS9Z6h/zHdRC8loW89O15JZ4Q1ULusb8ksJJzL0mtTykcZAh3HWkDE1zdxhxxh0AHt58MBiTR1ae+sSl+ISXCyGCPseyHVRtO7QY3P3+C/HIv70k+QsLqCfMMM4PInb0AqISVxypylPyoWbzzCCIekBN5hvV+nC8o62Me1JI4NSuAo4M17BivZtnNa07PrJGNHTNcroksYZ3X3S8/zhPp6eSXysjGkIoQ74u/eUa+iuW75jNgiltuTtM3dehV205iAc3HUDJ1GEKZQ76ylZs84Yf/INbZijOyRrHIDsLRqRjj4jDQ9WYHqHueCmqRZYN4mBo8U0oAK9UbUEHYwxTugpYt6vPZ+MiyKDR9RYN4UDFwvZDw3hg4wFM6jATF6xu4bvYDsfxM7vRWTBw8txoboXfQNzhkSYUsyeV/Fo4WSAp58Qc/iSxYxmhaoeZu1zuIi4bmmRPuXT0wmlBjL4YBdQItm3EFCtrZ7SNcbe8sLc4UOnVP3rGPWmiiCn3cohgM/B/V5+D+z54kTemYOx5moGUQp1lsrbv4ef7y5an7ebbjssLXVdR98MD3/DNle6xgoGC7i44nHP0l2uh+ieE85fMwJabrsAbXrAg8lxc1IapJzs6AZcNz5kU3mWVhNBSIF/6usvc4xl533CwC7nq7PnY01cJdRAi+Mad2tgJhlBcDKifZ9L3oUxNNxLKnRfkA5zZU8TX3/h8AAEZOTBQQdXKTn5Lwx/+5QLc/s5zM88jIx7S3GtORE+34zR3PczuAUQ6hH3+Naf7j2t2dvhyPchTQ6id0D4OVZ7M3MVmvMtSskvF4ltjES510txeP5pBZOt5CkiVTA37BmrQNY7ejPZmMrO/Y/V2ODxZnvLHN6cX63b1xRh3w2fupsZQhVtyuGTq6Bu2vPh7nitS44RZPdiwpx8A0BGjuSc1ZAZcx/C+/gpm9oYZKi3MFPkjd7yPQ5rmvlf4DJKr4trnBbKMe739mHAhxwII97TtMHWce9w0DFYsrPQkmfuf2YdF07vc3aj3Xs8/dgr+8w3Pw0uXzfIN3dJZPSiZGtbu7Iu0TKwXciBCEmTm/ujWQ/jj+r2hXaChMcnhGi11QYljYt19ADhxTi8mdZg4MlxDzXaw2Wts0kjjfrREy7QVc0+SMzoKmt9S7DKh4YQMUZYZzY2SF2JUjGjQ88gyHV67vuGqlRmnLGvy33/wOQBB+dUk3PL2F+D/rj4nsiPqLATNMIjB95RMV24o1/DEdret4bIYKUHGwulCqnzMeJLKFQPwfA7RRaQoMfc8iV6Gnqy57+0v+/HpJB0RE991ZBgHB92wxkCWCTN3Merjy1ed4fbzFMrZnji7Bzf//VnBB7JAbqFdHGMMV5w2JxKVMrmjgP5yLbY2fzNQkJj7bQ+7BWDFQm3ytZQzWP3zNBZ6HeDGyD/20UvQYeqo2Rwf+smT7uc2gGwdN8Ntu0k9CdodbcPcXY99/OTuLATp0HJVPBGiccnqSNMIhJsahzXLLJSEhCvZKSoj6f3kOiMypnYV8ILF0dT0glCrhljWyXN7cXCwCs5daQEIO4mTICanxMsyyVp4v1ecqjuiuQehkO+59VHs6csuZ6trWiJz7y9b/gLiO5S9xe3cT9+NMxdMxk/efV5QN56YuxBZIibQlUw3y5VzDstxi2LFlWy2HScz1p2yQSuW3RB2mwWZue+NScbSJf8FhdfKspHGWKh+jlyqWNztLJ01uvwSAJjWXcSWm64Y9ftMFLQNcy97iTRxECdVWgx5R4aRbCYMvT7mXvR6sebJaE0ybHkdqjLEbjyXnezuhP7mzHm+gdrnRYLkqg8unBMvy0SLnhGo2YfsnBWTmH7uNZLIk8SUxNwHypa/gNAC9P4fPeYvro962bi0OJBRZ4yhYGioCNpx0XSLfrkt3+L15A//9Els2jeABzYeyJwLZNzzlFhoBAInsVfhMkaeMoWcAc453n3LagDRxVtk7lecNge/FxIQZ/QUsePwMBZO68QrTp87JgtXu6FtrthwSpML0aCnsdw8afLNQr2au9to28FQ1UZnhiMt6f1GatxFqaRmOzhmSgcYY4Fx95ha2g1JZlS84ZOZe7xxp1Zu0WgZj7kLCU5zvGqVSZATb/xxco6BahBZJF4zuX5K0KtVSPryfAa+IdfdRCjOgSGP/dPvIzr0L/6PPwGI1/ZFdJdM9FesUWvueeE3UPFkJbk2PYBQSOwhoRx0l3Tv6VrA3F915jwsmt7lP3f8jG5s3j84Zt+rHdE2V22oasUmwQBBfRkgvY7GcTNcp5I4ycYKokHIFS1jumnwNdvJjOwRGbRY0TJLlkl7P7G7FbFvMu57cxh3gth1vl5ZhjBPysKkHZzYpeqqDJ3VSIiWGara4Dy4Vl1FA7N6izh2Wqe/KyDEVTAkeYHCMsWKjsRa6fe59epzIp8/nFLqAAA6TR3lqp0qSzYScjOW4ZhyHdSKDwh6vAIxzF3X/BpEsmQzrbuAQ4NVr+F4879XO6JtNPfhmpPYak5k9GkTZf7UDnzs5ctwycnJTtdG4L4PXhRJ+zfqjpbR/RsoazEQJZ/3Xnw8Pvwzt+PNaIw7lZgVnZWTPH/Ght1umGAeJ5h4w8eNJ02WISycFl6Mycgd8Byd779kaeYuRWSbIqjJwxTBV3Ph0pm45+m9+PWTu0LnxvVqJeMu1jWi38NvTehdv7hdZRwzFtHhtfLT2NiE72oa80JendD4RB+VWGNIXJw6pd+AHNFANCt7SmfB//0Ucx8Z2se4Vy3M7o1P2mAJFQ7jznvLeYsaPjYZYjMFgl6n5i7eDFlON3HRE9ldXJJRHhQMJsgyQZccatL99J4B77wUWSYmHK0eWYZe/96Lj48W3NIZNAbc8tBWANm1egB3ASzHlPL9+2+7cfxixvLkThOHhmp47VnT8Jcth/xxk0ETk77IP0ENObqLun+9qGZO2vjSipQBXphvzYbhGd2xQNV28N/3PItDg1V/9yHOwJLhRrqIceoAUuXDtHaAY7FotSPa5qodHqr5xkUGaaOLp3fheQsmj+WwcsMUo2Vyau4EPSO6ZsnMINJANLgjdqgKBlfMjJR3TmnG5pVeZ5yXLpvlH4uLUKJyxTL8Dj0xBoMxhpKp++e8+KRZkXNkJEXL0K7gBYuCqKFJnSaqloNDQy6zLAvt4IDwYlvQ3TaPFIrbWzL960VF6tKMe5Ys0+HlcFStbHmu0bjtL9t86Uv86cgw7z5Sxo9WuaGSX3/j81LrDMkLdFofWIV8aAvmzjnHwcFqqIGBCDJorzt7/piEOI4E4mTOx9zjY+TjIMpSRUPDmQsm49Gth0enuXuMbKhq+XKCvLVOY+4nz52UKyzN1BmqtoNHnjuIWb0lHDPF3fWUEzpJEYqGhqGqjaWzuv3SuKmfk5KheomQOAS4kgEQ6MkOd3cwcoYqEMgyRDB6SqYv29AxKjAGuI7fXYJOnZVN2enJMpzzlhpB8b4i404lpgH4v1sS0nxhWYX0FOLRFsb98FANlsMT46qvOns+HIfHpryPF4ilWfNkR4sGu56G3QVDw21XnzOqFGzTCByqQ1Xbr2EiL5x5FikA+OeXLMWm/fFlWEmW+dv/fhAA/AUh6CQVbxTcxa+WS5Khsco7hP996Dms390fqeFPMe+7+wIj/I4frMJS7zzRx2F6+nRfuQZDYyiZgeZO9cXFkrs/+IfleMkX7vX/zsqm7Ci4bSQHq3ZLtelvv/ls/3Gc7yvLuS6XqhBnjmLuI0NbGPeNXn1minaRYeoa3vzChWM4ovohGvcjObrQi0wnT9LTyXN7sXZnH2xn9Ayvt2Sib7gGy3ZDMbuEsSye3oVN+wfxoxx1Sgjve8mSxOdMXYtd7Cj6pJRgNGhnkzfSIi5D9d9+6mZHyhUc6dofGgx+pxUb9mGFF5ppxjhUa15IH2OBNk6aeynFqZxWNREIf79WGcFPv+rUUOnoyTEtL7P8AV0pzF05VEeGtrhqBwZc7XNWb3os83iGyHpffNLMlDNdiAk/eZj7Qi+8s2ylR57kwXEzulC1HWw/NOzKMoJB+sHbX4BPvfIUnL1w6qg/B0heuG73tNwk4+03qc5Zq1yXil05wmNRJhHf88BgfNleccxFw9XcxfIYJNvQIh4uQxH+LbPKoIivHatEn2+KpRIQZeqTR8Dc5V2fGHSgHKojQ1tcNduJxhdPZOTJ7Awx9xzyx2xv4YtraFIvaAs9WLUiSVTzJnfgjeccO+rPICQxvq/evTE0Fhkk1+SVZeTCYYeF3ZPcqYjeMy66BpDi3D3/hFsEzP0uZPy3HRoGECYl9c7hTsmfMhaQycdZC6eE/q5Xlvknodx18BmBE1zJMiNDW8gyFIKWV+NtB9Qry3zgkhMwZ1IptXBaXvgp6JZbEKuZzCrL2CWFc5JcU4/mLjL3A54Us3zRVHz+1aeHzq2nUFupoGOoaoeYO8k2Ww8MosPUQzH0aU2x4yDuXKb35KvfPlrILHtmT3jHHFe/Kc7gUyvGD1x6QurnKYfqyNAWV832k3kmtnF/5Rlz8frl83OdK8oNeatIvv38xQ25RkHtFo+R1uHQrReysdt+aCj0t5awE/E195yyjFymlurjXPviJVgwLRzpESf1zBPi4MXF9pjJHdhxaBhVy/F3TbRgbTkwhLmTSyFjmWcXljSWWWNk3EXccPmJkWOMMfzDi4J8kZvf9PzY3egd73oh/vKh7G5RirmPDBOeuT+x/YivuU904/6lq87MfW6aTtts0M3m13RvohwmJ/Hc8JMn8f23LUd30cBAxcIpCWWFaQHKqrtDkJn73evcxi5yrXgguhv4wmtPx0uXzcKpH/sdgPDvMWdSCVXbwYHBSqzmfppQTwYIS3JzJ5XwrguPSx33LGF8k1IqnjYaP3n3C7H90DBefvrc2OfpVnzPxccnZnwXDC1UDiMJWTslhXhMaONesx28/Gv3+3+PtZFrJcRU9TyyTCNBhpPa39UrJdQDOXLo3qf3YePeAUzvLuDiE2cmfjYx97wOVbEeymPbDuOb928GAMzoiTrp5TIBr3reMQCAX733fHzzvk2x9Xv6hmv+7ySWlpZlJZGgPHDdxZl5GYunBxFiSbWVmoEzF0zBmQumJD5PGbnTc7b6S4NcO0ghHya0Ndx1OBzFMNGZez0Qdch6t/KjhW/cpaqGzcDhoWhY6H/8bgMGq3Zqm0AaUt5QSJG5i9ExcV2uxGv/3osDZ+Cyub34wuvOCM1D37iXLf+42Dg8rswxIU/CneiozLuQjQUo2mg0Tt5PvfIULJ7RheldYy83tQMmtHE/PFwN/T3WRq6VEG/8RjQPrgcF37iHqxo2Ay8RyhO89iyXIbudoKxICVkRJLDkjnMXil2JiDOw4rELTkgPWyVJoW+45s9PsexDIyWHcWXcvRjO0QRnvfGcY3H3+y9MLVugkIwJbdxPO2ZyyDN/tE6CsW5kQGz0K398BkBzZaELls7wH3/iylMAuD1gh6p2avmE1y9fgNm9pdSeuSJE5i42s85CVjROwNxrfvXOZhn3pISuVoDcF+O13MfRgPEzG0aIekvltiPGmrmTjtrvN8lu7uefOs91OpZMHfOnBvprWuGzcxZPw0M3vBjnHhdtExiHiuWgZnMMViy/u9BPrzkv83VxbF8ElSoo1xx/fuoaqzvJKg+a6fuoF9zbOyVFMyk0H5mzgTH2bcbYXsbYk8KxMxhjDzHG1jDGVjHGlgvPXc8Y28gY28AYu7RZAyfUW3CrHdHq9OxmO3Rvu/ocPHzDiwEAXxYiihppGL/lOVBv+8s2v1b9wmnJxa4++vJl6CrooYqbcZgtJCjFzc+RVuYU8bNrzsOHX7Zs1O/TSFBm7VF6S44L5LEK3wVwmXTsswA+zjk/A8BHvL/BGFsG4CoAJ3uv+S/GWFOFQLHUt2LuYwcxyiOr/slo0VU0MNMzks3KyHzHXy0G4DpLk3qbinjreYuw9hOXZS4wvR1imYhgflKziykxdVjqxenzJ4fiyscDGqG5K4wOmXcH5/xeAAflwwBIzJwEgPqNXQngNs55hXO+GcBGAMvRRIhFnY5W5t6K5sH/d3VQGIx6po4FmlVL5T0vdouXDZQtn7k3YtFkjPk16+PmpzmOdPJGwvGZ+9F5T44HjHRmXQvgc4yxbQA+D+B67/g8ANuE87Z7xyJgjF3tSTqr9u3bN8JhRN6zIe8z0dCKmjrL5vbiE1eeDAChbjvNhlybvlHoKuiY3Gli8/5BVG0HjDVuJ0jSjChf0XufODsq65w+f3w2lKkH3GfuR+c9OR4wUsHvXQD+mXN+B2PstQC+BeAlCJdhJsTu2TnnNwO4GQDOOuus5u7r2xytYO4AcNXZC3BgoIqrPUljLCCGPzZSjmKMYcnMbmzaP4hJnSZMXWuYYaKIGTH5aeO/Xw7H4bERXj9+57mjqrc/HvCGFyzALx7fhbOOTU50UmguRnp3vBnAnd7jHyGQXrYDEIujHINAslFoMKZ7nadatfUtGBr++aVLR9zRaSQQP6vRi1pvycRgxXJb1jVw4aCMVarfTkgK3TV1LXd8/njFC4+bji03XRHqPaswthjpDN4J4ALv8cUAnvEe3wXgKsZYkTG2CMASAA+PbogKSbj1H8/B1X+1OLEDVbuj0Q2hC4aGtTv78PPHdjVUC7/8VLe2SlrSlYJCo5E52xhjtwK4EMB0xth2AB8F8I8AvswYMwCUAVwNAJzztYyx2wE8BcACcA3nPL3Dr8KIsWRWD264/KRWD2PMMaXTxKGhWsOdkQNe3P7+gUoohHG0mDOpA99/23KcnFDkTEGhGcg07pzz1yc89fyE828EcONoBqWgkIY5kzpwKKbmzGgh9is1jcZKXX8lZNoqKIwF2jMOS6GtccnJbmhh3kYcefGy04LytY1+bwWFsYYy7goTDu+9eAl+9d7zcVLOujF58frlCzBnkivHdCh9XGGCY8Ib94++fHylXSs0H5rGsKxJ+vXlp84B0PqSDgoKo8WEn8FvPW98pV0rTGxQPZn+cv7KkAoK4xET3rgrKDQSC6Z1AQD29ZczzlRQGN9Qxl1BQcDi6a5xP35md8aZCgrjG8prpKAgYP7UTvzw/7d3fyFWlHEYx78PmqaWqFlhrqSSBF6lSalFREaZiduloGSU9/25KGWvurQiIqIitOiPKWFWIgRGBd2UZf9sS801S7csN6KSglT6dTGvOrsct9atnXmn5wPDmfmdM8vz7p7z2zPvzNldeeV/NqdvNlTc3M36mH/JxKojmA2ap2XMzBrIzd3MrIHc3M3MGsjN3cysgdzczcwaqBFXyzy+bPbf/qNiM7P/k0Y09xN/D8TMzAqeljEzayA3dzOzBnJzNzNrIDd3M7MGcnM3M2sgN3czswZyczczayA3dzOzBlJEVJ0BST3AN4P4EhOBH/+lOEPN2auTc/6cs0Pe+euU/eKIOL/VHbVo7oMlaUdEzKk6x5lw9urknD/n7JB3/lyye1rGzKyB3NzNzBqoKc39qaoDDIKzVyfn/Dlnh7zzZ5G9EXPuZmbWW1PeuZuZWUnWzV3SQkl7JHVJWlV1nr4kTZH0tqRdkj6XdGeqT5D0hqS96XZ8aZ/VaTx7JN1YXfqTeYZJ+ljS1rSdU/ZxkjZJ2p1+BvNyyS/p7vSc6ZS0QdLZdc4u6WlJhyV1lmoDzivpckmfpfselaSKsj+Ynjc7Jb0iaVwds/crIrJcgGHAPmA6MAL4FJhZda4+GScBs9P6ucCXwEzgAWBVqq8C1qT1mWkcI4FpaXzDKh7DPcCLwNa0nVP2Z4GVaX0EMC6H/MBkYD8wKm2/BNxW5+zANcBsoLNUG3Be4H1gHiDgdeCmirLfAAxP62vqmr2/Jed37lcAXRHxVUQcBTYC7RVn6iUiDkXER2n9CLCL4oXbTtF4SLe3pPV2YGNE/BER+4EuinFWQlIbcDOwtlTOJftYihftOoCIOBoRP5NJfor/kjZK0nBgNPAdNc4eEe8AP/UpDyivpEnA2Ih4N4pu+VxpnyHNHhHbIuJ42nwPaKtj9v7k3NwnAwdL292pVkuSpgKzgO3AhRFxCIpfAMAF6WF1G9MjwL3An6VaLtmnAz3AM2laaa2kMWSQPyK+BR4CDgCHgF8iYhsZZO9joHknp/W+9ardTvFOHDLKnnNzbzWfVctLfySdA7wM3BURv/YkX0WsAAAB40lEQVT30Ba1SsYkaTFwOCI+/Ke7tKhV+fMYTnGo/UREzAJ+o5gaOJ3a5E9z0+0Uh/0XAWMkLe9vlxa1Wr4WktPlrd04JHUAx4H1J0otHlbL7Dk3925gSmm7jeLQtVYknUXR2NdHxOZU/iEdxpFuD6d6ncZ0FbBE0tcUU17XSXqBPLJDkac7Iran7U0UzT6H/NcD+yOiJyKOAZuB+eSRvWygebs5Nf1RrldC0gpgMbAsTbVAJtkh7+b+ATBD0jRJI4ClwJaKM/WSzpavA3ZFxMOlu7YAK9L6CuC1Un2ppJGSpgEzKE7SDLmIWB0RbRExleJ7+1ZELCeD7AAR8T1wUNKlqbQA+II88h8A5koanZ5DCyjO1+SQvWxAedPUzRFJc9O4by3tM6QkLQTuA5ZExO+lu2qf/aQqz+YOdgEWUVyBsg/oqDpPi3xXUxya7QQ+Scsi4DzgTWBvup1Q2qcjjWcPFZ9tL2W6llNXy2STHbgM2JG+/68C43PJD9wP7AY6gecprs6obXZgA8X5gWMU72LvOJO8wJw05n3AY6QPWlaQvYtibv3E6/bJOmbvb/EnVM3MGijnaRkzMzsNN3czswZyczczayA3dzOzBnJzNzNrIDd3M7MGcnM3M2sgN3czswb6CxrZ+gNzzOdiAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(train_predict[:,0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([183.26593, 184.21674, 175.80016, ..., 223.11134, 218.05138,\n",
       "       213.77455], dtype=float32)"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_predict[:,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1a358913188>]"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(Y_test_Sca[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1a361a3f248>]"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(test_predict[:,0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8,4))\n",
    "plt.plot(history.history['loss'], label='Train Loss')\n",
    "plt.plot(history.history['val_loss'], label='Test Loss')\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epochs')\n",
    "plt.legend(loc='upper right')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "aa=[x for x in range(347)]\n",
    "plt.figure(figsize=(8,4))\n",
    "plt.plot(aa, Y_test_Sca[0][:347], marker='.', label=\"actual\")\n",
    "plt.plot(aa, test_predict[:,0][:347], 'r', label=\"prediction\")\n",
    "# plt.tick_params(left=False, labelleft=True) #remove ticks\n",
    "plt.tight_layout()\n",
    "sns.despine(top=True)\n",
    "plt.subplots_adjust(left=0.07)\n",
    "plt.ylabel('Global Sale', size=15)\n",
    "plt.xlabel('Time step', size=15)\n",
    "plt.legend(fontsize=15)\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}