MLProjects/kaggle_time_series_air_passenger/model.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",
    "from matplotlib.pylab import rcParams\n",
    "rcParams['figure.figsize'] = 15,6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>Month</th>\n",
       "      <th>#Passengers</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1949-01</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1949-02</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1949-03</td>\n",
       "      <td>132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1949-04</td>\n",
       "      <td>129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1949-05</td>\n",
       "      <td>121</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Month  #Passengers\n",
       "0  1949-01          112\n",
       "1  1949-02          118\n",
       "2  1949-03          132\n",
       "3  1949-04          129\n",
       "4  1949-05          121"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Grab and look at our data\n",
    "DATA_LOCATION = 'AirPassengers.csv'\n",
    "data = pd.read_csv(DATA_LOCATION)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>#Passengers</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Month</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1949-01-01</th>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-02-01</th>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-03-01</th>\n",
       "      <td>132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-04-01</th>\n",
       "      <td>129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-05-01</th>\n",
       "      <td>121</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            #Passengers\n",
       "Month                  \n",
       "1949-01-01          112\n",
       "1949-02-01          118\n",
       "1949-03-01          132\n",
       "1949-04-01          129\n",
       "1949-05-01          121"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert to DateTime\n",
    "data[\"Month\"] = pd.to_datetime(data.Month)\n",
    "data.set_index('Month', inplace=True)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Time series in numpy format\n",
    "np_ts = np.array(data).reshape(144)\n",
    "\n",
    "# Time series in pandas format\n",
    "pd_ts = data.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\TSB\\Miniconda3\\envs\\myenv\\lib\\site-packages\\pandas\\plotting\\_matplotlib\\converter.py:103: FutureWarning: Using an implicitly registered datetime converter for a matplotlib plotting method. The converter was registered by pandas on import. Future versions of pandas will require you to explicitly register matplotlib converters.\n",
      "\n",
      "To register the converters:\n",
      "\t>>> from pandas.plotting import register_matplotlib_converters\n",
      "\t>>> register_matplotlib_converters()\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1da38dd2888>]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(pd_ts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split up our dataset\n",
    "split_time = 100\n",
    "x_train = np_ts[:split_time]\n",
    "x_valid = np_ts[split_time:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Naive Forecasting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1da3907fd48>]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Naive forecasting\n",
    "naive_forecast = np_ts[split_time - 1:-1]\n",
    "plt.plot(naive_forecast)\n",
    "plt.plot(x_valid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "41.72727272727273"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calculate our Mean Average Error as a baseline\n",
    "errors = naive_forecast - x_valid\n",
    "abs_errors = np.abs(errors)\n",
    "mae = np.mean(abs_errors)\n",
    "mae"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dense Model Forecasting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Utils\n",
    "\n",
    "def window_dataset(series, window_size, batch_size=32,\n",
    "                   shuffle_buffer=1000):\n",
    "    dataset = tf.data.Dataset.from_tensor_slices(series)\n",
    "    dataset = dataset.window(window_size + 1, shift=1, drop_remainder=True)\n",
    "    dataset = dataset.flat_map(lambda window: window.batch(window_size + 1))\n",
    "    dataset = dataset.shuffle(shuffle_buffer)\n",
    "    dataset = dataset.map(lambda window: (window[:-1], window[-1]))\n",
    "    dataset = dataset.batch(batch_size).prefetch(1)\n",
    "    return dataset\n",
    "\n",
    "def model_forecast(model, series, window_size):\n",
    "    ds = tf.data.Dataset.from_tensor_slices(series)\n",
    "    ds = ds.window(window_size, shift=1, drop_remainder=True)\n",
    "    ds = ds.flat_map(lambda w: w.batch(window_size))\n",
    "    ds = ds.batch(32).prefetch(1)\n",
    "    forecast = model.predict(ds)\n",
    "    return forecast"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/500\n",
      "3/3 [==============================] - 0s 82ms/step - loss: 271.8661 - mae: 272.3661 - val_loss: 505.8083 - val_mae: 506.3083\n",
      "Epoch 2/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 264.9019 - mae: 265.4019 - val_loss: 484.3098 - val_mae: 484.8098\n",
      "Epoch 3/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 252.7449 - mae: 253.2449 - val_loss: 459.5409 - val_mae: 460.0409\n",
      "Epoch 4/500\n",
      "3/3 [==============================] - 0s 15ms/step - loss: 240.2713 - mae: 240.7713 - val_loss: 445.2593 - val_mae: 445.7593\n",
      "Epoch 5/500\n",
      "3/3 [==============================] - 0s 15ms/step - loss: 234.7380 - mae: 235.2380 - val_loss: 439.4569 - val_mae: 439.9569\n",
      "Epoch 6/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 232.3914 - mae: 232.8914 - val_loss: 434.1690 - val_mae: 434.6690\n",
      "Epoch 7/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 229.5039 - mae: 230.0039 - val_loss: 428.0587 - val_mae: 428.5587\n",
      "Epoch 8/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 226.1541 - mae: 226.6541 - val_loss: 420.4857 - val_mae: 420.9857\n",
      "Epoch 9/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 221.9135 - mae: 222.4135 - val_loss: 410.7528 - val_mae: 411.2528\n",
      "Epoch 10/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 216.5273 - mae: 217.0273 - val_loss: 398.2069 - val_mae: 398.7069\n",
      "Epoch 11/500\n",
      "3/3 [==============================] - 0s 15ms/step - loss: 209.5912 - mae: 210.0912 - val_loss: 382.1219 - val_mae: 382.6219\n",
      "Epoch 12/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 200.5868 - mae: 201.0868 - val_loss: 361.0789 - val_mae: 361.5789\n",
      "Epoch 13/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 188.9043 - mae: 189.4043 - val_loss: 332.4605 - val_mae: 332.9605\n",
      "Epoch 14/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 172.7329 - mae: 173.2329 - val_loss: 293.5743 - val_mae: 294.0743\n",
      "Epoch 15/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 150.6249 - mae: 151.1249 - val_loss: 239.6270 - val_mae: 240.1270\n",
      "Epoch 16/500\n",
      "3/3 [==============================] - 0s 18ms/step - loss: 119.5182 - mae: 120.0182 - val_loss: 163.1781 - val_mae: 163.6781\n",
      "Epoch 17/500\n",
      "3/3 [==============================] - 0s 15ms/step - loss: 75.0523 - mae: 75.5523 - val_loss: 62.0292 - val_mae: 62.5292\n",
      "Epoch 18/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 29.5400 - mae: 30.0355 - val_loss: 87.7864 - val_mae: 88.2864\n",
      "Epoch 19/500\n",
      "3/3 [==============================] - 0s 19ms/step - loss: 48.2491 - mae: 48.7487 - val_loss: 125.9318 - val_mae: 126.4318\n",
      "Epoch 20/500\n",
      "3/3 [==============================] - 0s 15ms/step - loss: 57.6143 - mae: 58.1143 - val_loss: 90.4777 - val_mae: 90.9777\n",
      "Epoch 21/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 33.2537 - mae: 33.7454 - val_loss: 51.3877 - val_mae: 51.8832\n",
      "Epoch 22/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 26.7287 - mae: 27.2281 - val_loss: 63.7683 - val_mae: 64.2683\n",
      "Epoch 23/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 35.8266 - mae: 36.3216 - val_loss: 58.0335 - val_mae: 58.5335\n",
      "Epoch 24/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 27.6385 - mae: 28.1216 - val_loss: 50.9082 - val_mae: 51.3875\n",
      "Epoch 25/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 24.6842 - mae: 25.1842 - val_loss: 62.9936 - val_mae: 63.4859\n",
      "Epoch 26/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 28.3220 - mae: 28.8167 - val_loss: 61.9798 - val_mae: 62.4798\n",
      "Epoch 27/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 26.2102 - mae: 26.7068 - val_loss: 50.7160 - val_mae: 51.2160\n",
      "Epoch 28/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 22.5669 - mae: 23.0600 - val_loss: 49.6758 - val_mae: 50.1758\n",
      "Epoch 29/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 24.4321 - mae: 24.9220 - val_loss: 49.2939 - val_mae: 49.7939\n",
      "Epoch 30/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 23.1372 - mae: 23.6315 - val_loss: 49.9636 - val_mae: 50.4636\n",
      "Epoch 31/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 22.6659 - mae: 23.1559 - val_loss: 52.5233 - val_mae: 53.0233\n",
      "Epoch 32/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 22.9460 - mae: 23.4310 - val_loss: 49.7654 - val_mae: 50.2647\n",
      "Epoch 33/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 22.3227 - mae: 22.8192 - val_loss: 48.2324 - val_mae: 48.7324\n",
      "Epoch 34/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 22.2731 - mae: 22.7657 - val_loss: 47.9464 - val_mae: 48.4464\n",
      "Epoch 35/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 22.3831 - mae: 22.8741 - val_loss: 47.6719 - val_mae: 48.1719\n",
      "Epoch 36/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 22.0527 - mae: 22.5487 - val_loss: 47.6310 - val_mae: 48.1310\n",
      "Epoch 37/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 22.3382 - mae: 22.8329 - val_loss: 48.8708 - val_mae: 49.3708\n",
      "Epoch 38/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 21.9270 - mae: 22.4125 - val_loss: 47.1424 - val_mae: 47.6424\n",
      "Epoch 39/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 21.8263 - mae: 22.3203 - val_loss: 46.5738 - val_mae: 47.0738\n",
      "Epoch 40/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 21.9597 - mae: 22.4534 - val_loss: 46.2974 - val_mae: 46.7974\n",
      "Epoch 41/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 21.6622 - mae: 22.1520 - val_loss: 46.2643 - val_mae: 46.7643\n",
      "Epoch 42/500\n",
      "3/3 [==============================] - 0s 15ms/step - loss: 22.0928 - mae: 22.5886 - val_loss: 46.5074 - val_mae: 46.9889\n",
      "Epoch 43/500\n",
      "3/3 [==============================] - 0s 16ms/step - loss: 21.3585 - mae: 21.8580 - val_loss: 45.4746 - val_mae: 45.9721\n",
      "Epoch 44/500\n",
      "3/3 [==============================] - 0s 15ms/step - loss: 21.4842 - mae: 21.9802 - val_loss: 45.2174 - val_mae: 45.7174\n",
      "Epoch 45/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 21.3711 - mae: 21.8637 - val_loss: 45.2161 - val_mae: 45.7161\n",
      "Epoch 46/500\n",
      "3/3 [==============================] - 0s 23ms/step - loss: 21.2700 - mae: 21.7659 - val_loss: 46.5541 - val_mae: 47.0541\n",
      "Epoch 47/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 21.3453 - mae: 21.8391 - val_loss: 45.4600 - val_mae: 45.9600\n",
      "Epoch 48/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 21.1439 - mae: 21.6319 - val_loss: 45.3669 - val_mae: 45.8612\n",
      "Epoch 49/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.9301 - mae: 21.4204 - val_loss: 44.5570 - val_mae: 45.0570\n",
      "Epoch 50/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.9137 - mae: 21.4112 - val_loss: 43.6219 - val_mae: 44.1196\n",
      "Epoch 51/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 20.9056 - mae: 21.4037 - val_loss: 43.5163 - val_mae: 44.0163\n",
      "Epoch 52/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.7782 - mae: 21.2708 - val_loss: 43.8434 - val_mae: 44.3434\n",
      "Epoch 53/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.5193 - mae: 21.0175 - val_loss: 43.6874 - val_mae: 44.1874\n",
      "Epoch 54/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.4962 - mae: 20.9962 - val_loss: 43.1953 - val_mae: 43.6953\n",
      "Epoch 55/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.3569 - mae: 20.8524 - val_loss: 43.0371 - val_mae: 43.5371\n",
      "Epoch 56/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.3203 - mae: 20.8173 - val_loss: 43.1354 - val_mae: 43.6354\n",
      "Epoch 57/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.4946 - mae: 20.9881 - val_loss: 44.3989 - val_mae: 44.8989\n",
      "Epoch 58/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.3857 - mae: 20.8857 - val_loss: 42.9665 - val_mae: 43.4665\n",
      "Epoch 59/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.2082 - mae: 20.7028 - val_loss: 41.3625 - val_mae: 41.8612\n",
      "Epoch 60/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.1148 - mae: 20.6113 - val_loss: 41.5026 - val_mae: 42.0026\n",
      "Epoch 61/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 20.0211 - mae: 20.5187 - val_loss: 43.2069 - val_mae: 43.7047\n",
      "Epoch 62/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 20.0685 - mae: 20.5667 - val_loss: 42.7780 - val_mae: 43.2549\n",
      "Epoch 63/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 19.7746 - mae: 20.2746 - val_loss: 41.1488 - val_mae: 41.6488\n",
      "Epoch 64/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 19.4789 - mae: 19.9751 - val_loss: 39.9085 - val_mae: 40.4085\n",
      "Epoch 65/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 19.8853 - mae: 20.3800 - val_loss: 39.9053 - val_mae: 40.4053\n",
      "Epoch 66/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 19.7656 - mae: 20.2585 - val_loss: 41.1507 - val_mae: 41.6507\n",
      "Epoch 67/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 19.4631 - mae: 19.9609 - val_loss: 41.0337 - val_mae: 41.5337\n",
      "Epoch 68/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 19.2458 - mae: 19.7441 - val_loss: 39.9247 - val_mae: 40.4247\n",
      "Epoch 69/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 19.2395 - mae: 19.7385 - val_loss: 38.9186 - val_mae: 39.4186\n",
      "Epoch 70/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 19.3983 - mae: 19.8928 - val_loss: 39.7305 - val_mae: 40.2305\n",
      "Epoch 71/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 19.0110 - mae: 19.5035 - val_loss: 39.9783 - val_mae: 40.4783\n",
      "Epoch 72/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 19.1222 - mae: 19.6220 - val_loss: 40.0905 - val_mae: 40.5905\n",
      "Epoch 73/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 18.9902 - mae: 19.4902 - val_loss: 39.0536 - val_mae: 39.5536\n",
      "Epoch 74/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 18.7837 - mae: 19.2693 - val_loss: 38.5736 - val_mae: 39.0736\n",
      "Epoch 75/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 18.7511 - mae: 19.2363 - val_loss: 38.4372 - val_mae: 38.9372\n",
      "Epoch 76/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 18.6280 - mae: 19.1121 - val_loss: 38.9946 - val_mae: 39.4946\n",
      "Epoch 77/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 18.6835 - mae: 19.1806 - val_loss: 39.0065 - val_mae: 39.4861\n",
      "Epoch 78/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 18.4926 - mae: 18.9872 - val_loss: 38.1479 - val_mae: 38.6479\n",
      "Epoch 79/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 18.3852 - mae: 18.8801 - val_loss: 37.4041 - val_mae: 37.9041\n",
      "Epoch 80/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 18.3213 - mae: 18.8059 - val_loss: 37.5607 - val_mae: 38.0607\n",
      "Epoch 81/500\n",
      "3/3 [==============================] - 0s 9ms/step - loss: 18.2718 - mae: 18.7639 - val_loss: 37.8740 - val_mae: 38.3740\n",
      "Epoch 82/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 18.1646 - mae: 18.6609 - val_loss: 37.9984 - val_mae: 38.4984\n",
      "Epoch 83/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 18.1353 - mae: 18.6335 - val_loss: 37.8383 - val_mae: 38.3383\n",
      "Epoch 84/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 18.0452 - mae: 18.5438 - val_loss: 36.8255 - val_mae: 37.3253\n",
      "Epoch 85/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 17.8956 - mae: 18.3945 - val_loss: 36.6914 - val_mae: 37.1914\n",
      "Epoch 86/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 18.3186 - mae: 18.8141 - val_loss: 37.3418 - val_mae: 37.8415\n",
      "Epoch 87/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 17.6991 - mae: 18.1915 - val_loss: 34.9576 - val_mae: 35.4576\n",
      "Epoch 88/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 18.1341 - mae: 18.6273 - val_loss: 35.0580 - val_mae: 35.5580\n",
      "Epoch 89/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 17.5359 - mae: 18.0238 - val_loss: 37.1074 - val_mae: 37.6072\n",
      "Epoch 90/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 17.5872 - mae: 18.0816 - val_loss: 38.0691 - val_mae: 38.5691\n",
      "Epoch 91/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 17.6661 - mae: 18.1650 - val_loss: 36.0530 - val_mae: 36.5530\n",
      "Epoch 92/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 17.1835 - mae: 17.6716 - val_loss: 33.9970 - val_mae: 34.4970\n",
      "Epoch 93/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 17.6034 - mae: 18.0928 - val_loss: 34.3608 - val_mae: 34.8608\n",
      "Epoch 94/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 17.3103 - mae: 17.7935 - val_loss: 37.7342 - val_mae: 38.2342\n",
      "Epoch 95/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 17.5027 - mae: 18.0026 - val_loss: 36.7959 - val_mae: 37.2959\n",
      "Epoch 96/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 17.0857 - mae: 17.5775 - val_loss: 34.3730 - val_mae: 34.8730\n",
      "Epoch 97/500\n",
      "3/3 [==============================] - ETA: 0s - loss: 19.1238 - mae: 19.611 - 0s 10ms/step - loss: 17.2126 - mae: 17.7044 - val_loss: 32.9452 - val_mae: 33.4330\n",
      "Epoch 98/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 17.2318 - mae: 17.7153 - val_loss: 34.6835 - val_mae: 35.1748\n",
      "Epoch 99/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 16.8614 - mae: 17.3483 - val_loss: 36.6522 - val_mae: 37.1378\n",
      "Epoch 100/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 17.3502 - mae: 17.8424 - val_loss: 35.1961 - val_mae: 35.6961\n",
      "Epoch 101/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 16.8471 - mae: 17.3378 - val_loss: 34.6734 - val_mae: 35.1734\n",
      "Epoch 102/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 16.5688 - mae: 17.0639 - val_loss: 32.1836 - val_mae: 32.6748\n",
      "Epoch 103/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 16.9663 - mae: 17.4598 - val_loss: 32.4516 - val_mae: 32.9516\n",
      "Epoch 104/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 16.5369 - mae: 17.0333 - val_loss: 36.0674 - val_mae: 36.5674\n",
      "Epoch 105/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 17.0327 - mae: 17.5302 - val_loss: 36.1704 - val_mae: 36.6704\n",
      "Epoch 106/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 16.5939 - mae: 17.0920 - val_loss: 32.8386 - val_mae: 33.3386\n",
      "Epoch 107/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 16.6553 - mae: 17.1535 - val_loss: 31.5060 - val_mae: 31.9889\n",
      "Epoch 108/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 16.5042 - mae: 16.9990 - val_loss: 33.1242 - val_mae: 33.6242\n",
      "Epoch 109/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 16.1900 - mae: 16.6787 - val_loss: 33.6231 - val_mae: 34.1122\n",
      "Epoch 110/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 16.1753 - mae: 16.6708 - val_loss: 33.0998 - val_mae: 33.5998\n",
      "Epoch 111/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 16.2847 - mae: 16.7735 - val_loss: 32.2187 - val_mae: 32.7187\n",
      "Epoch 112/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 16.0797 - mae: 16.5720 - val_loss: 33.2867 - val_mae: 33.7867\n",
      "Epoch 113/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 16.0323 - mae: 16.5299 - val_loss: 32.4003 - val_mae: 32.8967\n",
      "Epoch 114/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 15.9859 - mae: 16.4697 - val_loss: 31.4558 - val_mae: 31.9558\n",
      "Epoch 115/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 15.9755 - mae: 16.4663 - val_loss: 31.3526 - val_mae: 31.8526\n",
      "Epoch 116/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 15.8411 - mae: 16.3298 - val_loss: 32.8375 - val_mae: 33.3375\n",
      "Epoch 117/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 15.8607 - mae: 16.3526 - val_loss: 32.5154 - val_mae: 33.0154\n",
      "Epoch 118/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.8319 - mae: 16.3224 - val_loss: 32.3623 - val_mae: 32.8623\n",
      "Epoch 119/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.9133 - mae: 16.4119 - val_loss: 32.2506 - val_mae: 32.7506\n",
      "Epoch 120/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.8105 - mae: 16.3001 - val_loss: 30.3848 - val_mae: 30.8669\n",
      "Epoch 121/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3/3 [==============================] - 0s 11ms/step - loss: 15.7665 - mae: 16.2566 - val_loss: 31.5247 - val_mae: 32.0247\n",
      "Epoch 122/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.7876 - mae: 16.2771 - val_loss: 34.1970 - val_mae: 34.6970\n",
      "Epoch 123/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 15.8239 - mae: 16.3205 - val_loss: 29.7869 - val_mae: 30.2848\n",
      "Epoch 124/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.9011 - mae: 16.3986 - val_loss: 28.9909 - val_mae: 29.4909\n",
      "Epoch 125/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 15.8517 - mae: 16.3455 - val_loss: 30.7543 - val_mae: 31.2543\n",
      "Epoch 126/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.5239 - mae: 16.0134 - val_loss: 34.1309 - val_mae: 34.6309\n",
      "Epoch 127/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.7893 - mae: 16.2857 - val_loss: 30.1250 - val_mae: 30.6233\n",
      "Epoch 128/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.3611 - mae: 15.8502 - val_loss: 28.4716 - val_mae: 28.9716\n",
      "Epoch 129/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 16.0946 - mae: 16.5899 - val_loss: 28.5204 - val_mae: 29.0204\n",
      "Epoch 130/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.7102 - mae: 16.1992 - val_loss: 31.5259 - val_mae: 32.0259\n",
      "Epoch 131/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 15.3707 - mae: 15.8659 - val_loss: 29.9592 - val_mae: 30.4472\n",
      "Epoch 132/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 15.2737 - mae: 15.7658 - val_loss: 28.3643 - val_mae: 28.8641\n",
      "Epoch 133/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.3008 - mae: 15.7899 - val_loss: 28.7894 - val_mae: 29.2894\n",
      "Epoch 134/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.1524 - mae: 15.6482 - val_loss: 30.1518 - val_mae: 30.6473\n",
      "Epoch 135/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 15.3034 - mae: 15.7975 - val_loss: 29.6749 - val_mae: 30.1725\n",
      "Epoch 136/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 15.0980 - mae: 15.5870 - val_loss: 30.3176 - val_mae: 30.8175\n",
      "Epoch 137/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.0165 - mae: 15.5116 - val_loss: 28.1953 - val_mae: 28.6764\n",
      "Epoch 138/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.1979 - mae: 15.6835 - val_loss: 27.6566 - val_mae: 28.1566\n",
      "Epoch 139/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.1651 - mae: 15.6626 - val_loss: 29.0283 - val_mae: 29.5283\n",
      "Epoch 140/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 15.0798 - mae: 15.5734 - val_loss: 29.3383 - val_mae: 29.8264\n",
      "Epoch 141/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 14.9361 - mae: 15.4313 - val_loss: 27.4836 - val_mae: 27.9836\n",
      "Epoch 142/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 14.9442 - mae: 15.4325 - val_loss: 28.3182 - val_mae: 28.8182\n",
      "Epoch 143/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 14.7955 - mae: 15.2919 - val_loss: 30.7510 - val_mae: 31.2491\n",
      "Epoch 144/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 15.1375 - mae: 15.6374 - val_loss: 29.7002 - val_mae: 30.1981\n",
      "Epoch 145/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.8991 - mae: 15.3927 - val_loss: 27.1630 - val_mae: 27.6513\n",
      "Epoch 146/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.7542 - mae: 15.2471 - val_loss: 27.6312 - val_mae: 28.1167\n",
      "Epoch 147/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.5373 - mae: 15.0353 - val_loss: 30.2205 - val_mae: 30.7103\n",
      "Epoch 148/500\n",
      "3/3 [==============================] - ETA: 0s - loss: 17.0996 - mae: 17.586 - 0s 10ms/step - loss: 14.9651 - mae: 15.4599 - val_loss: 29.5434 - val_mae: 30.0390\n",
      "Epoch 149/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 14.9066 - mae: 15.4036 - val_loss: 28.7087 - val_mae: 29.1977\n",
      "Epoch 150/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 14.6904 - mae: 15.1818 - val_loss: 26.7065 - val_mae: 27.2065\n",
      "Epoch 151/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 14.5668 - mae: 15.0590 - val_loss: 26.9992 - val_mae: 27.4989\n",
      "Epoch 152/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 14.2906 - mae: 14.7808 - val_loss: 28.6841 - val_mae: 29.1602\n",
      "Epoch 153/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 14.5521 - mae: 15.0520 - val_loss: 28.3887 - val_mae: 28.8703\n",
      "Epoch 154/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.5371 - mae: 15.0309 - val_loss: 26.9168 - val_mae: 27.4012\n",
      "Epoch 155/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.2325 - mae: 14.7125 - val_loss: 26.4789 - val_mae: 26.9789\n",
      "Epoch 156/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.2306 - mae: 14.7230 - val_loss: 26.2296 - val_mae: 26.7251\n",
      "Epoch 157/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.2055 - mae: 14.6994 - val_loss: 26.4162 - val_mae: 26.9162\n",
      "Epoch 158/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.1233 - mae: 14.6037 - val_loss: 26.7106 - val_mae: 27.2050\n",
      "Epoch 159/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.1513 - mae: 14.6509 - val_loss: 26.2702 - val_mae: 26.7702\n",
      "Epoch 160/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.0933 - mae: 14.5728 - val_loss: 25.7251 - val_mae: 26.2251\n",
      "Epoch 161/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.2326 - mae: 14.7241 - val_loss: 25.9112 - val_mae: 26.4102\n",
      "Epoch 162/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.9938 - mae: 14.4826 - val_loss: 25.8764 - val_mae: 26.3760\n",
      "Epoch 163/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.9258 - mae: 14.4081 - val_loss: 25.7677 - val_mae: 26.2677\n",
      "Epoch 164/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.9365 - mae: 14.4278 - val_loss: 25.7941 - val_mae: 26.2941\n",
      "Epoch 165/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 13.9082 - mae: 14.4004 - val_loss: 25.5379 - val_mae: 26.0379\n",
      "Epoch 166/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.7699 - mae: 14.2561 - val_loss: 26.1703 - val_mae: 26.6693\n",
      "Epoch 167/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 14.2743 - mae: 14.7743 - val_loss: 27.4281 - val_mae: 27.9280\n",
      "Epoch 168/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.8726 - mae: 14.3636 - val_loss: 25.3624 - val_mae: 25.8624\n",
      "Epoch 169/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.6295 - mae: 14.1056 - val_loss: 24.9379 - val_mae: 25.4379\n",
      "Epoch 170/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.8635 - mae: 14.3519 - val_loss: 25.2571 - val_mae: 25.7571\n",
      "Epoch 171/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.6091 - mae: 14.1021 - val_loss: 25.7598 - val_mae: 26.2459\n",
      "Epoch 172/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.9167 - mae: 14.4151 - val_loss: 25.4713 - val_mae: 25.9483\n",
      "Epoch 173/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.7429 - mae: 14.2412 - val_loss: 25.3295 - val_mae: 25.8043\n",
      "Epoch 174/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.6755 - mae: 14.1702 - val_loss: 24.7941 - val_mae: 25.2941\n",
      "Epoch 175/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.4613 - mae: 13.9426 - val_loss: 25.0830 - val_mae: 25.5649\n",
      "Epoch 176/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 13.5731 - mae: 14.0710 - val_loss: 25.5703 - val_mae: 26.0697\n",
      "Epoch 177/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.4171 - mae: 13.9110 - val_loss: 24.7032 - val_mae: 25.2032\n",
      "Epoch 178/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 13.3789 - mae: 13.8668 - val_loss: 24.3128 - val_mae: 24.8128\n",
      "Epoch 179/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 13.3512 - mae: 13.8412 - val_loss: 25.4260 - val_mae: 25.9260\n",
      "Epoch 180/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.3688 - mae: 13.8645 - val_loss: 26.1222 - val_mae: 26.6222\n",
      "Epoch 181/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3/3 [==============================] - 0s 10ms/step - loss: 13.5694 - mae: 14.0694 - val_loss: 25.8322 - val_mae: 26.3154\n",
      "Epoch 182/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.2000 - mae: 13.6992 - val_loss: 24.1472 - val_mae: 24.6472\n",
      "Epoch 183/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.4243 - mae: 13.9072 - val_loss: 23.8955 - val_mae: 24.3892\n",
      "Epoch 184/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.0042 - mae: 13.5026 - val_loss: 25.4790 - val_mae: 25.9790\n",
      "Epoch 185/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.9159 - mae: 14.4128 - val_loss: 28.5987 - val_mae: 29.0987\n",
      "Epoch 186/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.5712 - mae: 14.0610 - val_loss: 24.1575 - val_mae: 24.6534\n",
      "Epoch 187/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.3716 - mae: 13.8569 - val_loss: 23.7050 - val_mae: 24.2049\n",
      "Epoch 188/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.2427 - mae: 13.7220 - val_loss: 24.5271 - val_mae: 25.0270\n",
      "Epoch 189/500\n",
      "3/3 [==============================] - 0s 9ms/step - loss: 13.1849 - mae: 13.6804 - val_loss: 27.2852 - val_mae: 27.7852\n",
      "Epoch 190/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 13.4474 - mae: 13.9412 - val_loss: 24.7319 - val_mae: 25.2319\n",
      "Epoch 191/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.8934 - mae: 13.3921 - val_loss: 23.4282 - val_mae: 23.9282\n",
      "Epoch 192/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.8833 - mae: 13.3780 - val_loss: 23.4374 - val_mae: 23.9126\n",
      "Epoch 193/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.8670 - mae: 13.3577 - val_loss: 24.2055 - val_mae: 24.6851\n",
      "Epoch 194/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.9455 - mae: 13.4392 - val_loss: 23.8033 - val_mae: 24.3033\n",
      "Epoch 195/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.5873 - mae: 13.0811 - val_loss: 24.5168 - val_mae: 25.0168\n",
      "Epoch 196/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.7155 - mae: 13.2154 - val_loss: 24.3160 - val_mae: 24.8159\n",
      "Epoch 197/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.6164 - mae: 13.1046 - val_loss: 23.5825 - val_mae: 24.0824\n",
      "Epoch 198/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.5246 - mae: 13.0162 - val_loss: 23.2989 - val_mae: 23.7989\n",
      "Epoch 199/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.5449 - mae: 13.0365 - val_loss: 23.5102 - val_mae: 24.0057\n",
      "Epoch 200/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.5623 - mae: 13.0530 - val_loss: 23.5566 - val_mae: 24.0554\n",
      "Epoch 201/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.3288 - mae: 12.8123 - val_loss: 22.8634 - val_mae: 23.3439\n",
      "Epoch 202/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.6252 - mae: 13.1189 - val_loss: 22.9295 - val_mae: 23.4295\n",
      "Epoch 203/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.5140 - mae: 13.0104 - val_loss: 23.4423 - val_mae: 23.9394\n",
      "Epoch 204/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 12.4381 - mae: 12.9298 - val_loss: 23.1390 - val_mae: 23.6390\n",
      "Epoch 205/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.2462 - mae: 12.7399 - val_loss: 22.5807 - val_mae: 23.0807\n",
      "Epoch 206/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.6462 - mae: 13.1349 - val_loss: 22.8647 - val_mae: 23.3647\n",
      "Epoch 207/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 12.4039 - mae: 12.8931 - val_loss: 25.0346 - val_mae: 25.5346\n",
      "Epoch 208/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.5314 - mae: 13.0235 - val_loss: 23.3902 - val_mae: 23.8899\n",
      "Epoch 209/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 12.1288 - mae: 12.6203 - val_loss: 22.2594 - val_mae: 22.7430\n",
      "Epoch 210/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 12.7364 - mae: 13.2315 - val_loss: 22.4439 - val_mae: 22.9439\n",
      "Epoch 211/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 12.1403 - mae: 12.6304 - val_loss: 24.4633 - val_mae: 24.9633\n",
      "Epoch 212/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 12.3521 - mae: 12.8416 - val_loss: 23.1292 - val_mae: 23.6169\n",
      "Epoch 213/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 11.7764 - mae: 12.2669 - val_loss: 22.1143 - val_mae: 22.6143\n",
      "Epoch 214/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 12.4722 - mae: 12.9703 - val_loss: 22.0392 - val_mae: 22.5392\n",
      "Epoch 215/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 12.3456 - mae: 12.8430 - val_loss: 23.6137 - val_mae: 24.1137\n",
      "Epoch 216/500\n",
      "3/3 [==============================] - 0s 14ms/step - loss: 12.2469 - mae: 12.7366 - val_loss: 23.5710 - val_mae: 24.0710\n",
      "Epoch 217/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 12.2427 - mae: 12.7362 - val_loss: 22.2466 - val_mae: 22.7466\n",
      "Epoch 218/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.9867 - mae: 12.4813 - val_loss: 22.4064 - val_mae: 22.9064\n",
      "Epoch 219/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.7876 - mae: 12.2822 - val_loss: 22.6816 - val_mae: 23.1816\n",
      "Epoch 220/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.6494 - mae: 12.1382 - val_loss: 23.2191 - val_mae: 23.7186\n",
      "Epoch 221/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.8760 - mae: 12.3710 - val_loss: 22.5545 - val_mae: 23.0545\n",
      "Epoch 222/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 11.5640 - mae: 12.0553 - val_loss: 22.1304 - val_mae: 22.6304\n",
      "Epoch 223/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.7733 - mae: 12.2692 - val_loss: 22.0229 - val_mae: 22.5229\n",
      "Epoch 224/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.7202 - mae: 12.2202 - val_loss: 22.4809 - val_mae: 22.9809\n",
      "Epoch 225/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.6541 - mae: 12.1474 - val_loss: 22.6033 - val_mae: 23.0880\n",
      "Epoch 226/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.3586 - mae: 11.8457 - val_loss: 21.8460 - val_mae: 22.3460\n",
      "Epoch 227/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.9701 - mae: 12.4633 - val_loss: 21.6155 - val_mae: 22.1155\n",
      "Epoch 228/500\n",
      "3/3 [==============================] - 0s 13ms/step - loss: 11.7680 - mae: 12.2612 - val_loss: 22.4829 - val_mae: 22.9693\n",
      "Epoch 229/500\n",
      "3/3 [==============================] - 0s 11ms/step - loss: 11.4631 - mae: 11.9514 - val_loss: 22.2543 - val_mae: 22.7543\n",
      "Epoch 230/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.4029 - mae: 11.8874 - val_loss: 22.1904 - val_mae: 22.6904\n",
      "Epoch 231/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.3520 - mae: 11.8395 - val_loss: 21.7819 - val_mae: 22.2819\n",
      "Epoch 232/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.4524 - mae: 11.9376 - val_loss: 22.1132 - val_mae: 22.6132\n",
      "Epoch 233/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.2629 - mae: 11.7553 - val_loss: 22.1095 - val_mae: 22.6095\n",
      "Epoch 234/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.2789 - mae: 11.7637 - val_loss: 22.0426 - val_mae: 22.5426\n",
      "Epoch 235/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.2404 - mae: 11.7320 - val_loss: 21.8242 - val_mae: 22.3242\n",
      "Epoch 236/500\n",
      "3/3 [==============================] - 0s 10ms/step - loss: 11.2075 - mae: 11.6967 - val_loss: 22.0701 - val_mae: 22.5679\n",
      "Epoch 237/500\n",
      "3/3 [==============================] - 0s 12ms/step - loss: 11.2869 - mae: 11.7650 - val_loss: 22.1825 - val_mae: 22.6594\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x1da28fce148>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "keras = tf.keras\n",
    "\n",
    "# Create Windowed Datasets\n",
    "window_size = 20\n",
    "train_set = window_dataset(x_train, window_size)\n",
    "valid_set = window_dataset(x_valid, window_size)\n",
    "\n",
    "# Simple Dense Model Forecasting\n",
    "model = keras.models.Sequential([\n",
    "  keras.layers.Dense(10, activation=\"relu\", input_shape=[window_size]),\n",
    "  keras.layers.Dense(10, activation=\"relu\"),\n",
    "  keras.layers.Dense(1)\n",
    "])\n",
    "\n",
    "optimizer = keras.optimizers.SGD(lr=1e-5, momentum=0.9)\n",
    "model.compile(loss=keras.losses.Huber(),\n",
    "              optimizer=optimizer,\n",
    "              metrics=[\"mae\"])\n",
    "early_stopping = keras.callbacks.EarlyStopping(patience=10)\n",
    "model.fit(train_set, epochs=500,\n",
    "          validation_data=valid_set,\n",
    "          callbacks=[early_stopping])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1da3f6ad8c8>]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA20AAAFlCAYAAAB4PgCOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdeXxV93nn8c+5ulrQhvYNhAQCIcQiFoGxsLENxmsSx0kdO4njNXHSOuk002mTzqSdTtq0adNkstvj1HbsJLazNN5is9hgMJuNJMwisQihBbTvK2i9Z/44ulgYLVfibpK+79fLryPO/Z17HrWRdJ/z+z3PzzBNExEREREREfFPNl8HICIiIiIiIqNT0iYiIiIiIuLHlLSJiIiIiIj4MSVtIiIiIiIifkxJm4iIiIiIiB9T0iYiIiIiIuLH7L4OACAuLs5MT0/3dRgiIiIiIiI+UVhY2GSaZvxIr/lF0paenk5BQYGvwxAREREREfEJwzAqR3tNyyNFRERERET8mJI2ERERERERP6akTURERERExI8paRMREREREfFjStpERERERET8mJI2ERERERERP6akTURERERExI8paRMREREREfFjStpERERERET8mJI2ERERERERP6akTURERERExI8paRMRERER8RctZTDQ6+soxM8oaRMRERER8Qed9fDTdfDs7dbXIkOUtImIiIiI+IOqQ+Doh5oj8ItNUFfk64jETyhpExERERHxAwPnDtFPIHs3/hpMBzxzK5Rs93VY4geUtImIiIiI+IHO0oMUOdJ46C3Yv+l3ELsQXrwPDv4cTNPX4YkPKWkTEREREfG1wQHCmo5TZCxiSXIEj/6xig82vwBZd8L2v4M3/jsM9vs6SvERJW0iIiIiIj5mNhQTZPbQl7SaXz68jqTIEB76TTElN/wMrvs6FDwDv7kHLrb5OlTxASVtIiIiIiI+VlO0F4DkZRuJCw/mV49eQ5DdxgPPFFCd+w246+dQsQ+e3mJtCyAzipI2EREREREfaztzkCYzknUrVwKQGhPK84+so7tvgC88/T4tmffAA69AdyP8YjNUHvRxxOJNStpERERERHwssvkI5cFLiIsIuXRuSXIk//lALlWtF3n4l/l0J6+HL+6E0Bh4/hNw5EUfRizepKRNRERERMSHGhrqSB2sYjAl94rXrlkQy08/u4rjVW38+W8O0zd7PnzxbZi3Hl75Cuz8J3A4fBC1eJOSNhERERERHzqRvxuAlGXXjfj6LUuT+NdPLefdkkb+x++P4giOgvv/CKsfhL3/AX94CPoueC9g8Tq7rwMQEREREZnJ2s4cwIFB6rINo465d+08mrv7+Pdtp4kND+IfPpaN8fEfQVwm7PgWtJ2Hz74IEUlejFy8RTNtIiIiIiI+0tM/SEzrURpD5mOEzB5z7J/fkMEjG+bz7P4Kfr77LBgG5H0V7nsBGk9bDUrqjnspcvEmJW0iIiIiIj5ysLSJ5ZTiSFkz7ljDMPjWnUv45MoUvrf9NC8dOme9kHUHPLINMOHpW+H0Vs8GLV6npE1ERERExEeOHCsk2ugiLmv0pZHD2WwG37snhxsy4/mfLx9ne3Gd9ULyCvjSLojPhBc/Cwd/7sGoxduUtImIiIiI+IBpmnSesfZbC0y7xuXrAgNsPHH/albMjeJrL37Ae2XN1gsRSfDQm5B1J2z/O23CPY0oaRMRERER8YGTtZ2k9ZykPyAU4hdP6NrQIDvPPrSWeTGhfOm5Ak7UdFgvBIXCDd+wvq4+7OaIxVeUtImIiIiI+MDOk/WsspVizlkNtoAJXx8dFsTzj6wjPMTOg88e4lzzUNv/+CwICIK6Y26OWHzFpaTNMIwowzD+YBjGKcMwThqGca1hGDGGYbxlGMaZoWP00FjDMIwfG4ZRahjGMcMwVnv2WxARERERmXr2njxPtu0cQRNYGvlRKVGzeP6RdfQPOvjCM+/T2NkL9iBIWAK1StqmC1dn2n4EbDNNMwvIAU4C3wR2mqa5CNg59G+A24FFQ/89Bjzh1ohFRERERKa4xs5eHDVHsDMIc3Kv6r0WJUbwzENraejo5aFnD9HZ0w9JK6D2KJimmyIWXxo3aTMMIxLYCDwNYJpmn2mabcBdwHNDw54DPjn09V3A86blPSDKMIxkt0cuIiIiIjJFvXO6gZVGqfWPuVeXtAGsnhfNz+9fzem6Th57vpD+hBVwsQU6qq/6vcX3XJlpWwA0As8ahvGBYRj/aRhGGJBommYtwNAxYWj8HOD8sOurhs5dxjCMxwzDKDAMo6CxsfGqvgkRERERkalk58l68oLLMKPSIDxh/AtccNPiBL53zwoOljXzw+IQ62TtUbe8t/iWK0mbHVgNPGGa5iqgmw+XQo7EGOHcFfOypmk+ZZpmrmmaufHx8S4FKyIiIiIy1fUODLL3TBOrA85iuGGWbbi7V83lb29bzDOlYZiGTXVt04QrSVsVUGWa5vtD//4DVhJX71z2OHRsGDY+ddj1c4Ea94QrIiIiIjK1vVfWQkRfI1H9DTB3rdvf/5EN8wmaFUFdYKpm2qaJcZM20zTrgPOGYTg3j9gMnABeAx4cOvcg8OrQ168BDwx1kVwPtDuXUYqIiIiIzHS7TtazLnBo42sPJG0hgQHcvWoOh3pScShpmxZc7R75NeA3hmEcA1YC/wJ8F9hiGMYZYMvQvwHeBMqAUuAXwF+4NWIRERERkSnKNE3ePtnAHTHV1l5qScs9cp/P5KZyfDANW2cNdDd55B7iPXZXBpmmeQQYacHt5hHGmsDjVxmXiIiIiMi0U1LfRXXbRdaklFpt+e3BHrlPdkokv4ldCh1g1h7FWHjFx3aZQlydaRMRERERkav09sl6AhgkvuOER5ZGDpezdiMAdacPefQ+4nlK2kREREREvGTXqQbuTGzFGLjolv3ZxnLb2iyqzHgaS5S0TXVK2kREREREvKC5q5fD51q5K26osbqHk7bIkEBaIpcQ2XaSi32DHr2XeJaSNhERERERL9h9uhHThNW2UgiLh6g0j98zOiOXdKOWHYfPePxe4jlK2kREREREvGDnqXoSI4OJaj0Gc3LBMDx+z7lLrgGg8NBej99LPEdJm4iIiIiIh/UNOHi3pIk7Fs7CaCrx+NJIJyNlJQC2+mOUNXZ55Z7ifkraREREREQ87FB5C129A3w8rtY64eHOkZdEJDEYGs9yWwW/K6jyzj3F7ZS0iYiIiIh42Nsn6wm221hungEMSFnltXsHpKxk3awq/lBYRf+gw2v3FfdR0iYiIiIi4kGmabLzVD0bFsYRWFsICUsgJNJ7ASSvYE7/OTq7OnnnVIP37ituo6RNRERERMSDShu6ON9ykU2L46G6wGv1bJck52AzB1gfVs9v8897997iFkraREREREQ8aOfQ7NYtyd1wsdXqHOlNSSsA+Oy8Nt453UB9R4937y9XTUmbiIiIiIgH7TxZT3ZyJAntx60T3mpC4hSdDsGzyQurxmHCHwrVkGSqUdImIiIiIuIhrd19FFa2cvOSBKjKh6AIiF/s3SAMA5JXENl6gmvmx/C7gvM4HKZ3Y5CroqRNRERERMRDdpc04DBh05JEK2mbswpsAd4PJGkF1Bfx2dxkKpsv8H55i/djkElT0iYiIiIi4iE7TzYQFx7MioRAqC/2/tJIp+QcGOjh9uQuIkLs/Db/nG/ikElR0iYiIiIi4gH9gw72lDSyKSseW90xcAz4MGmzmpEENxbzyZVz2FpUR/vFft/EIhOmpE1ERERExAPyK1ro7Blgs3NpJHi/c6RT7CKwh0DtUe5dm0rvgIPXjlT7JhaZMCVtIiIiIiIesPNkA0EBNq5bGGftzxaVBuHxvgkmwA6Jy6DuGMvmzCY7OZKXtGfblKGkTURERETEA3adauDajFjCgu1QVeC7pZFOySug9hiYJvetS6W4poOi6nbfxiQuUdImIiIiIuJmZxu7KG/qZvOSBOiogY5qmOujpZFOyTnQ2w6tFdyVM4cgu43fFWi2bSpQ0iYiIiIi4ma7TjYAsCkrwZplA9/PtCVZzUioO8bs0EDuWJbEyx9U09M/6Nu4ZFxK2kRERERE3Oztk/VkJUUwNzrUakISEARJy30bVEI22OxQexSAz6xNpbNngG1Fdb6NS8alpE1ERERExI3aL/RTUNlqLY0Ea6YtaQXYg30bWGAIxGdZdW3A+vmxzIsJ5SXt2eb3lLSJiIiIiLjR7pIGBh0mm7ISYXAAaj7w/dJIp6QVl2babDaDe9em8l5ZC5XN3T4OTMaipE1ERERExI12nWogNiyIlalR0FAMAxd934TEKTkHuhug01oS+enVc7EZqCGJn1PSJiIiIiLiJgODDnafbuTGxQkE2IwPN9X2m6RtqBnJ0Gxb0uwQblycwO8LqhgYdPgwMBmLkjYRERERETcprGyl/WI/N1+qZyuEsHhrY21/4GyGMlTXBnDv2lQaOnvZU9Loo6BkPEraRERERETcZOepBgIDDK7PjLdOVOVb9WyG4dvAnIIjICYD6o5eOrUpK4G48GBeytcSSX+lpE1ERERExE12nqxn/YJYwoPtcLEVms/AnDW+DutyyR82IwEIDLDx6TVz2HWqgYbOHh8GJqNR0iYiIiIi4gYVTd2cbexmc9bQ0sjqQuvoL50jnZJzoO2clVQO+UxuKoMOkz8ervZhYDIaJW0iIiIiIm6w81QDAJuXJFonqgoAA+as9l1QI0lyNiP5sK4tIz6ctenR/C7/PKZp+igwGY2SNhEREfG9plJorfB1FCJXZefJejITw0mNCbVOVOVDwhKrjsyfJOdYx7pjl52+d+08ypq6ya9oHeEi8SUlbSIiIuI73c3wp6/Dz9bC7x/ydTQik9bR08+h8hZrQ20A07Rm2vyl1f9wYXEQOeeymTaAO5YnER5s56X8cz4KTEajpE1ERES8b7Af3nsSfrIKCp8b6mZ3HPrVBEGmpndLGhlwmB+2+m8+Cz1t/lfP5pScc1kzEoDQIDufWJnCm8dr6ejp91FgMhIlbSIiIuJdZ9+BJ6+Dbd+AlNXw5wdg07fAMQANJ3wdncikvHWinujQQFbNi7ZOODfVnuOHM21g1bU1n4G+7stO35ubSk+/g9eP1vgoMBmJkjYRERHxjpZyeOnz8KtPwkAP3PcCfOFlSMiClJXWmNojvo1RZBJO1nbwp2O1fDwnhQDb0H5s1QUQFAHxi30b3GiSV4DpgPriy06vmDubrKQIfqs92/yKkjYRERHxrN4u2Plt+Nk11izb5n+Av3gfsu78cMPhqDQIibpiuZaIv3M4TL71ShGzZwXy9ZszP3yhKt/qGmkL8F1wY3E2I/nIz5xhGNy7NpVjVe2crO3wQWAyEpeSNsMwKgzDOG4YxhHDMAqGzv2jYRjVQ+eOGIZxx7Dxf2cYRqlhGKcNw7jVU8GLiIiIHzNNOPpb+Gku7P0+LL0bvlYA1/81BIZcPtYwrA+RNZppk6nlD4VVFFa28s3bs4gOC7JO9l2AuiL/bELiFDkHZsVc0UES4JMr5xAUYNNsmx+xT2DsTaZpNn3k3P81TfM/hp8wDCMbuA9YCqQAbxuGkWma5uDVhSoiIiJTRvVh2PoNqDoEKavgM89D6rqxr0nOgfefhIE+sAd5J06Rq9Da3ce/bj3J2vRo/mz13A9fqD0K5qD/NiGBDx+UjDC7HR0WxK3Lknj5g2q+eXsWIYF+Ols4g3hieeRdwEumafaaplkOlALj/JYWERGRaaGrAV59HH6xydp37a6fwRd3jZ+wgfUBcrAPGk96PEwRd/i3bafo6Bngnz65DJuzlg38vwmJU/IKaDhpPSj5iHtzU2m/2M+OE/U+CEw+ytWkzQR2GIZRaBjGY8POf9UwjGOGYTxjGMZQqxzmAMPnUquGzomIiMh0NdAHB34CP1ljLYnM+yp8rRBW3Q82Fz9upKyyjqprkymgsLKVl/LP88iGdLKSIi9/sSrfqtMMj/dNcK5KWjH0oOTUFS/lZcQyN3oWvy/QEkl/4GrStsE0zdXA7cDjhmFsBJ4AMoCVQC3w/aGxxgjXmx89YRjGY4ZhFBiGUdDY2DjxyEVERMTnTNOk58RWeOJa2PEtmHct/MV7cMs/Q0jk+G8wXPR8q9uekjbxcwODDr71ShFJkSH81fDmI07Vhf69NNIpeahr6wh1bTabwR3Lk3m/rIWLfapy8jWXkjbTNGuGjg3Ay8A60zTrTdMcNE3TAfyCD5dAVgGpwy6fC1yx0YNpmk+ZpplrmmZufLyfP4UQERGREb37ylOE/O4+Klsu8KOkf+FHid9hZ2ME9R09mOYVz2zHZrOpGYlMCc8frORkbQf/++PZhAV/pEVEezV0VE+NpC1mAQSFj/qg5NqMWPoGHRRWtno5MPmocRuRGIYRBthM0+wc+voW4NuGYSSbplk7NOxuoGjo69eAFwzD+AFWI5JFwCH3hy4iIiK+1NnTz4Wjr9JsxPCjzOc4WnOBsp0lOHO1uPBgls2JZFnKbJbNmc2yOZHMiZqFYYy0KGdIcg4UPA2DAxAwkX5pIt5R39HDD94q4YbMeG5blnTlgOoC6+jPnSOdbDZIWg61V860AaxLj8FuM9h/tonrFsV5OTgZzpXfhonAy0O/YO3AC6ZpbjMM41eGYazEWvpYAXwZwDTNYsMwfgecAAaAx9U5UkREZPr5xbtlfN4sxpaxkR981lpw0907wMnaDoqq2zle3UFxTTt7zzQx6LAyuajQQJalzGbpsGQuLSb0wyYOKSutjbebTkPiUl99ayKj+qc/naBv0MH/+cTSkR9AVBVAQJCVDE0FSSvgg1+Dw3FF/WlYsJ2VqVEcONvso+DEadykzTTNMiBnhPNfGOOa7wDfubrQRERExF81dfXy1r79/HdbGyzZdOl8WLCd3PQYctNjLp3r6R/kVF0nRdXtFNe0U1TdwbP7KugbdAAQHmwnOyWSRzakc9vwDX+VtImf2XumkT8dq+XrN2eSHhc28qCqAmvG2B7s3eAmKzkHDv0/aDkLcYuueDkvI5afvlNKR08/kSGBPghQYGL7tImIiIgA8NNdpax2FFvV8enXjTk2JDCAlalRrEyNunSub8DBmYZOiqs7KKpp553TDXzzj8fZ/Hc3ERgYZtW1rfych78LEdf19A/yD68Wkx4bypdvWDDyoMF+qPkA1jzk1diuSvIK61h7dOSkbWEcP95VyqGyFm7OTvRycOLkiX3aREREZBo733KBF94/xz1xlRCeCLELJ/weQXYbS1Nm85m1qXz7rmV8685s2i70835F+1CNjTpIin956t0yypu6+fZdy0bfbLrhBAxcnBr1bE7xWdZyzlF+5lbNiyLYbmP/2SYvBybDKWkTERGRCfnh22cwDJPl/cetWbaxGou46IbMeEKDAthaVGst16o7Bg6VxIt/qGzu5qfvlHLnimQ2Zo7R9dy5qfZU6BzpFBAICdkjtv0HCLYHsDY9hoOqa/MpJW0iIiListN1nfzxgyq+vjqAgO66cZdGuiokMICbshLYXlyHIzkH+i9Ac6lb3lvkapimyT++VkygzeDv78wee3BVAYTFQ9Q87wTnLskrrA6So2zTcW1GLKfqOmnq6vVyYOKkpE1ERERc9h87ThMebOeB5CrrRJp7kjaA25cl0dTVR5Ej3Tqh/drED2wvruOd0418fUsmSbNDxh5cVWDNsrlh9tmrknPgYgu0V4348oaFVrt/zbb5jpI2ERERcUlhZStvnajnKzdkEFrzHoQljNi4YLJuWpxAsN3GK1XhYA9RXZv4XHfvAP/n9RNkJUXwUF762IMvtEDzmalVz+aUNNS1dZQlkstSIokItqv1vw8paRMREZFxmabJv207RVx4MA/npUHFPrfVszmFBdu5ITOeN4obMROXKWkTn/vxzjPUtvfwnbuXYQ8Y52Nz9WHrOGcKJm2JS8GwjfozZw+wcc2CGA6qGYnPKGkTERGRce0paeRQeQv/bfNCQrvOQWcNpG9w+33uWJ5MfUcvjRFLrA+QDofb7yHiitN1nTy9r5x7c1NZkxYz/gVV+YABc1Z7PDa3CwqFuEyrrm0UeRlxVDRfoLrtohcDEyclbSIiIjImh8Pk37edZl5MKPeunQeV+60X0q93+702LUkgMMDg0MW50NcJreVuv4fIeEzT5FuvHCc8xM43bs9y7aKqfKsLY3CEZ4PzlKQVY85u5y2MBeBAqWbbfEFJm4iIiIzpT8drOVHbwV/fkkmQ3WYtjQyLt57Mu1lkSCDXL4rnD7VDbdVrPnD7PUTG81+Hq8mvaOXvbs8iJixo/Av6LlgPM+a7/0GG1yTnWDPo3SMnZZkJEcSGBakZiY8oaRMREZFR9Q86+P6O02QlRfDxFSlWS3AP1LMNd9uyJPZ3xOGwjb7hr4intF3o41/fPMnqeVHcsybVtYvKdsNAD2Te5tHYPCp5hXUc5WfOZjNYnxHLgbPNmKNsDSCeo6RNRERERvXb/PNUNl/gG7dlYbMZ0FoBHdWQ5v56NqdbshMxbYHUz8qAWrX9F+/69+2nab3Qxz9/crn1v3lXlGyF4EiP/lx4XNLYSRvAhow46jp6KGvq9lJQ4qSkTUREREZ0sW+QH+08w7r0GG5cPLRcsWKfdfRAPZtTVGgQ12bEkt+bill7dNQNf0Xc7YNzrbx46BwP5c0nOyXStYscDijZDhmbwO7CUkp/NSsKotJGbfsPkJcxVNemJZJep6RNRERERvTsgXIaO3v529sWYziXQlbsg9A4iF/s0XvfviyZ9y6mYvS0W7N7Ih426DD51itFJEQE8/UtE9h/sPYD6KqHxbd7LjhvSR67GUlabCgps0PU+t8HlLSJiIjIFdov9PPk7rPcvCSB3PShdudeqGdzumVpIsXmfOsfqmsTL/jVwQqKazr4+49lExES6PqFJdutPc4WbvFYbF6TnAMtZdDTMeLLhmGQtzCOg2ebcTg0A+5NStpERETkCk/sOUtn7wD/49ZhM2ptldBRZSVtHhYXHkxk2goGCFDSJh7X0NHD93eUcP2iOO5cnjyxi09vhbnrICzWM8F5U1KOdawvGnVIXkYsrRf6OVk3cmInnqGkTURERC5T197Ds/vLuXvlHLKShtX1XKpn83zSBrBlRRqnHXPpriz0yv1k5vrOmyfpHXDw7buWfbgU2BXt1VYN2OIp3DVyuOShpG2s/doy4gDU+t/LlLSJiIjIZX686wwO0+TrWz6yD1vFPgiNhXgXNxu+SrcuTaLIMR+bmpGIB71X1syrR2r4yo0ZzI8Lm9jFJdusY+Y0qGcDiEiE8ESoHb0ZSdLsEBbEh6kZiZcpaRMREZFLypu6+W3+eT5/TRqpMaGXv1ix32pp7uF6NqfEyBA6opcya6AN2qu8ck+ZeV4/WkNEsJ2/uDFj4heXbIPodI835vGqpLGbkYC1RPL9smb6Bx1eCkqUtImIiMgl399xmmC7jcdvWnj5C62V0H7Oo63+R5KweB0ADSXve/W+MnMUVLSyOi2akMCAiV3Y1w1le6xZNi89yPCK5BxoPAX9PaMOycuIo7tvkGNV7V4MbGZT0iYiIiIAFFW386djtXzxuvnERwRf/qKX69mcctdvZMC0ca74oFfvKzND+4V+Ttd3sjY9euIXl+2Gwd7pU8/mlLwCzEFoODHqkPULrKYrav3vPUraREREBIB/336a6NBAvrhxwZUvermezWlOXAxV9lTMmiNeva/MDIXnWgA+3NZiIk5vheBImJfn5qh8zIVmJDFhQWQnR6quzYuUtImIiAgHzjbxbkkjj9+0kMiR9qiq3AdpeWDz/keH3vjlpPedoaql2+v3luntUHkrgQEGOXOjJnahwwFndkDGJrAHeSY4X4lKg5DZVlfMMeRlxFJQ2UpP/6CXApvZlLSJiIjMcKZp8u/bTpM8O4T716ddOaC1Etq8X8/mFJ+5jnijnXcPH/fJ/WX6KqhoYdmc2cwKmmA9W+0H0FUPi6dJ18jhDMO1ZiQLY+kbcHC4stVLgc1sStpERERmuB0n6jlyvo2v35w5cjOGyv3W0cv1bE4xGVYzksqiAz65v0xPPf1WI421k1oauQ0MGyy6xf2B+YPkHKgvhsGBUYesmx9LgM3QEkkvUdImIiIygw06TL63/TQZ8WF8avWckQdV7IdZ0RC/xLvBOSUtx8QgpLGIuvbRO9qJTMTx6nb6Bh3kpk2iCUnJVki9BkInkfBNBUkrYKAHmkpGHRIebCdn7mz2qxmJVyhpExERmcH+eLiK0oYu/ubWxdgDRvlYULHX2p/NB/VsAASH0x+VwTJbBduL63wTg0w7+RVWE5I1E03a2qug7jhkTrOukcM5m5GMW9cWx7Gqdjp7+r0Q1MympE1ERGSG6ukf5IdvnyFn7mxuXZo08qC2c9BW6bN6Nqeg1NWstFfw5vFan8Yh00dBRSsZ8WHEhgePP3i4km3WcTrWsznFLQL7LKgdJ2lbGMugw7yUAIvnKGkTERGZoX7z/jmq2y7yjduyMEbbHLjCt/VslyTnEG82U1ZRTmNnr29jkSnP4TApqGiZfD1b9HyIy3R/YP7CFgBJy8ZtRrJ6XjRBdhsHSlXX5mlK2kRERGag7t4BfvZOKdcviiNvYdzoAyv3WfVsCdneC24kKSsBWGqUs+OElkjK1TnT0EVHz8DE92fr64byd61ZttEedEwXSSus5ZEOx6hDQgIDyE2LZr+akXickjYRmXEaOnqobrvo6zBEfOoPhVW0dPfxVzePM1tQsc+39WxOScsBuC68mq3HlbTJ1XEu51ubPsF6trLdMNgLmbe6Pyh/k7wCejugrWLMYXkZsZys7aClu887cc1QStpEZEYZdJjc94v3uP7fdvEXvynkg3PaX0ZmHofD5Nn95ayaFzV2E4b2Kmit8P3SSLA2+43JYGNYNQfLmmnVB0S5CgUVLcRHBDMvJnRiF57eCsGRMC/PM4H5E2czknHr2qyZ+vfKNNvmSUraRGRGeetEPWWN3WzJTmTfmSbu/vkB7nnyADuK63A4TF+HJ+IVu041UNF8gUevmz/2QH+pZ3NKziG9v5RBh8lbJ+p9HY1MYfkVraxNjx69lnMkDgeUbIeFm8Ee5Lng/PKtIxcAACAASURBVEVCNtjs43aQXDFnNuHBdvaXqvW/JylpE5EZ5Rd7y5gbPYuffW41B/5uM//wsWxq2np47FeF3PyDPfzm/Up6+gd9HaaIRz29r5yU2SHcNlrHSKeKvRASBQlLvRPYeJJzCOqqIjuqn61F6iIpk1PTdpHqtovkpk2wnq3mA+hugMxp3DVyOHuwtTdjzZGxhwXYWDc/hoOqa/MoJW0iMmMUVrZQWNnKF6+bjz3ARniwnUeum8+ev7mRn3x2FWHBdv7Xy0XkfXcXP3y7hOYudaiT6ae4pp2DZc08mJc++r5sTv5Sz+Y01Izk/nlt7Cttov2i9oaSiSuotJbFT7hzZMlWMGywaIsHovJTaXnW74ELY7f0z8uIpaypm9p21Yt7ip/8FhYR8bz/t6eMqNBAPrM29bLz9gAbH89J4bWvbuClx9azKjWKH759hrzv7uJbrxynvKnbRxGLuN8z+yoIDQrgvnXzxh7YXg2t5f6zNBKsbnbADRHV9A+a7DqlJZIycQUVLYQFBbAkOWJiF57eBqnrIXQS2wRMVasfsBqvHH1xzGF5GVZdm2bbPEdJm4jMCGWNXbx1sp4vrE8jNMg+4hjDMFi/IJanH1rL2/99I3evmsPv8qvY9P3dfPlXBRRWDnvSWHsMfvUpqCr00ncgcvUaOnt4/WgN96yZy+xZgWMPrnTWs23wfGCuCo2BqDRSLpaQFBnCm+oiKZOQX9HK6rTo8Weah2s7D/XHYfFtngvMHyUtg7lroeBZMEev+85KiiA6NJD92q/NY5S0iciM8J/7ygkMsPHAtekujV+YEMF3P72Cfd+8ia/etJD3y1v49BMH+dTP9nHilf/A/M/NcHYnFP2XZwMXcaNfv3eOfoeDhzaM04AEhurZZkPiMs8HNhHJORi1R7ltWRJ7Shrp6h3wdUQyhXT09HOqrmPi9Wwl26zjTKlnG27Nw9B85sMHOSOw2QyuzYjl4NkmzDGSO5k8l5I2wzAqDMM4bhjGEcMwCobOxRiG8ZZhGGeGjtFD5w3DMH5sGEapYRjHDMNY7clvQERkPE1dvfyhsIpPr55DfETwhK5NiAjhr29ZzIFvbuK7d6TyVy3/RPaRf+IgK2gLz8AxTlctEX/R0z/Ib96rZHNWAvPjwsa/4FI9W4Dng5uIlJXQWs7HMkPpG3DwzqkGX0ckU8jhylZMcxL7s5Vsh+j5ELfIM4H5s6V3Q/Bsa7ZtDHkZcdS091DZfMFLgc0sE5lpu8k0zZWmaeYO/fubwE7TNBcBO4f+DXA7sGjov8eAJ9wVrIjIZDx/oIL+QQdfvH7BpN8jtP4D7iv8PNebBZxc/g2+F/2PvNmWRm/V0TGXjIj4i9eO1NDc3ccj47X5B+iogZYy/6pncxraO2pV4DniwoPVRVImpKCilQCbwcp5Ua5f1NcN5e/C4tthIlsETBdBoZBzH5x8DbpHX/6YlxELwP6zav3vCVezPPIu4Lmhr58DPjns/POm5T0gyjCM5Ku4j4jIpF3oG+D59yq5eUkiGfHhE38DhwP2/RCevQ0MA+ORHSz59P/kj49fh5G8nFkDHXTUV7g9bhF3Mk2TZ/aXk5UUwbULYse/wLk/W5of1bM5JVsdJAPqjnLbskTeOdXIxT5t0yGuya9oYVlK5Ki1zSM6+47VjCNzhtWzDZf7MAz2wZHfjDpkflwYSZEhHFAzEo9wNWkzgR2GYRQahvHY0LlE0zRrAYaOCUPn5wDnh11bNXRORMTr/lBYRduFfh7bOIlZtu4meOEeePt/Q9ad8OV3Ye4awGpacm3eTQDs3vO2O0MWcbv9pc2cquvk0evmu7aZcMVeazlU0nLPBzdRYXEQORdqj3L7smQu9g+yp0RLJGV8fQMOjpxvI3cyrf6DZ1vt72eqhCVW58zCX466usQwDPIWxvLe2WYcDq1AcTdXk7YNpmmuxlr6+LhhGBvHGDvSX4Mr/j9nGMZjhmEUGIZR0NjY6GIYIiKuG3SY/OfeclbNiyI3bYL1C+V74YkN1vHOH8A9z8Gsy5fTpGevxYHBuZOH6OzRflHiv57ZX05ceBCfWJni2gUV+6wPqP5Wz+aUnAO1R7hmfgzRoYFsLVIXSRlfUU07vQOOidWzORxQsgMWboaAcTquTne5D0PLWWup6CjyMuJo7u6jpKHTi4HNDC4lbaZp1gwdG4CXgXVAvXPZ49DR+ZirChi+CdJcoGaE93zKNM1c0zRz4+PjJ/8diIiMYltRHedaLvDljQtcm10AcAzCO/8Kz38CgiPgSzth7aMj1zEEh9M/ez4Zg+U8f7DSvcGLuMnZxi52nWrg/vVpBNtdSMI6aq0PZv7U6v+jUlZCcyn2/i5uyU5i58kGege0RFLGll9ubduyZiKdI2sOQ3eDVc8202XfBSFRUDh6Q5JLdW1q/e924yZthmGEGYYR4fwauAUoAl4DHhwa9iDw6tDXrwEPDHWRXA+0O5dRioh4i2maPPXuWdJjQ9mSneTaRR018NwnYM93YcV98NjucZeHBc/NYU1wFf+5t4xutR4XP/Ts/nKC7DbuX5/m2gWX9mfzwyYkTkPNSKg7zu3Lk+jqHWDfGTU/kLHlV7QyPy5sYl2ET28FIwAW3uy5wKaKwFmw8nNw8k/QNfIquZSoWcyPC+OgmpG4nSszbYnAPsMwjgKHgDdM09wGfBfYYhjGGWDL0L8B3gTKgFLgF8BfuD1qEZFxHCpv4WhVO49ev4AAmwuzbGfegievg5oP4JNPwt1PQLALjUuSlpMwUMvAhXZ+/Z5m28S/tF3o478Kq/nkyhTiwl38oFqxF4IjIWmFZ4O7GkPNSKg9Sl5GHBEhdm20LWNyOEwKK1smvlS+ZDvMW29t7C6w5iFw9MORX4865NqMWN4va2Fg0OG9uGaAcZM20zTLTNPMGfpvqWma3xk632ya5mbTNBcNHVuGzpumaT5ummaGaZrLTdMs8PQ3ISLyUU+9W0ZMWBD3rJk79sDBftjx9/CbP4OIZGt2beVnXb/R0Afbz8xr5xd7y9TFTvzKi4fOc7F/0LU2/04V+/27ng0gIhHCk6D2CEF2G1uyE3nrRB19A/qQKCMra+qi9UI/ayfShKTtPNQfh8xbPRfYVBO/2OoqW/hLq95vBHkZsXT2DnC8ut27sU1zV9PyX0TEL52p72TnqQYeuDaNkMAxPni2VsIzt8GBH0Puo/DFtyE+c2I3S1wGwAPzO2jq6uPFQ+euInIR9+kfdPDcgQo2LIwlKynStYs666D5jH+2+v+olJVQexSA25cl09EzwMEy1dHIyPIrWgHInUgTkpJt1jFT9WyXyX0EWiugfPeILzu3FVHrf/dS0iYi084v9pYRbLfxwLXpow868So8eT00lVidIT/2A2u9/kRFJEFoHGl9Z1m/IIYn95ylp1+zbeJ7W4vqqOvo4dEJzbLts47+XM/mlJxj/fz2dXP9ojjCggLYpo22ZRT5FS3EhgUxPy7M9YtKtkHMAohb5LnApqIlH4fQWCgYuSFJbHgwWUkRHFTS5lZK2kRkWmno6OGVD2q4J3cuMWFBIw/a8z343QMQtxC+sheWfnLyNzQMq1lJ3XH+ctMiGjp7+X3B+fGvE/Eg0zR5el85C+LCuDEzYfwLnCr2+X89m1PySjAdUFdESGAAm5Yksr24XnU0MqKCilZy06Nd7yTc22W1ts+8feTuwTOZPdhqSHLqDWt2fgR5GXHkV7Soq6sbKWkTkWnllwcq6Hc4+OJ1o2ym3X8R9v8QFt8BD2+D6PSrv2nScmg4ybXpkeSmRfPE7rOqrRGfOnyulaPn23h4Qzo2VxrxOFXut5ouBNg9F5y7ODtIDi2RvGNZEi3dfRwaausu4lTf0cO5lgsTq2crewcG+2DxbZ4LbCpb8zCYg/DBr0Z8OS8jlt4BB4cr27wc2PSlpE1Epo2u3gF+/V4lty1NIn20JTClb0NfF6x7DOyjzMRNVNIKGOzDaD7D1zYvoqa9h/86XOWe9xaZhKf3lRMZYufT4zXiGa6z3lpuOBWWRgJEpkBYPNQeAeCGxfGEBNq00bZcoeBSPdsEkrbT2yB4Nsy71kNRTXGxGTB/IxQ+b+1v+hHXLIghwGao9b8bKWkTkWnjt/nn6egZ4EsbR5llAyj6I4TGQfr17ruxcy+3uuNsXBRHztzZ/Hx3Kf1apiU+UNV6gW1FdXz2mnmEBk1gxqxyCtWzgbVkLTnn0kxbaJCdmxYnsK24DofD9HFw4k/yK1oICbSxNMXFhjwOB5zZDotuhoBAzwY3la15GNrPwdldV7wUERLI8jmz1YzEjZS0ici0MDDo4Jl95axNj2b1vFG6g/VdsArLsz/h3uVfsQvBHgJ1xzEMg7/cvIjzLRd59UiN++4h4qLnDlRgGAYPjtWIZyQV+yEoApJyPBKXRyTnQMNJa9kzcPvyZBo7eyk81+rjwMSfFFS2sCo1msAAFz/21hyG7kZ1jRxP1ses2e5RGpLkZcRy5Hwb3b0DXg5selLSJiLTwhvHa6luu8hjGzNGH3RmO/RfgKWfcu/NA+yQkA11xwDYlJXA0pRIfvZOKYN64i9e1NU7wEuHznPH8mRSoibYDbVi39SpZ3NKXmnV1dSfAKyfvaAAG9u0RFKGdPUOcKKmg7UTafV/eisYAbBws+cCmw7sQbDy89bD0I4rH1LmZcQx4DA5VKE6U3dQ0iYiU55pmjz1bhkL4sPYnDVGp7zilyEswdo42N2GOkhimhiGwdc2LaS8qZs/HdNsm3jPHwrO09k7wCMb0id2YVcDNJ2eOksjnS41I/kAgPBgO3kLY3nrRD2mqQcmAh+ca8VhTrCerWSb9QAjdALXzFRrHrQenBy+siFJbno0QQE2tf53EyVtIjLlHTjbTHFNB1+6fsHonfJ6u6BkB2TfBbYxNtyerKTlcLEVOqoBuCU7icWJEfxkV6nqa8QrBh0mzx6oYPW8KFaNtkR4NJf2Z3Njrac3RM2DWdGX6toAtmQncq7lAiX1XT4MTPxFfkUrNgNWp7n4M9F2DuqLIFNdI10SswAW3ASHr2xIEhIYwOq0KA6oGYlbKGkTkSnvqXfLiAsP5u5Vc0YfVLINBi7CMjcvjXRy7mtVdxwAm83gq5sWUtrQpW524hU7T9ZT2XyBR0fb7mIslfshKPzDmaupwtmMpObIpVM3L0kE4K0T+rkTKKhoITslkvBgF5f9lmy3jotVz+ay3IehowrOvHXFS3kZcRTXdNB2oc8HgU0vStpEZEo7VdfBnpJGHspLIyRwjBm04pchIhlS13smkMRswLiUtAHcsTyZjPgwfrLrjGbbxOOe2V/OnKhZ3Lo0ceIXT8V6NqfklVYzkoFeABIjQ8hJjeKtE/U+Dkx8rX/QwQfn2shNm0ir/60QkwFxizwX2HSz+A4IT4TCKxuSbFgYi2nCe2VaInm1lLSJyJT21LtlzAoM4P71aaMP6umwngBmfxJsHvq1FxxhLRMZakYCEDA023aqrpO3TuoDpHhOcU0775W18GBeGnZXO+Q5dTVC46mpV8/mlJwDjn4rcRtyS3YiR6vaqWvv8WFg4msnajq42D/o+qbavZ1QsVezbBMVEAir7oczO6D98j1KV8yNIjQoQK3/3UBJm4hMWbXtF3ntSA33rk0lKnSMjbJPb4XBXlh6t2cDcjYjGebjK1JIiw3lJ7vOqDGCeMzT+8oJDQrg3rXzJn5x5X7rONXq2ZxSVlrH2g+XSG7JHloiqYclM1r+UNfCXFc7R559Bwb7VM82GasfBNO0atuGCQywsW5+DPtLVdd2tZS0iciU9cv9FThMk0evmz/2wOKXIXIuzF3r2YCSlkNrhTWzN8QeYOPxmxZSVN3B7tONnr2/zEgNnT28frSGz+SmMnvWJDYCrtgHgWFTr57NKXo+BM++rBnJooRw0mNDtURyhiuoaGVeTCiJkSGuXVCyHUJmW0uFZWKi06wtEg4/D4OX78u2ISOOs43d1Hdo5vtqKGkTkSmps6efF94/xx3Lk0mNCR194MU2KH0blnpwaaSTsxlJffFlp+9eNYe50bP40U7Nton7/fpgJQMOk4fy0if3Bpfq2SaR8PkDw4DkFZc1IzEMgy3ZiRw820RnT78PgxNfMU2TgsoW12fZHA5rL8+FN0/dnwVfW/MwdNZa/3cc5tqMWAB1kbxKStpEZEp68dA5OnsHeGzjOJ3yTr9p1bu4e0PtkSQtt44fWSIZGGDjz2/M4Mj5NvZpiYi4UU//IL9+/xybsxJJjwub+Bt0N0Hjyalbz+aUnGM9LBn8MEHbkp1E/6DJnhLNcM9EFc0XaOrqc72erboQuhshU/Vsk5Z5m9Xwq+DyhiTZyZHERwTzxrFaHwU2PShpE5Epp2/AwTP7Kli/IIYVc6PGHlz0R2svpzmrPR9YRBKExl3WjMTpz9bMJXl2CD/WbJu40atHqmnp7ht/ifBopno9m1PySqtutfHUpVNr0qKJCQvSEskZKr/cqmdb6+pMW8lWMAJg0c0ejGqaC7DDqi9Yq1taKy+dttkM7lkzl12nGqhpu+jDAKc2JW0iMuX86VgNdR09fHljxtgDL7RA2TtWAxJjlE233ckwIGnZFTNtAMH2AL5yQwb5Fa28V9bi+Vhk2jNNk6f3lbMkOZL1CybQ0nw4Zz2bs5nHVHWpGcmHdW0BNoNNWQm8c6qB/kGHjwITX8mvaCE6NJCM+HDXLji9DeZda23WLpO3+gHrb+Hh5y47/dl18zCB3+af901c04CSNhGZUkzT5Kl3y8hMDOfGxfFjDz71J3AMeL5r5HBJy63W44NX1tHcuzaV+IhgfrLrjPfikWlrX2kTJfVdPHrdfIzJPpSo2Afzrpn6NTwxGdbm4MPq2sDqItnRM8Chcj0omWkKKltZkxbj2s9G2zloKIbF6hp51aJSYeEW+ODXl/0dTI0J5fpF8fw2/zwDeogyKUraRGRKefdME6fqOvnS9QvG/2Nc/LLVWS7Zi7MISSusZVpNVyZmIYEBfHnjAg6cbaagQh8i5eo8va+cuPBgPp6TPLk36GqAhhOQtsG9gfmCzWb97A2baQO4flEcwXablkjOMI2dvZQ3dbu2NLK9GrZ+w/pa9WzukfswdNVbNeXDfG7dPOo6enhHnZQnRUmbiEwpT717loSIYD6xMmXsgd1NULbHe0sjnUZpRuL0+WvSiA0L4se7Sr0Xk0w7pQ1d7D7dyBfWpxFsD5jcmxT90Tpm3em+wHwpOcf6uRvWbjw0yM71i+J460S9aklnkMJK5/5sYywb7u+BPd+Dn+ZC6U7Y/A8Qt9BLEU5zC7dA5JwrGpJsXpJAQkQwL7xfOcqFMhYlbSIyZRRVt7O/tJmHN8wf/4PqydfBHIRlXugaOVzsIggIHrEZCcCsoAC+tHEB75Y0cuR8m3djk2nj2f3lBNltfH79JDbTdjr6ojU7lbDEfYH5UspKGLgIzZfPcm/JTqS67SInajtGuVCmm/yKVoLtNpbNibzyRdOEE6/Bz9bCO/9s7S321UNw/V97P9DpKsBubbZd9g60lF06HRhg4961qewuaaSq9YIPA5yalLSJyJTxo51niAi287lrXPigWvwyxC6ExGWeD2y4ADskZo860wZw//o0okID+alq22QSegcGee1IDR9bkUxcePDk3qTxNNQegZz73BucLzk3B//IEslNWYkYBloiOYMUVLSQkxp15cO9+hPw/Cfgd1+waiAfeA3u/TVEp/skzmlt9ResbpyFlzckuXdtKqCGJJOhpG0mGOizimzPvW/tQyIyBRVWtvLWiXoe27iA2bPGaZrQ1QAVe6292by5NNIpabmVtI2yHCs82M6jG+bz9skGiqrbvRycTHX7S5vo7B3g4yvGWSI8lqMvWR+olv2Z+wLztbhMsM+6ohlJfEQwq+dFK2mbIS70DVBU03F5PduFFnjjf8CTG6D2GNzxH/DlvbDgBt8FOt1Fplj7th35jfU5dMjc6FBuWpzAb/PPq6vrBNl9HYBcBdOEnnZr9/mOmqFjLXTWXH7sbgScHx4N+MsPIGaSe/qI+IBpmvzbtlPEhQfziCv7UZ18DUyHd7tGDpe0Ag4/b/1czp4z4pAHN6Tz1N4yfrqrlCe/sMbLAcpU9saxOiJC7GxYGDe5N3A44PjvIWMTRCS6NzhfsgVYD0w+MtMG1hLJ7249RXXbReZEzfJBcOItR861MegwrXq2wQEofBbe+Y71eSn3Ebjpf0HoJLfIkInJfRhOv2F1ch5WqvC5dfP44vMF7DzZwG3LknwY4NSipG0qME2aDv2e6PYiAjprL0/S+kdYEzwrxnrCEZFsLReJSIHIZLCHwMtfhpJtsP7Pvf99iEzS7pJGDpW38O27lhIW7MKvraKXIW6x72p1hjcjGSVpiwwJ5OEN8/nxzjOcqusgK2mE2guRj+gbcPDWiTq2ZCcSZJ/kYpnK/dB+Hm7+R3eG5h+Sc6xaPYfD6ig5xJm0vX2ingfz0n0Xn3hcfkUrhgFrKYanvgX1Rdbm8bd919pHU7wnYxPMnmclzsOSthsXx5M8O4QXDp1T0jYBStqmgNMHX2fxji/Rjx1HRDKBUXMgeYU17RyZbCVnziQtIhkCQ0Z/s70/UNImU4rDYfLv204zLyaU+9a6UMvWWWd9KL3xm75ZGgmQuNQ61h0fc9+fRzak8/TQbNtPP7faS8HJVLb/bBMdPQPcuXySbf4Bjr0EQRGw+A73BeYvUlZC/i+g5SzELbp0OiM+nAXxYbylpG3aKz97kufDnyT8xf1WwvCZ52HJJ3z392AmswXAmgdg1z9D81mIzQDAPtSQ5Idvn+Fc8wXmxYb6ONCpQTVtfs7hMOl450c0M5vrjOdY1vo9fpn1JI5PPwu3/QvkfQ2W/xmk5VlLHsdK2AAyb4WK/dCjLloyNbx+rIaTtR389S2Zrs0snHgVMH23NBIgOAJiFozaQdIpKjSIB/LSeeN4LdVtF70UnExlbx6rJSLYznWLJrk0sv8iFL8K2Z+AoGn4QcnZjOQjdW1gzba9V9ZM+8UrN76XaaDvAo6d/8x3ax7hmoECaxnkVw9B9l1K2Hxp1RfAZrdm24a5d20qNgNezD/no8CmHiVtfm7Hu++ytr+Axqz7ef3rN5OXEcs/vn6CB589RF17z8TfcPHt4OiHszvdH6yIm/UNOPj+jhKWJEe63nSh+GVIWArxiz0b3HiczUjG8enVczBN2KPNRmUc/YMOdpyo5+bsxMnvzXbqDejrhBX3ujc4fxGfBcGRULb7ipduyU5kwGGy+3SD9+MSzzFNOP4H+Gkutr3fY4cjl923bIUb/hYCVb/ocxFJ1mfPIy/AQO+l08mzZ7EpK5HfF5ynb0ANSVyhpM2PdfUO0PXuz+gjkMUf+ysSIkN45qG1fOfuZRRUtHLL/93Dq0eqJ/amc9fBrGg4vc0zQYu40W/zz3Gu5QJ/e9tibDYXnpS2V8O5g76dZXNKWg6t5ePOamfEhzMnahZ7SvRBUsZ24Kw1S3T71dSAHPuttelt+vXuC8yfBARam4Wfev2yjnUAK1OjiQsPYoe6SE4ffd3w3Mfhvx6F0FjezH2Gv+z/Gsuzs30dmQy35mG40GztnzrM56+ZR1NXnzq7ukhJmx97+q1C7hjcTWfm3RjhCQAYhsHnr0lj63+7noUJ4fy3l47w1RcO03ahb5x3GxJgh0W3wJkd4Bj0YPQiV6e7d4Af7Sxl3fwYbsyMd+2iE69aR79I2lZYx/riMYcZhsHGzHj2lzar/bGMaevxWsKCAtjo6s/DR3U1QOlOWH7PZU06pp2ln7I6BZa9c9npAJvB5qxE9pxupHdAf/+mhcPPW9u73P49eGw3b7TPZ07ULJJna4bNryy4CcKTrJn+YTZmxjMnahYvHKr0UWBTyzT+rT21nWu+QP/7zxJq9BK7+a+ueD09Lozfffla/ubWxWwrquPWH77LnhIXl1dl3gYXW+D8ITdHLeI+z+4vp6mrl2/cloXhaj1C8R+tGa64hZ4NzhXDO0iO44bMeLp6Bzhc2erhoGSq6h90sL24js1LEgkJnOTSyKL/AnNwem2oPZIFN0JIFBT98YqXblmaSFfvAO+VtXg9LHGzwQE4+HOYdy1c8ximYSO/ouXy/dnEP9hs1s9l+R6rs+uQAJvBfWtT2V/aTEVTt8/CmyqUtPmpf33jGPfbttM77/oPO9F9hD3AxuM3LeSVxzcQGRLIg88c4h9eLeJi3zhPEBdutopCS7Z6IHKRq9fa3cf/21PGluxE1qS5+Ae47RxU5VtP2f1BRDKExo7bjAQgb2Esdpvh+oMXmXHeL2uh9UI/d1xN18ijL1mNOny1FYa32INgycfg9JvQf3nt94aFccwKDOCtE3U+Ck7c5uRr0H4Orv0qAOdbLtLQ2Wvtzyb+J+Mma4lkfdFlpz+zNpUAm8GLh9SQZDxK2vzQgbNNBJ56nSSjheDrvjbu+GVzZvP6167j0evm8/zBSu788V6OnG8b/YKQ2ZC2AUq2uzFqEff5+e5SuvoG+JtbJ9BMpPgV6+gPSyPB6lbmYjOSyJBAVqdFK2mTUb1xvJbQoABuXDzJpZENp6D2CKyY5rNsTkvvht6OK5puhQQGsDEzjrdPNGCapo+Ck6tmmnDgJ1aX3sW3A5BfYc2erlXS5p/m32AdP9IkKDEyhJuXJPD7wiotWx6HkjY/M+gw+fZrxXwleBuOmIWwcItL14UEBvD3H8vmhS9eQ0//IJ9+4gD/962S0WtkFt8OjaegpdyN0YtcvZq2izx3sJJPrZpLZmKE6xcWvwwpq6ytL/xF0nJoOGkt4xnHDZnxFNd00NA5ia6wMq0NDDrYUVzHpqyEyS+NPPYSGAHWFjEzwfwbYFaM9XvhI7ZkJ1HX0cPx6nYfBCZuce4g1ByGax+39gIDCipbiAyxsygh3MfByYgik63urh+pNQX43DVptHT3sb1YDUnGWhpbOAAAIABJREFUoqTNz7yUf47QhsNkm6XY1n9lwsXieQvj2PpXG7krJ4Uf7TzDp584QGlD15UDM2+1jiXqIin+5UdvnwETvr5l0fiDnVrKrT/g/jLL5pS0AgZ7ofnMuENvGGousbekydNRyRRzqLyF5u6+yW+o7XDAsd9DxiYYamo17QUEwpKPw+mt1t50w2zKSsBmoI51U9mBn1pJec7nLp3Kr2glNz3GtU7D4hsLboTKg1csW75+YRypMbN44X01JBmLkjY/0n6xn+/vKOFvZ7+NGTIbcj47qfeZPSuQH9y7kic+v5rzLRe488d7+eX+chyOYUtBYhZA3GLrD5qInyht6OT3hee5f30ac6MnsPHvCT9bGumUuMw6urBEMjs5krjwYC2RlCu8WVTLrMAAblw8yYSrch90VE3/BiQf9f/ZO+/4qKr0/7/vTHonHZIQSEhIoZMQikAEaYKCXbGtChYsu2vZ/tum67bv6ror6lrBtesKoiBKDb2EnhBIr6SH9D5zf3+cJBJaZiaTzCRz3q8Xr0vu3HPmgVfmzn3O+TyfZ8zN0FoPGVu6nfZ2dSBuhLdM2gYqlVmiXjH+oa4G8VUNrWSW1RMnTUism7Brob0JCrsb4Wk0CndNGc6B7KrLbzRIACOSNkVRtIqiHFMU5ZuOn9coipKjKMrxjj8TOs4riqL8S1GUTEVRTiqKMqmvgh9s/GtbBi6NRSS07EOZ/CNw7N0W/6KxQ/nuJ7Ou3JB79ELI2yuskSUSK+D/vkvHxcGOx68NN25gypcQFAdew/smMFPxjQCto0FmJBqNwqxIX3ZnlKPTy1obiUCnV9mcUsqcKH+cHUyURp74FBzcYfT15g3O2gm9Blx8havsRcyPCeBMSR0FVY0WCEzSK/avFjup8Su7TiXLeraBwYgZQqZ9UV0bwG2TQ7CThiRXxZidth8DaRede05V1Qkdf453nFsERHT8eRh4vfdhDn6yyutZuy+XF4P2o6DAlIfNMm9nQ+4XbxpLcu55Fr2yi7LajsQtchHo20XfHonEwhwvqGZzagkrZ4bh4+Zo+MDKLJEUjbES18gL0doLpz4DdtpASCTPN7bJWhtJF4dzq6iob2HRWBMbarc2iv6FMUu7diVsBq0dxNwoTLdau9uJz4sJAJCNtgcaDZVw/EMYdwe4B3SdTs47j4NWw9ggTwsGJ+kRR3cIjr9s0ubn7siC2ED+d7SQ5jZpSHI5DEraFEUJBhYDbxtw+VLgfVVwAPBSFKUXHsW2wQvfnMbHvpVrajeKL1fPYLPNrSgKyxOGs/7xGdS3tPPy1o76mpApQhMu69okFkZVVf767Rl8XB14aKaRRiKdq+gxS80fmDnodJA0wKluZoQfigJJZ6VEUiLYdKoYJ3sN15oqjTy7CVrrYPwd5g1soBB7M7Q1XuKWHOrjSmSAm7T+H2gkvwPtzV02/50czq1iXLCn6UY9kv4jLBHOHYOmS/uSLk8YTnVjG9+mFPd7WAMBQ3fa/gn8DLjYivBPHRLIlxVF6VwaDwIKLrimsONcNxRFeVhRlGRFUZLLy237AWXn2TJ2nC3npchUNK11wg2pDxgd6M7dCaF8ejifjNI64bgUMR8yvjfI3U4i6St2Z1SwP7uSJ+eMws3RzrjBqeshZKpZFzrMSuA40ZumrucvIW9XB8YFe5GUXtYPgUmsHb1e5duUEhIj/XE19nPRyclPwSNYSAVtkdDp4BZwBRfJAA7nnqe6sdUCgUmMpq0ZDr0pXLX9o7pON7XqSCmqkf3ZBgphiaDqIWf3JS9NC/NhhI8LHx2UEsnL0WPSpijKEqBMVdUjF730SyAKiAe8gZ93DrnMNJcsMauq+qaqqnGqqsb5+ZnYd2YQ0KbT8/w3pwn3cWJ6xecQPAWC4/rs/Z6aG4Grgx1/+faMODF6oVjtKDzcZ+8pkVwNvV7lr5vPEDzEmbsSjKxJK08XjTqtzYDkQgLHiqMREsnjBdXyQVJCct55yutauH6ciWKV+jIhfx93m9FOxIMGjVbswmd8Dy3dDQ7mxQSi06tsPyMXSQYEJz+FhnKY3r1/7YnCatp0KvHShGRgEBwHDm6XlUh2GpIczj1Pemld/8dm5RhyF58B3KgoSi7wCTBHUZQPVFUt7pBAtgDvAVM6ri8EQi4YHwycM2PMg4oPDuSRVd7ASxNKUM7nwNTH+vT9vF0dWHXtKLadKWN/ViWEzwWNHaRLF0mJZdh4qpjUc7U8PS8SRzsjpS2p6wDFeqWRAAGx4miAGQmIpE2vwp5Maf1v62w6VYyDnYY5USZKI099AarOdhpqX4nYm4Sk7qJSgHFBnvi7O0oXyYGAXi8MSALHwshZ3V7qNCGZHCqTtgGB1h5GXHPZpA3g1snBOGg1crftMvSYtKmq+ktVVYNVVR0B3AlsV1X1ns46NUVRFGAZkNIxZANwX4eL5FSgRlVVKU69DFUNrby8JZ2ZEb6MK/pISFiib+zz931gxgiGeTrx4qY09A7uEDoDzsq6Nkn/06bT84/vzxIV6M7SCZeoqHsm9Ushf/Kw4rJZJw8YMtLgnbbxwZ54OtvLujYbR0gji0mM9DNeMtzJyU9g6PhuUjKbJGQquA+9RCKp0ShcFxNAUnq5ND6wdjK3QsVZmPYkKN0FXYdzzxMZ4IaXi4OFgpMYTVgiVGVB9aWJmY+bIwvGBPKlNCS5hN7oJT5UFOUUcArwBV7oOL8JyAYygbeAVb2KcBDz8pZ0Glp1vDBVRcndAwkPC7erPsbJXsuzC0ZzqqiGr0+eg9GLxM2wKrvP31siuZBPDxeQW9nIcwtGozW2IWpZGpSfsW5pZCedZiQGYKfVcE2EL0np5agGmJdIBifHCs5TWtvC9aY21C47A8Un5C4bCGlozDLRr625tttL82ICaGzVCeWJxHrZ9y9wH3aJS7BOr3I077ysZxtohCWK4xV225ZPGU5tczvfnJR7PhdiVNKmqupOVVWXdPx9jqqqY1VVHaOq6j2qqtZ3nFdVVX1cVdXwjteT+yLwgc7Zkjo+PJjHPQnDCU1fC/auMOn+fnv/ZROCiB3mwd82n6UlbH5HUHK3TdJ/NLXqeGVbBnGhQ0yTf6WuA0XTL7vTvSZwnFgUaTFMoz870o+yuhbOlEhNv62y8WQJDloNc6NNlEae/ET0Qxp7q3kDG6iMuRl0LXC2eynA9HAfXB200vrfmik+Abm7YeqjQlrXQeH5RlasPUxdSzvTw30sGKDEaPyiwC3wiknb1DBvwvxc+ehgXv/GZeXYaGWyZVFVlT9+k4q7kz0/neoJKV/AxLvB2avfYtBoFH51fTRF1U2sPYP4AMm6Nkk/8t6+HMrrWvj5oigUxchdNlUVDbVDZ3Tr1WO1dJqRlKYadPnsSGHOlJQuJZK2SKc0clakL+5O9j0PuHQCOPkZjJoLbiYmfYONoDhRgnBRo21HOy2zR/uxNa0UvWxqb53se1UYV3QsbLfr9Ly9O5t5L+3iYE4Vv1kczfVjrFgiL7kURRG7bdlJ4n51ycsKy6cM52h+NWnFtZe8bqvIpM0CbDldyt7MSn56XQReqe+Drg0SHu33OGaM8iVxtB+vbs+kOWwe5O2DZtnUV9L3VDe28vrOLOZG+RNviqylNBUqM6yzofblMNJBMsDDiahAd1nXZqMcL6ymuKbZdGlk7m6oLRINiCUCjQZilwk3zabqbi/NiwmgvK6F44XVVxgssRg1hZDyP5GwOXtxqrCGZa/t5YWNaUwP92HL07NZMTMMjbHyeonlCUuExgoou/xi5i2TgnGwk4YkFyKTtn6mpV3HnzalMcrfjbvjAkSjyMiF4BNukXh+uSia+pZ2Pq6OBX27KPaVSPqY15OyqG9p57mFo02bIPVLIf0aCNJIAI9hopG9gQ6SALNH+5GcV0VDi+yhaGt8e6oYe63C3GgTd5FPfgoO7hC12LyBDXRibwZ9G5zZ2O30taP90WoU6SJpjRx8A4DGSSv549enWbp6D2W1Lbx29yTevj+OIC9nCwcoMZmw2eKYteOyLw9xdWDx2KGsP1ZEY6v8HgSZtPU7a/bmklfZyP9bEoN96hei6e40y3m1jA505/a4EP6S4o7OaQikf2exWCS2QUlNM2v25nLThCCiAj2Mn0BVRT3byFng6mv+APsCRTHKjASERLJNp0qDBBtDVVU2nSphZoQfns4mSCNbG+H0V6INhr18oO1G0CTwGn6Ji6SXiwNTRnjLpM3aaK6FI2spDV7AvHdzeHdvDssThrP1mdlcP3ao8bJ6iXXhMQx8R1+xrg1gecJw6lra+fqE7BwGMmnrV8rrWvj39kzmRvkzO8IX9r8GAWNgxEyLxvX0vEg0GjuS7eNEA1KdXNGQ9B2vbEtHr6r8dF6kaRMUnxCmHgPBNfJCAsdC6WmDP19xod64OGhlXZuNcbKwhqLqJhaNCTRtgrOboLUexktp5CUoirhvZO+AxqpuL82PDSCzrJ6cigYLBSe5mLr970JLLSsyp+PqqOV/j03jhWVj8TClzlNinYRfK0pz2lsu+3Jc6BAi/N2kRLIDmbT1I//33Vla2nX8enG0WFkoT4Opqy7pOdLf+Hs4sXJWGGsro6HpPBQesmg8ksFLVnk9nyUXcndCKCHeLqZNkrpONISPvsG8wfU1geOEe11lhkGXO9hpmB7uy870Mmn9b0NsOlWMnUZhfoyJSduJT4ThRug15g1ssBB7sygFOPNNt9PzYoQUdcvpEktEJbkAvV7lw31Z1CX9m0NqNAvmLeSbJ2cyOVTa+g86whKhvQkKLv/cqSgKyxOGc6KwhpQi6bkgk7Z+IqWohs+OFHD/tBGE+bnBgdfA1Q/G3GLp0AB4ZFYYqc7xtGOHela6SEr6hpe+T8fRTsMTc0aZNoGqinq2sERwGWBf4EaakYCoayuoaiK3srGPgpJYE6qqsimlmBmjfPF0MWE3oa4UsrbDuNuF8YbkUoaOF83uU7q7SAYPcSF6qIeUSFqY9NI6bvvPfg588y7DqCB0yc94Yk4EDnby93lQEjpD1KdfRSJ588RgHO00fHRI7rbJT0E/oKoqf/z6NENcHHhybgRUZAgZYvwKsHeydHgAuDra8fD88ezXRdFw6pueB0gkRnKysJqNp4pZMTMMXzdH0yY5dxSq88Vq+UDDNwK0jsaZkUR0WP+fLeurqCRWREpRLQVVTSw21TUy5QtQdTBeNtS+IooiXGdzdkFDRbeX5sUEcCTvPJX1l5dqSfqO5jYd//fdWRb/azfZZXX80W8Hqs8oAiYvs3Rokr7EyQOC44Rk+Qp4utizZNwwvjpWRL2NG3PJpK0f2HiqmEO5VTw7f7QoLD/wOmgdIO4hS4fWjTviQjjhMg23umzayjMtHY5kkPG3zWfxdnVg5cyRpk9yZqNYlYu63nyB9Rdae/CPgpIUg4cM93FhpK+rrGuzETalFKPVKF1SPaM58QkMnQB+Jrqy2gqxN4nkNm1Dt9PzYwLQq7DtjFwk6U/2ZVaw8J+7eHVHJjeMH0bSHQ4MqUlFmfa43DG2BcIS4dwxUZ5zBZYnDKehVceG47ZtSCI/DX1Mc5uOP286Q/RQD+6IDxHFzyc+hrG3g5ufpcPrhp1Ww4S5YoX2xNaPLRyNZDCRnFvFnswKViWGm9YsuJPsnWJVznmI2WLrVzodJI2oUZsd6cf+7Eqa23R9GJjE0gjXyGKmh/swxNXB+AnK0sQurtxl65mAMeATcYmLZOwwD4Z5OkmJZD9RXtfCM5+dYPnbBwH4cEUCL90+AY+j/wEXHxh/l4UjlPQLYdeCqofcPVe8ZNJwL6IC3fnoUF4/BmZ9yKStj3lrVzZF1U38dkkMWo0CR9dCW6NFbf6vxoz4yRTYhaI/u5m65jZLhyMZJLy3NxcPJzuWJww3fZKmarEaN3K2+QLrbwLHiWaidYabHcyO9KO5Tc/h3KqeL5YMWE4X15JX2Wh6Q+0Tn4hd6DG3mjewwUini2TuHqgvu+C0wnUxAezOKKepVS6SmANVVSmpaWZvZgVr9+Xy269SWP7WARJe3Er8n7by1fEiHr82nM0/mcWMUb6ifCT9247yEdmywiYIjgMHt6vWtXUakqQU1XKysLr373l6AyT9/YquldaKnaUDGMyU1jbz2s4sFsYGMi3cB3RtcPBN8dAZEGvp8C6Loig4xi5m4vH/8Pq2kzy1eLKlQ5IMcIqqm9icWsKKa0bi4tCLW07eXrEaF5ZortD6nwvNSDwMezhPCPPGwU5D0tlyZkZY1+68xHxsOiWkkQtiTXCN1Ovh1Ocwaq7VKTisljE3w66/iZ52U1Z2nZ4XE8D7+/PYk1lhukzVBmnT6cmvaiSzrJ6s8npxLKsnq7yhWx2Su6MdYf5uXDPKj1H+blwX7U9EgPsPE+1/VdT+xq+8zLtIBiVae2FIcoUm250smxjEnzed4cMD+Yy71cv091NVSPor6HUw61nT57EAMmnrQ/7x/Vna9Xp+eX2UOHH6K6g7B0tetmxgPeA/eSmceI2cg+spuSaWQE/rMEuRDEze35+LqqrcOy20dxNl7wR7FwiON0dYlqFzsabkJETON2iIi4MdCSO9SUov5zd9GJrEcnQ21J4a5o23KdLI3N1QWwTz/mj+4AYr/tHgFyUkkhckbQkjfXB3tGPL6RKZtF2BxtZ2vkstIbOsIzkrbyCvsoE23Q+y70APJ8L9XbllUhDh/m6M8nMj3N8Nf3fHKzfFbqgQO8bj75SLD7ZGWCJkfCeMxrwur8jxcLLnhvFD2XDiHCtmjuye7BtD9k4oTYEbX7V4yy1jkUlbH5F6robPjxSy4pqRhPq4isx+/2rwDocIwx7WLEZwPDonbxIbj/KP78/y99vGWzoiyQClsbWdTw4VsCA2kOAhJvZl6yQ7CYZPAzsTHmqtBSdPGDLCKNt/EBLJFzamUVTdRJCXlAwNNs6U1JFT0cAKU016Tn4Kjh4Qtdi8gQ12Ym+GnX+G2uKunW8HOw2JUf5sSytDp1dFWYOkG099fIytaWXYaRRCfVwI93NjfkwA4X5ujPJ3I8zP1bTa5cNvQ3szTHvC/EFLrJuwRHHMToJJ917xshUzw/g2pYSFr+zmzvgQfjov0ng36v2vgqu/aI0ywJA1bX2Aqqr8aWMans72PHFthDhZcEjYlU99zPrdkDRatKMXMN/+BOuO5pFWXGvpiCQDlHXHiqhpauOBGb1wjATxUFVxFsIGcD1bJ51mJEYwO1KsOu+SLpKDkm9PFaNRME0a2dooVBwxN8oaIGOJXQaol7hIzosJoLKhlWP5V3azs1V2ni1ja1oZT8+L5PQfF7LtmUTevC+Ony2M4pbJwYwP8TItYWtrgkNvQeRC8Is0f+AS68Y/GtwCrlrXBhAZ4M7OZxO5J2E4nxwuIPHvO1m9I9Nwo67S05C5FRIeBjsTWw9ZECvPHgYm28+UsS+rkp/MjfihQeqB1WKVfaC4IUUuxFlXx0zHbP787RlLRyMZgKiqypq9ucQO8yB+RC/dHnN2iWNYYm/DsjyB46AqG1rqDB4yyt+NYZ5OJJ2VSdtgQ1VVNp4qJmGkj2n9C89shNZ6GCddI43GbzT4x17SaDtxtB/2WkW6SF5Em07P89+cZqSvK4/ODjdvw+sTnwiTJrnLZpsoivh+z94panSvgo+bI39YOobvfzqLaeE+/P27s1z7fzv58mghen0Pzsz7V4Ods9W13DIUmbSZmTadnj9tSiPM15W7p3bU8FTnQ9rXMPlH4Ohm0fgMJnwOaOz56fAsdqWXsztDPixKjGNPZgUZZfU8MGPklWsYDCV7Jzh7Q8BYs8RmUQLHAqpY8TMQRVGYPdqPvZkVtOmu/oUmGVhkdNQEXT/WhF02gJOfgGeIKOSXGM+Ym6DgANQUdZ3ycLJnapgP358uRTWiPcdg57/788gqb+A3i6PNm7Dp9eJheuh4GHGN+eaVDCzCEkXiXpZq0OXhfm68dV8cnzw8FT93R57+7AQ3vLqHfVkVlx9QVwKnPoOJd4OLt9nC7k9k0mZmPj6UT3Z5A7+6Php7bcd/78H/AApMediisRmFkweMuIYxDfsJ8XbmxU1n0PW0giGRXMB7e3PxdXPghvEmWph3oqqQkwQjZ1q/tNgQuhwkTxo1bHakH3Ut7RzLN4PdscRq2HiyGEWBBWNMSNrqSiFrO4y9bXB8NixB7M3ieHp9t9PzYgLIqWggq7zBAkFZH5X1Lby8NZ1ZkX7MifI37+QZ30NlBkx/asAZQ0jMSFiiOPYgkbyYqWE+rF81g1funEB1YxvL3zrIirWHySy7SM1y6E3h4j7VOltuGYK8y5uRmqY2Xt6SzrQwH+ZGd9zUWurg6H8hZil4Bls2QGMZvQhNZQa/n+FEWnEt644V9TxGIgFyKhrYfqaMuxNCcbTT9m6yyizhjBeWaI7QLI9HkGgObmRd2/RRvmg1CknpZT1fLBkwfJtSTPwIb/zdTXDpTflCtMGQDbVNxydcSJYvarR97WjxHS7rSAUvbUmnsVXHb5dE9145cTH7/g0eweI5SWK7eAwD39FGJ20AGo3C0glBbHtmNj9fGMXB7CoW/HM3v1l/ior6FmhtgMPvCLMmn3Dzx95PyKTNjKzekUl1Uxu/XnzBTe34R9BSA9Met2xwphC5AIA5ylHGB3vyj+/PGl7sKbFp1u7LxV6rcPfUXjTT7iRnpzgO5KbaF6IoJpmReDjZM3n4EJLkQ+SgIbOsjvTSehb3pqH20AmiNktiOrE3QeFhUcrQQYi3C2G+rvLzBpw+V8vHh/K5b1ooo/xNtFm/EueOQd4emPqo6NclsW3CEiFvn8lNr53stTyWGM7O54RZyceHhFlJ0mevQHM1TH/SrOH2NzJpMxP5lY2s2ZvLrZOCGRPk+cMLR9ZAUJzo+D7QGDIC/KJR0jfzq+ujKa5p5p09OZaOSmLl1Da38XlyATeMG2ba7sHFZO8UNTveYb2fy1oIHAdlp0HX3vO1FzB7tB8pRbWU15n2hSaxLjadKkFRYKEp0sjS00JiO1DMrayZ2JvEMbW7RHJWpB8HsitterFSVVX+8HUqXi4O/GRuH7g67ntVtKuYdL/555YMPMISoa1RLKL0ggvNSqaHeRGavoYUJYIvy4N6NiuxYmTSZib+uvkMWo3CswsuWPGsLhAPZrHLLBdYbxm9EPL3kzBUy3XRAby+M4vKevnAKLkynycX0tCq673NP4BeBzm7xS7bYKp1CBwr+hFVZho1rNP6XxoDDQ42nSomLnQIAR4mLG6c/AQULYy5xfyB2RreI2HYREjt7iI5e7QfLe16DuVUWSgwy/NtSgkHc6p4Zn7kD27Y5qK6QMhSJ90n6uglkhEzxH0ta4dZpgv3c+PN+FJGaEr52vVWnv78JDeu3sP+rEqzzN/fyKTNDCTnVrHxVDGPzA7r/uWbtU0cR82zTGDmIHIR6Nshcyu/WBRFU5uOf23LsHRUEitFp1dZsy+HuNAhjA327HlAT5ScFJKGsMTez2VNdJmRGCeRjBnqga+bg5RsDQKyyus5U1LHojEmSCMrMuDwu0LC7uZn/uBskdibhVSv6gc1ydSRPjjYaWz289bcpuNPG9OICnTnzngzSN0v5uAb4pjwqPnnlgxMnDwhaLJJdW1XZP+r4BXKz3/yLP+8YwJV9a3c9dYBVqw9TFPrwNpFl0lbL9HrVZ7fmEaAhyMPz7pIvpWxRRTXDuR6g+A4cPGB9M2M8nfjrikhfHgwn+zyektHJrFCtqWVUlDVZJ5dNvjhxj1ylnnmsxZ8I0HrYLSDpEajMCvCj13p5dLNdYDz7aliABYZa/XfUgef3gN2DnD93/sgMhulUxFzgSGJs4OWhJHeNmtG8taubIqqm/jdDbFoNWZWOpSni2baY28FrxDzzi0Z2IRfC+eOQpMZnJILDkHBQZi6Co2dHcsmBrH92UR+vjAKR3stzg69NErrZ2TS1ku+PnmOEwXVPLcgChcHux9e0LVBdhKMmjuwZV0aLUQsEAmorp0fz43E0U7DP7akWzoyiRXy3t5chnk6sSA2wDwTZieBXzS4m2k+a0FrD/7RRu+0gZBsnW9sI6Wopg8Ck/QXG0+VMGm4F0M9nQ0fpKrw1eNQkQ63vjfwHImtGa/hEBx/iYvkrAg/MsrqKapuslBgZqS2WPRDa6nr8dLimiZe25nF9WMDmRbuY944dO2w/lFwcIF5z5t3bsnAJyxRuOLm7un9XPv+LXbvJt7TdarTrGT18km9n7+fkUlbL2hu0/G3zWeJHebBzRODur9YcAha6yBiAEsjOxm9UEjUCg7g5+7IXVOG831qCecbWi0dmcSKSCuuZX92JfdOG4Gd1gy3lvYWyD8w+KSRnXQ6SBrZvPeaUb4oCjYr2RoM5FQ0kFZcy/XGukbu+xec/gqu+wOEDRI3VWsi9iax+12Z1XVq9mghPx3Qu216vbA7Xz0FvvsVbHymxyF//fYMOlXll4uizR/Pvleg6Ags/sfgW5CT9J6gOLB3hexe1rVV5cCZbyDuQXB0M09sFkYmbb3gnT05FFU38evF0Wgulg5kbgWN3eCQdYXPEVKu9M0A3DQpiDadysYOeY9EArBmby5O9hrummImqUvBIWhvGrwPp4HjoLEC6kqMGubj5si4IE+ZtA1gNnVJI41I2rJ3wtbfQ8yyAW9bbbXEdEokfzAkifB3Y6in08BN2srOwHuLYOPTwmxlyiNw8lPRLuIKHMmrYv3xczwyK4wQbxfzxlOaCjv+LP6vpYmO5HLYOQhDkt7WtR14XZiaTHnELGFZAzJpM5HyuhZe25HJvJgApof7XnpB5hYISRDbsgMdR3cYcQ2cFUlbzFAPIgPcWC+bbUs6qGpoZf3xIm6aGIyXi4N5Js3eKW64oTPMM5+1ETBGHE2RSEb6cSz/PDWNbWYOStIffJtSzIQQL4K8DJRGVhfAFw+KWsilqwe25N6a8QxU6sE3AAAgAElEQVSCkKmQ8oNEUlFEHemezAradXoLBmck7S2w40V44xqoOAvLXof7voKFf4bQa+Cbp7vtKHai16v84evTBHo48ViimZsQ69pg3aPg7AWLXzLv3JLBRdi1wl25usC08Y1VcOy/ombSw8Q+mFaITNpM5OWt6bS06/nloqhLX6wrEQ9io67r/8D6isiFUJkBlVkoisKyiUEk550nv7LR0pFJrICPD+XT0q7ngRkjzDdpThIETRq8VtCBHUlbqWl1bXoV9mRWmDkoSV+TX9lISlEt1xtqQNLWDJ/dKx547/hg0Mh8rJYxN0NZKpSf7To1e7Qfdc3tHC8wgzFCf5C7F16fAUl/FZLPJ5JhwnKR7Gu0cPObYjfjiwegvXuZwxdHCzlZWMMvFl1Up28Odv9DyE+XvAyuZq6TkwwuwhLFMSfJtPFH3hP93qY9Ya6IrAKZtJnA2ZI6PjmUzz1TQwnzu8wXaNZ2cRxsSRvA2W8BWDpB1PB9dVzuttk6bTo9/92fx8wIXyID3M0zaXMtFB0dvPVsIHbhvUJN2mkbH+yFh5MdSellfRCYpC9Z33HPNMjqX1Vh0zPCiv6mN8A3oo+jkxB9I6B0MySZEe6LZiDUkTZVw4anYM31oGuBe/4Ht7wFrhepgTyDxI5t8QnY9oeu03XNbfxt81kmhw5h6YRh5o3t3HHY9XcYeztE32DeuSWDD/9ocPU3TSLZ3gIH3xS7dZ2Lo4MEmbSZwJ82peHmaMeP517hCzRjC7gF/NCLaTAwJBT8Y7rq2oK8nEkY6c2640WoRhopSAYX36aUUFLbbN5dtry9oOpEU+3BTKcZiZHYaTXMjPAjKb1cfv4GEDVNbbyzJ4c5Uf6G1QodWQPHPoBZz0HU4j6PT4KQUoXO6Ja0ebrYM3H4EOtN2lRVxLt6ivh9mf4UrDpw9YXjqMUQv1L0sMrYCsCrOzKpqG/hdzfEoJhTgtveAusfAxdfuP5v5ptXMnhRFLFom71TGOkYw6kvoL4Epg+uXTaQSZvR7Dxbxq70cp6aG8EQ18vU7uh1YqctfIBb/V+OyIWQtw+azgNw08QgsssbOCWtx22a9/bmMNLXlcRIf/NNmr0T7JwhZIr55rRGAseJupIW4/sezo70o7S2hbOlPdt3S6yDt3dnU9PUxjPzI3u+uDAZvv2ZePBO/GXfByf5gdhlUH4GSk93nZod6cepohoq61ssGNhlqCmEj++Ez38E7oHw8A6Y/zw4uPY8dv7z4B8L6x8lPy+Hd/fkcNvkYMYFe5k3xp1/gbLTcOO/wXmIeeeWDF7CEqGhXPzuGIqqirYW/jHiOXyQIZM2I2jX6XlxUxqhPi7cN23E5S8qOirs8SMGkTSyk9GLxO5H5jZAOJ85aDWsP3bOwoFJLMWx/PMcy6/m/mmhlzqo9obsJBg+FewczTenNRI4FlCN+1LqYFaksCJPOmulq/+SblTUt/DOnhyWjBtK7LAeDKrqy+DTe8F9KNz8lqhDkvQfMUtB0XTbbZsV6YdqTXWkeh0ceANWJ0DOLpj/J1ixHYaON3wOe2e49V1oqafu44dw1MJzC0ebN87CZNj7T9EnK3K+eeeWDG7CEsXRGIlk1nZRkzrticG3cYJM2ozi0+QC0kvr+eWiKBzsrvBfl7lV3OzDru3f4PqDoMlC3tAhkfR0tmdOlD8bTpwbWK5aErPx3t5c3B3tuDXOTDb/AHWlUJ42uOvZOumUUJecNH6opxNRge7WK9mSdOP1nVk0t+n46bwedtl07fD5A9BUJYxHXLz7J0DJD7j5C8fk1HVdfRTHBnkyxMXeOj5vJSnwzjzY/HOxuLXqgJCCaU0wDvGPIn3Sr4ltPsJbEQfwd3cyX5xtTcIt0n0YLHjRfPNKbAPPIOGYa0zStv9VUZ409tY+C8uSyKTNQOqa23jp+3SmjPBmQexVXL8yt3QkN4Pwi1ajhcgFomZP1w7AsolBVNS3sDer0sLBSfqbkppmNp0q5ra4ENwczegy1ukWNVj7s12IZzA4eZlU1wZCspWce56GlnYzBya5LJVZYofDSM5VN/HfA3ncOjmY8MuZV13I1t9B3h644V8wdJyJgUp6TezNwjH53FEAtBqFmRF+7EqvQK+3UB1pW5Po1ffmbDifB7e8A3d/IWrOTZ1Sp2dV2lh2aqczNWc1FB4xX7zbXxD/h0tfHRztjyT9T1iiqHG/yOX0spSkiJ22hEcGrUpHJm0G8vrOLCobWvnNkugrF+g2VAp55Kh5/RtcfxK5UMg/Cw4AcG2UHx5OdrJnmw3ywYE8dKrKj6aPMO/E2UkikQm0gQdWRTHZjARE0taq03MgWy6a9DmFyfDvSfD6dEj7umsHxhD+vT0DVHjqSuZVnaT8T6wUT3kExt/Ry4AlvWLMzWDvCoff6To1K9KPivoWThfX9n88uXvgtWmw52UYdyc8cVjsJvRSAvbBgTwyyxtQb3gFxX0o/O9B4d7bW/L2idqiuIcgfBAqjyT9Q1iisO4vPNTztftXg70LTH6gr6OyGDJpM4DC8428vSeHmyYGXb1AN3sHoA4uq/+LCb8WtA5d1v+OdloWjxvG5pQSudpvQzS36fjoUD5zowIY7mOAC56hqKrYaRs5y3bqeALHQWlq1+61MUweMQQXB611SLYGO2lfg8ZO7LR9eg+8PVcsMPRATkUDnyUXsjxhOMFDrvJZKT0NXz0Bw6fB/BfMGLjEJJw8YfydwomusQqAWRHCOn9XRj9/3hoq4YOOBO2+DbBstVnUPFUNrby8JZ2ZEb4kjo+AW94WzYw3Pm3UosQltDbA+lViB3DeH3sdp8SGGXENKNqeJZK1xXDqc1E7ORiVbh3IpM0A/rb5LArw3IIeCnQztoCzNwyb0C9xWQRHdxgxE85s7Lqp3zQxiKY2HVtOl1o4OEl/seH4OaoaWnnQnDb/AFXZUFNgG9LIToImQXtz1+61MTjaaZke7iOTtv4g/TsInS7qh258VdRevn8jvL8Uiq4sKXt5SzoOWg2PXzvqynM3VcOnd4v7621rRONjieWZslL0Ozv6PgD+Hk5ED/Xof/OfI+9CexPc+bFZ740vbTlLQ6uO3y7psPgfPlU4lZ76HE58YvrEW34H53Nh6WuyGbykdzh5ipKjnpK2Q/8BfTtMfaxfwrIUBidtiqJoFUU5pijKNx0/j1QU5aCiKBmKonyqKIpDx3nHjp8zO14f0Teh9w/H8s+z4cQ5Hp4VxjAv5ytfqNdD1jYInzP4dwhiboTzOV3mCXGhQwjycmadlEjaBKqq8u7eHEYHuDMt3Me8k3femEcmmndea2b09UIOevhtk4bPjvQjr7KR3IoGMwcm6eJ8njDHiVwozB4m3QtPHoEFfxbS1rfmCLfH8vRuw9KKa/n65DkemDECP/cr1Fjo9bDuEajOh9vfF7btEuvAP1osUh5+p6uWcXakH0fyzlPfX8qS9lY49LawL/ePMtu0acW1fHQwn3unhhIR4P7DCzOfhtBrYOMzUJFp/MTZO+HwW+LhecQMs8UrsWHCEsXCWPMV2ku11EPyu6Jpu3dYf0bW7xiz0/ZjIO2Cn/8KvKyqagRwHnio4/xDwHlVVUcBL3dcNyBRVZUXNqbh5+7Io7PDr35xyUnRTyJiENezdRJ1g9iuTl0PgEajsGziMPZkVlBeZ2U9bCRm50B2FWdK6nhgxgjzNmAFIY30CAKfHj5vgwkHFyHpSPsaao1vnzG7oz+e3G3rQzK+F8eIBT+cs3eCaavgqeNidyJrO7yWAOsfFxIz4B/fp+PmaMcjs67y+7z7/4Qj78K/iJ0OiXUxZSXU5Hf9DsyO9KNdr7Kvv6z/U78UjYKnrTLblKqq8sevT+PpbM9Pr7vIzVSjhZvfFLu9XzwgGmMbSnOtkPj6jIK5vzVbvBIbJywRVD3k7L7868c/FAnd9Cf7MyqLYFDSpihKMLAYeLvjZwWYA3zRcclaYFnH35d2/EzH63MVsz/Z9Q+bTpVwJO88z8yLxLUnd7zMreIYPqfvA7M0rj4wciacXt8lkVw2IQidXuWbk7Jn22Dnvb05DHGxZ9nEIPNOrO+4KYclDsr+Klcl/iGxkn9kjdFDh/u4MNLXVSZtfUn6ZvAOB9/LSBydPCDxF/DjE5DwmJCW/XsSpZ//lKNpGTw6OxxPF/srzPs97HgRxt8F8Sv69t8gMY3Ri4Vl/aE3AZgcOgRXB23/1LV1Ngr2HW3WRsHfpZawP7uSp+ePvvzvpmcQLF0tFqO3GVGT9v2vobYIlr0hesBJJOYgOF6YAl1OIqnXic9I8BQImdLvofU3hu60/RP4GdDZjMsHqFZVtVMfUAh0PsEFAQUAHa/XdFzfDUVRHlYUJVlRlOTycut72Ghu0/GXzWlEBbpzmyE9qDK3iqaWbv59H5w1ELNM1B91uN5FBLgTO8xDukgOcvIrG9mSVspdU4bjZG9mGXDpKdGbaqQN1bN14h0mdumPrDHM2vgiZkf6sT+rkuY24+3oJT3Q2iAWEyIXXv06V19Y+CI8dRTG3YFf6hp2O/2ElbpPL+/GV5UNX66AwDGw5GXbW6gYKGjtIO5BsZNakYmDnYZp4b7sPFuO2huzDkPI2ycSp6mPme33o7lNxwsbxbPNXfFXebaJWgxTHhZuphlbep44Y4uo/Zv+FITEmyVWiQQQu74jZlw+aUv7GqrzbGKXDQxI2hRFWQKUqap6YaX15e4eqgGv/XBCVd9UVTVOVdU4Pz8/g4LtT9buy6WgqonfLI5Bq+nhZtlUDQWHBrfV/8VEd0gkT6/vOnXTxCBOFNaQVV5vwcAkfcna/bloFYV7p5neF+iKdNWzzTL/3AOBKQ9DfSmkbTB66OxIP5radCTnnu+DwGyc7CRhRhE537DrPYPZG/t75rX8lTL/a3DY8zd4ZbxYDW5rFte0NooaOBTRQFvuSlg3k+8HjX1X3ensSF8KzzeR09d1pAdeA+chMM587R/e3p1N4fkmfrskBjttD4+A854H/1jRILuu5MrXNZ2HDU+CXzRc+yuzxSqRdBGWKHr+1RR2P7//VRgyUiwy2ACG7LTNAG5UFCUX+AQhi/wn4KUoSqdmMBjo1MUVAiEAHa97AlVmjLlfmBw6hEdmhXFNh8XvVclJAlU3uK3+L8bVV1ixpv4gkbxh/DA0Cnwld9sGJfUt7Xx2uIBFY4cy1LMPHjKzk4QMyGOo+eceCITPFV8+JhiSJIR542CnISm9rA8Cs3HSN4ODOwyfbtDlqqry9+/O0uQRzrCHP4OVO4QK47tfwb8ni92Ir58SbR5ufQeGjOjb+CW9x80fYpeJ2pmW+q460l19KUmuyhEuzXEPirpXM5Bf2cjqHVksjA1k+igDnm3sneDWd8Vu87pHhIT9cnz7C6gvg5teH7RNjSUWJixRHC9ss5J/EAoPw9RVg98AsIMekzZVVX+pqmqwqqojgDuB7aqq3g3sAG7tuOx+4KuOv2/o+JmO17erfa4hMD9xI7z55fXRhl2csQUcPYXu1paIXQZVWVCaAkCAhxMzRvmy7nhR38tGJP3O/44UUtfSzgPmtvkHUeyev/+HG7MtotGIuqb8/VB80qihLg52JIz0ZmtamfzsmRNVFQYUo+YYbMO/Na2M4wXV/Pi6CBzttKKlw33r4f6vhTPkhidF3duc39jWQt9AZ8rD0FILpz7rnzrSg/8RD6LxK80yXUV9C/e9exAHOw2/Xmzgsw0Ix8pFfxFKiH3/uvT1Mxvh5Ccw61kYNtEssUokl+AfA65+Hf2QO9j3L+G8PPFuy8XVz/SmT9vPgacVRclE1Ky903H+HcCn4/zTwC96F6KVo6qQuU30TtH2YFYy2Ii6ARRNl4skCEOSgqomjuZLmdZgQq9XWbMvl/EhXkwaPsT8b1B4GNoabas/2+WYeDfYOQvLbCO5cfwwcioaSM6Tnz2zUXIS6oq7u0ZeBb1e5f++O8tIX1dumRTc/cWRs2DFVrjjQ9Fw+Jqn+yBgSZ8RHA+B4+DQW6CqzIrw5UB2Vd/UkTbXwrEPIPZmsygPGlraeXDNYUpqm3n3R/GEeBu5czfpflHHvv15KEy+YOJK+PrHEDgWZj7b6zglkiuiKGJRN3uneO6uzBILBvEPgYOrhYPrP4xK2lRV3amq6pKOv2erqjpFVdVRqqrepqpqS8f55o6fR3W8nt0XgVsNZWlQd842rP4vxs1PSCQvcJFcMCYQJ3sN649JF8nBRFJ6OTkVDeZvpt1JdpJYAAi18b4+zkNg3G1w8nNRJ2IEi8cNxc3Rjs8OF/RRcDZI+neAYvD9/euT5zhbWsfT8yIvXy+kKBC9BGb8WOysSgYOiiJ228pOQ95eZo/uwzrSY/+F1jqz2Py3tut59IMjpJ6rZfXySUwONWHRTVHghleEi+YXD/7QL2vTM6Km/6b/yIbwkr4n7FrRWqvsNBx4HbT24jNpQ8hvjd7SZfVvPjveAUXMMqjMFB8iwM3RjnkxgXxz8hyt7VfQv0sGHO/uzcHf3ZFFY/qo3iwnCYZNAmevvpl/IBG/Etqb4NiHRg1zcbDjhvFD2XiquP8a/w520jdD0GSDXIHbdHpe2pJO9FAPFo+10brMwc7YW8XCyqG3mBrmg4O2D+pI9To4+Iaooeyl3FCvV3nuixPszqjgzzePZW50gOmTOXvBLW8LI4hvnoaULyF1nWh3ERDbqzglEoPoVOKkfCl2osfeJiTnNoRM2npL5hahtfU0c8+qgUL0jZdIJG+aOIzzjW19W6Qt6Tcyy+rYnVHBvVNDcbDrg1tGc62Q3Ni6NLKToeNg+DRhSHKlwv8rcFtcCI2tOr45IXe6e019GRQdhUjDpJFfHCkkr7KRZ+dHounJcVgyMLF3hon3QNrXuDSXET9yCLvSzdxk+8xGqM4XNv+9QFVV/rQpja+On+O5BaO53ZDWRT0xPEE0kk/5AtY/JhY0Zvyk9/NKJIbgGQw+EbD3n2Jhc9oTlo6o35FJW29oqYe8/bZdTO7mJyRtF0gkZ0b44e3qwLrj0kVyMPDu3lwc7DQsTxjeN2+Qt0+4r9pif7YrEb8CzudA1jajhk0M8SLC343PkqVEstdkbAFUg5K25jYdr2zNYNJwL+ZE2UivTlsl7iFQ9ZD8HrMj/ThbWkdxTZP55j/wGniF9trC/M1d2byzJ4cfTR/BqsRwMwUHzHwaRswU3/fLXre9Wn6JZQlLBH27ULcFxFg6mn5HJm29IXc36NtsO2kD4SJZkS7q+wB7rYYbxg1l6+lSapvbLBycpDfsOFvGJ4fyuXVyMD5ufWTlnJMEdk4QktA38w9Eom8EtwA49KZRwxRF4Y74EI7mV5NZVtdHwdkI6ZvBfagwn+iBDw7kUVLbzHMLolBkk+zBjfdIkcgfWcOscE8Adptrt63oqHCPTXi0Vxbm/ztSyJ+/PcOScUP57ZIY8/5OarRw9+fwxGHwG22+eSUSQ4hcKI4zfmzZOCyETNp6Q8YWsHeF4VMtHYll6ZRIXtBoe9nEIFra9WxOuUpDTolVk1Fax1MfHSN6qAe/McYi2liyk8RnyN6p795joGHnAJN/JO4xVcZ5OS2bGISdRuFTaUhiOu2tkLUDIuYLE4arUN/Szms7s5gZ4cu0cJ9+ClBiUeJXQkMZoyt3EODhaD7r/wOvi56AE+8xeYodZ8r42f9OMmOUD/+4fXzfSHXtnWFIqPnnlUh6IuI6+EmKzZZTyKTNVFRV1LOFzZbNJN38hUTygrq2CSFejPBxYb1stD0gqWpo5aG1yTjaa3nrvjhcHPpIAlNfBmWpUhp5OSY/IFa1D7/T87UX4OvmyHXRAXx5tIg2nTQDMon8fcK9r3NV9yq8uyeHqoZWnpkvdx1shvA54B2GcvgtZkf6sTujnPbeftZqz0HqlzDpXnDyMGmKY/nnWfXhUaIC3XnjnsmiT6BEMtjwMkN95gBFJm2mUpklioVH2ahr5MXELIWKs10SSUVRWDYxiP3ZlZTUNFs4OIkxtLbreeyDI5TUNvPWfZMZ5uXcd2+Ws0scbXTV7Kp4DIXoGzrsvxuNGnpHfAiVDa1sSzOzs52tkP4daB17/L2sbmzlrV3ZzI8JYEKIdD61GTQasdtWcJAl/uXUNrdzorCmd3Mefls4R5poYZ5ZVs+Daw7j7+HImgem4O5k37t4JBKJ1SGTNlOxdav/i4m+EVAuabStqrDhhNxtGyioqsrvNqRwMKeKv90yjol90Uj7QrJ3gpMnDJ3Qt+8zUIlfKXoipXxh1LCZEb4EeDhKQxJTSf9ONMPuoWnrG0nZ1Le2y102W2TCcrB3IaH8SzQKvZNItjZC8rvCfMR7pNHDS2qauf/dQ2g1Cu8/OAU/dxtX/0gkgxSZtJlK5hbwGWXSDXZQ4h7wg4tkByN8XZkQ4sU62Wh7wLBmXy4fHypgVWI4yyb2cRsLVRX1bCNm9qroflATOh38Y4UhSYc7qyHYaTXcOjmYnWfLKK2VO91GUZEJVVk9ukaW1TazZl8OyyYEMTrQvZ+Ck1gNzl4w7nYc075kRpC2dy1uTn4CTedh2uNGD61pauP+dw9R3djKmgemEOpz9YUGiUQycJFJmym0NUHuHhg1z9KRWBexy6D8DJSd6Tp108Qg0oprOVNSa8HAJIaQlF7O89+cZn5MAM/2x87B+RyoyRcWvpLLoygwZQWUnIKCQ0YNvW1yCHpV9A+TGEH6ZnGMmH/Vy17dkUm7TuUn10X0Q1ASqyR+JbQ3sdJtHycKqznf0Gr8HHq9MCAZOl70ZzSC5jYdK9cmk11Rz5v3xTEmyNP495dIJAMGmbSZQt5eaG+WVv8X0ymRvGC3bcm4oWg1CuvlbptVk1lWzxMfHWV0oAcv3zGhf5oDZyeJozQhuTpjbwdHT6Pt/0f4upIw0pvPkwtQjdils3kyvgP/mKu64xVUNfLxoXxujw+ROxu2TOAYGD6dhMp1oOrZnWmC9X/WdtEyZ+rjPTqVXohOr/LUx8c4nFfFS7dPYMYoX+PfWyKRDChk0mYKmdtEX6kRMywdiXXhHiDkXBfUtfm4OTI70o+vjheh18sHR2ukurGVFWsP42in4e3743B17KdmqTlJ4D4MfOVOxVVxdIOJd8Ppr6Cu1Kihd8SHkFvZyKGcqj4KbpDRXCOavfewy/bKtgwUReGpOfJ31+aZshLHunwWO6eaJpE88Bq4BULsTQYPUVWV36xP4fvTpfxuSQw3jB9m/PtKJJIBh0zaTCFjC4y4RvQqkXQnZhmUp0H52a5TyyYGUVzTzEH54Gh1tOn0rPrwKOeqm/nPvXEE9aVT5IXo9cI5Mmy2UavLNkv8CtC3wdG1Rg1bNGYo7o52fCoNSQwjazvo269q9Z9ZVseXRwu5f1oogZ6yt6DNE30DuAXymPN2dqWXG7erXXYGsrYJCbSdg8HD/rk1g48P5bMqMZwfzZB19RKJrSCTNmM5nwuVGVIaeSViLnWRnBcdgKuDVvZsszJUVeX3G1LZl1XJX24Zy+TQPnaKvJDSFGislNJIQ/EJF061ye+Crs3gYc4OWm6YMIxNp4qpbTZ8nM2S/j04D4Hg+Cte8tKWdJzttTyWOKofA5NYLVp7iHuA2MaDONfncaakzvCxB14Tqp3JDxo85IMDebyyLYPb44J5boF0LZVIbAmZtBlL5jZxlEnb5XEPFMXUF9S1OTtoWTAmkE0pxTS36SwYnORC3t+fx4cH83l0djg3Twru3zfP6ahnk/3ZDGfKw1BXDGe+MWrYHXEhNLfp+eZEcR8FNkjQ6yDje3Fv115eIpxSVMOmUyWsmBmGt6vhOyOSQc7kH6Fq7LhHu9Vw6/+GSjj5KYy/E1x9DBry7ali/t9XKcyN8ufFm8aiSJWCRGJT9FPxyiAicyt4DRd2/5LLE7sMvv0ZlKeDXyQgXCS/PFrEjjNlLBo71MIBSnZnlPPHb05zXXQAP7PEam12EvhGgoesxTCYiHni3nPobaPqX8YFezI6wJ1PkwtYnjC8DwO0Xs6U1JJb0YBGUdBqFDSKgkajoFUUNBrQKgoeFceJbqwg12cmdYU14nzXNWLMXzefwcvFnhUzpSRNcgHugSjRN3Ln6e94Mq2AR2eH9zzmyLvC0CzhMYPe4kB2JT/+5DgTQ7x4dfkk7LRyzV0isTVk0mYM7a3iYXP8nbIO52pE3wjf/lzsts3+GQDTw33xc3dk3bEimbRZmKzyelZ9eJQIfzf+eWc/OUVeSHurcGCdcHf/vu9AR6MVtW1bfgulqRAQa9AwRVG4PT6E5785zdmSOpvqKaaqKu/uzeVPG0/Tkw/S03afEaHVsHSzEzXsueJ1v1gUhbuTvZkjlQx4pjyMe+qXBBV+TUPLNVc3dGpvFYsv4XPBP6rHqY/kVbFibTLDfVx490fxODvIvpYSiS0ikzZjKDgAbQ1SGtkTHkNh+FRR19aRtGk1CkvHD2Pt/lyqG1vxcpHSIktQ09jGyrXJOGg1vHVfHG795RR5IUXJ0NYopZGmMPFe2PEiHH4blrxs8LCbJgbxl2/T+Cy5gP+3JKYPA7QeWtp1/L/1KXyWXMiC2ACe7HB61OlVdKqKXq+iV8XPelVl/MbnqbebxEtzEjvOgV5Vu17X6VXstBoWjQm07D9MYp0Mn0qDVxR3V21hf+ZzXBd7ld+T1HVQXwLLVvc47cHsSh5Yc5hADyc+eChBfndKJDaMTNqMIXMraOxh5ExLR2L9xCyDzT+HiowuS/dlE4N4e08OG08Vc3fClXsgSfqGNp2exz86SsH5Rj5aOZUQbxfLBJKdBIpGOLBKjMPFG8bcCic+gbm/A2cvg4Z5uzowLyaAdceK+PnCKBzsBre0qryuhUc/OMKRvPM8NWcUP7ku8uo7yjVFcD4NrvsDc6MD+i9QyeBBUXCY/ggxm37KvuPbIPYKSgJVhQOrwXe02Gm7CvsyK3hw7WGCh7jw0YoE/D2kW6lEYssM7m9uc1c5Sp4AACAASURBVJOxFUKngaPtyItMJuZGcbzARTJ2mAcR/m7SRdJCPP/NafZkVvDiTWOJH+FtuUBykmDoBOHSJzGeKSvFTuWJj40adntcCFUNrWxLM67X20AjpaiGpa/uIfVcDa8un8jT80f3LAHO+F4cr2L1L5H0hP2EO2jQuDEy+yqfzfz9UHwCpj521TKLXenlPLDmMKHerny8cqpM2CQSiUzaDKb2HJSlSmmkoXgMg5Cp3VwkFUVh2cQgDueep6Cq0YLB2R7/PZDH+/vzeHhWGLfFhVgukJZ6KDwspZG9YdgEYUl/+G3R785AZkb4MdTTaVD3bNt4spjb3tiPCnzx6HSWjDPQ6Cb9O2Hy4ict1CW9wMGVvJBlzGrfR0Fe9uWv2b9aLFiNu+OK02w/U8qKtcmE+7nx8cNT8XN37KOAJRLJQEImbYYirf6NJ2ap6MdVkdl1aukE8RD11XG529Zf7M2s4PcbUpkb5c/PF/Zc9N6n5O0TzYtlf7beMeVhqMyE7B0GD9FqFG6dHMyu9HKKa5r6MLj+R69XeWlLOo9/dJTooe5seOIaxgR5Gja4rQmyd4pdNmkwJeklHjMfw17RUbnrzUtfrMqBMxsh7kFwuLw8/fvUEh757xFGB7rz0coE2VpCIpF0IZM2Q8ncAu7DwN82ivjNQsxScTy9rutU8BAXpozwZt2xIlS1Bzs3Sa/JqWhg1YdHCfdz5Z93TkDb306RlwSUBFpHYVQjMZ2YpeDqB4feMmrYbZND0KvwvyOFfRRY/9PQ0s6qD4/yr20Z3Do52Pididw90N4EkQv6LkiJzRA8agwHtJMYkfsZ6C5qaH/ozQ4X2JWXHfvtqWJWfXiU2GGefLBCmo5IJJLuyKTNEHTtkLUTRs2VK7HG4BkEwVMg9atup5dNDCKrvIHUc7UWCsw2qKxv4aG1h9FqFN65P946bMqzk2B4Atg7WzqSgY2dI0y6H9I3w/k8g4cN93FhWpgPnyUXou/JA38AUFDVyC2v7+P70yX8ZnE0f791HI52Rtqhp28GexcIlcY4EvOQGXoXXrpK2i787muuhaP/hdibhcPyRWw4cY4nPj7G+BAv/vvQFDydreB+LZFIrAqZtBlCUTK01EhppCnELoPSU1CZ1XVq8dihOGg1rJOGJGalXafnSF4VL29J55bX9zHlxW0UVDXyxj2TLecUeSH15eJ3QUojzUPcA8KFM/kdo4bdER9CflUjB3Iq+yiw/uFQThVLV++lqLqJd38Uz4qZYSjGLqqpqqhnC7sW7KXRg8Q8BE5aQp7en8Y9b/xw8tgH0FoH01Zdcv2XRwv5ySfHmBw6hPcfnGIdC2wSicTqkEmbIWRuBUULYYmWjmTg0SmRTP1BIunpYs+1UX5sOHGOdp3hRgqS7qiqSm5FA//dn8vD7ycz8Y9buOX1/fx7ewbtepVVieGsf3wGU0Za0CnyQnJ3iWNYoiWjGDx4BkPU9WL1vs3wGrWFYwJxd7Lj8+SBK5H85FA+d799AC9ne9Y/PoPE0f6mTVSWBjUFUhopMSvTIvz5WJ2HZ9lhKEkBvQ4OvgHDp8Owid2u/exwAc98foKpYT6seSD+6k25JRKJTSPvDoaQsQVCphjcE0lyAZ7Bwunu9HqY9WzX6ZsmBvFdail7syqZHelnwQAHFjWNbezNqmB3RgW7M8opPC8e1oOHOLNk/DBmRvgyPdzHOmshspPA0VPY/UvMw5SHIe1rSPkSJl6hL9RFONlrWTphGJ8nF/L7G2MHlAyrXafnhY1prNmXy8wIX169axKeLr2IP32zOEbMN0+AEgng6mhHZtAyWko+x/HwW6IfW3UezH+h23UfHcznV+tOMTPCl7fui8PJ3khpr0QisSlk0tYT9eVQfBzm/MbSkQxcYpbB978WEkmfcACujfLHy8Wejw7myaTtKrS26zmWf549mRXsyqjgVGE1ehXcHe2YGu7DI7PCmBnhR6iPi/HSsP4me6doqK2Vtx2zMWIm+EUJg4MJyw2uub0jbjgfHMhnw4lz3Dt1YDS6r25s5fGPjrI3s5IV14zkF4uisNP2UiyS/h0MHX/ZGiOJpDfERYezrnA6d5z4FOXccfAKhajFXa+v3ZfL7zakMifKn9funiQTNolE0iPy6aknsraLo6xnM52YpSJpO/0VzHwaAEc7LcunDOeNpCwKqhqto+bKSmhq1fH5kQKSzpZzILuShlYdGgUmhHjx5JwIZkb4Mj7EC/ueHlhVVchyrCFJOvaBWGme9rilIxlcKArEr4BNz0LREQiOM2jYmCAPood68HlywYBI2jLL6lixNplz1c387dZx3G6OXoONVVB4CGY91/u5JJKLmB3px7Ob53Nn+06x8Lvgz8I5Enh7dzYvbExjXkwAq5dPwsFOVqpIJJKekXeKnsjcIqy1A8dbOpKBi1cIBMV1a7QNcO+0UBRFYe2+XMvEZYUcyK5k0Su7+O1XqWSW17NsYhBv3DOZY7+dz5erZvDTeZHEjfDuOWED2Pln+Hs4nN7Q94FfifZW2PgsfPU4jJwldoMk5mX8neDgLnbbDERRFG6PC+ZkYQ1pxdbt4rr9TCnLVu+jvqWdjx9OME/CBqJWWdVDhKxnk5ifqEB3yt1Gk+UUKz6fE+8B4PWdWbywMY3rxwby2t0yYfv/7d13eJRV2sfx70mnBQidBEikdwKhgyAKAqKCFAEBxYIiuLqWtezrFtvqWthdRREbCCKggigiilSldxTBUKVIb1ITkpz3jzMqgUASMpOZgd/nunKReeodfAxzzznnvkUk5/Tb4kIy0l1T7cpXQ4j+qvKkdlfYtdo1F/UoV7QAneuWY8Ky7RxLSfNjcP539NRp/jr5e3qPXESGhXF3NmXuI1fxbLe6dKxTNvfrjlKOwaIRcPoETOwP0x6B06d8E/z5HN0No6+HpW9B86HQbzJEFsnfGC4HkUVcMrx2spvOnUNdG8QSERrChKXbfRjcxbPW8uasdfxl9EwqxhRkytBWNKrkxaI6ydPdB3JnFYYQ8QZjDFdWK8XQU/eSPuAziIrm1ZkbeGH6em6oX57/9U7M2YdvIiIe+o1xIbtWwcmDmhrpDb832s482jawZTxHT6VdUs1+c2v2+r10GDaPD5ds485WCXz1wJW0qFIybxddM961qeg/GZoNcaMw71wD+zd6J+jsbF8Cb7aB3Wug+ztw7bOBMU3zUtX4TkhPhRWjc3xK8UIRtK9dhk9X7SQlLd2HweXe8ZQ0Hhi7mIZzBjA/6s9M6nSa2GJe7O2XnuZG2qpeqw/kxGeurFaKdaeKszojgVe+/omXZyRzU2Isw25ukPf1mCJy2dFvjQvZOBMwULmdvyMJfsUqQmwjWJs5aWtYsTgNKhRj1IKtl0Sz39w4eDyVB8avZOCopRSODOOTwS34vy61KBCRxwXp1sLiN90IQqWW0PE56DMejuyAkW1gzUTv/ADnu/fSd+C9zq7v1R0zoG4P391PnFLVXP+75aPcDIEcujmpAodPnGbGj3t8F1subd1/nG7Dv6NV8nM0DkkmIrokURN6w0/TvXeT7Yvh1BGV+hefal2lJMbAQxNX879ZG+mVFMeLPesTGhLgRaNEJCApabuQDTMgtiEUKuHvSC4Ntbq60cszpkiCG23bsv84c5L3+imw/GWtZeqaX2j/ylymrtnFn66uytQ/tSKxYnHv3GDzbNifDE3v+aOaYPVOcM93ULYuTLoLPh0Cqce9c7/fnD4Fnw2FLx50vdgGzYGydbx7Dzm/Rre5nmOb5+T4lJZVShJbrEDATJGcvX4v17/2He2PTqZn6Fxo8yjm7m+hTC2YcItrbeANydMhJBwqX+Wd64lkoXihCOrHFWPL/uP0bVqR52+qp4RNRC6akrbzOXEQdi7T1Ehv+n2K5JRMmzvXLUeZ6Ejem781/2PKZ3t+PcXdY5YzdNxKyhcrwOf3teLB9tWIDPNiuefFb7q1OrW7Zd5eNA5uneqq5a36AEZeBXt+9M49j+yA9zq5KpFXPgJ9J0ABLyWhkjM1rnN/5yvH5PiU0BBD90ZxfLdxPzsP57xBt7dlZFhenbmB20cv5YbC63mY96FGF2jzGBSMgQGfQVwT+OQO94zlVfJXEN9SayzF5x7tWIO/X1+LZ7vWIUQJm4jkgZK289k8x1UWU9LmPcUrQfmG56xrCw8NYUDzeL7dsJ/kPUf9FJxvWWuZsHQb17wyl7nJ+3i8Uw0m39uCmuWivXujA5vcG9Kk2yEs8tz9oWGu52D/yXDyELx1lZtSZ/MwNXXLt2792v4NcPMH7voh6jmU78Iiod7NsP4L96FTDvVsFAfAx8v8s6706KnT3DN2OS/PSObOmuk8k/YypnQt6PbmH+vNoqKh3yduBHfKEFic80qZ5zi4Bfb/BNU6eiN8kQtqXrkEA1smBH4fTREJeErazufAJihY0q3DEu+p3RV+WQmHtmba3KdJRSLDQi7J0bbtB0/Q/50lPPrJ99QsF830B67k7jaVfbMQfenbLmFKuv3Cx1W+CgbPh4rN4fP74ePb4VQuS79bCwuHw/s3utGQu2ZBzS4XH7vkXWJ/V5BkzYQcn1IhpiAtK5fko+Xb831d6ca9x+g6fD4z1+/l6WvjeOLIU5iQMOg9DiILZz44oqBbm1mjC3z5CHz7ysXddMPX7s+qHfIWvIiISD7K9l2jMSbKGLPEGLPaGLPWGPNPz/ZRxpgtxphVnq8Gnu3GGPM/Y8xGY8waY0xDX/8QPtHmEfjzDxox8LbzTJGMKRRBt8RYJq3YwaHjqX4IzPvSMyzvfreFDsPmsWr7YZ7pWofxdzUjoWQh39ww5aibOla7GxQpm/3xhUtDv0lw9d/cf483W8POFTm7V+px+ORO+OoJqNHZJWylquUtfsm7snVcAZoVY3I1etozKY4dh06ycPMBHwaX2ddrd9N1+HwOnTjN2IFJ9N/5FObQFrh5rBuVz0pYJPQcBXV7wsx/wsyncj9KnDwdSlSFEpXz/DOIiIjkl5x81J8CtLPW1gcaAB2NMc08+x6x1jbwfK3ybOsEVPV8DQLe8HbQ+SbciyWmxSke795UnlVFEmBgywRS0jL4cOm2/I/LyzbuPUrPEQt4auqPNL0ihq//fCX9mlXy7ZqG1eMh5VdXgCSnQkKg9UMwcJorg/5OB1j4+oXfCB/c7I774ROX8PUao7VBgSSxP+xd60a0c+ja2q4XYH4UJMnIsLwyI5lBY5aTULIQn9/Xiuab/+tK8F/3MlRqceELhIa7qZMNb4VvX4bpj0FGRs5unnIMtn6nqpEiIhJ0sk3arHPM8zLc83WhjzZvBN73nLcIKGaMKZf3UOWSUasr/LICDv2caXP1skVoWaUE7y/4mdPpOXwTFmBOp2fw2qwNdP7vd2zef5xhN9fnvdsaU96bPaaykpHhCpDENoK4pNyfX7EZ3PMtVG0PXz0O4/tmvS5qwzcwsq0rPNLvY5fwaa1GYKnbA8KiclWQJCo8lK4NyjN97W6OnDjts9COnDzNXe8v438zN9CjURwf3dOc2C2TYOFr0ORuVwEzJ0JC4fr/uh6Ei0fA5/flrNXB5jlu+qiSNhERCTI5WlRjjAk1xqwC9gIzrLWLPbue9UyBHGaM+a3qQSxw5se1OzzbRJzaXd2fZ02RBLi9ZQK7fz3F9B9253NQeXciNY2eIxby0tfJtK9dhm8ebEO3xLj8WYC+aRYc2JC7UbazFYxxa4k6Pu/aXYxoBT8vdPushXkvwQc9oGhFV85fRXoCU1RRNw35+48h9USOT+uZVIHUtAymrN7pk7CS9xyl6/D5zE3ex9M31ubFHvWI2rUMpj7gCoxc+1zuLmiMa9re5lE3LfiTOyE9m4QzeTpERru1nCIiIkEkR0mbtTbdWtsAiAOaGGPqAI8DNYDGQAzwqOfwrN6hnjMyZ4wZZIxZZoxZtm/fvosKXoJU8Xgo1+CcKpIAV1UvTXyJgrw7f8u55wUway2PT/qe1TsO89/eDRjetyElC2dRvdFXFo+AwmXcKGZeGAPNBsMdX0NoBIy6Dub+Gyb0g1lPu1GcO76GmATvxC2+kdjfTZVd91mOT6kTW5Ta5aN9MkXyy+930XX4fI6eSmPcXc3o3zwe8+tO91wVjYMe77nKprllDFz1BLR/CtZOggn9Xb/ArGRkuA8jqlztpliKiIgEkVyVr7PWHgbmAB2ttbs8UyBTgPeAJp7DdgAVzjgtDvgli2uNtNYmWWuTSpUqdVHBSxCr3RV2LofDmdevhYQYbmsRz8pth1m57ZCfgsu9d+dvZcqqX3i4Q3VubJDPA8v7N8LGGZB0B4RFeOeasQ3h7nnuv9PsZ+GnL+Haf8FNb7kqfhLY4ltB8YRc9zTrlVSBtb/8yvQfdrPz8ElOnc7BlMMLSM+w/Hv6egZ/sIJqZYow9b5WNEmIcSOA4/vC6ZOuImTBmDzdh5b3Q+eXIPlLGNcr68bxu1fDsd1QVVMjRUQk+GT70aYxphRw2lp72BhTALgGeMEYU85au8u4uV9dgR88p3wGDDXGjAeaAkestbt8FL8Eq1pd4Zt/uClcrR/MtKtHUgVe/jqZ9+ZvJbFi4DdoXrT5AM9NW8e1tctwb1s/VKRbMhJCwiFpoHevGxUN3d+BmjdAdHmo0CT7cyQwGAOJntHRg5sh5oocnda1QSzPf7mee8Yu/31bkcgwShaJpGThCEoWjqSE588/vjyvi0RSKCL09+nAh0+k8qfxq5iXvI8+TSrwjxtquyby1sKUe2HXGug7EUpV987P3OQuiCjk+riNuQlumeimiv4m+SvAuHWbIiIiQSYn81HKAaONMaG4kbmJ1tqpxphZnoTOAKuA3xbTTAM6AxuBE4CX30nKJSEmARKuhCVvQfOhmUaICkeG0atxBUYv2MoTnWtStmiUHwO9sF1HTjJ03AoqlSjISz3r538D1VO/wqoPoE53V8Lf24z5Yw2iBJcGfd0o6cqxrspnDhQtGM70B1qzYc8x9h9L8Xylsu9YCgeOpbBh7zEWbk7h8HmKlUSFh3gSu0j2/nqK/cdSeK5bXfo2rfjHQfNegrWT3ZTGal7uldagL4QXdOvbRl8P/SZDoRJuX/JXENcYCpX07j1FRETyQbZJm7V2DZCYxfZ25zneAkPyHppc8lr8yRW2WDsZ6t+cadetzeN5d/4WxizayiPX1vBTgBeWkpbO4LErOJmazvhBzSgS5Yd1MqvGQeoxaDoo/+8tgS26PFRp756Rtk/keM1YpRKFqFTiwr0EU9MyOHg8NVNit/9YCvuPutcHjqdSILwgr/VtSKNKZ4yWr/scZj8D9Xq7//99oXZXl7hN7A+jOkP/T8GEuIq17f7PN/cUERHxsYtY+S3iJVWugVI1YOGrUK9XptLxFUsUpH3NMoxbvI372lUlKjzwmpz/8/MfWbX9MCP6NaRKaT/0KcvIgCVvQlwTV+pf5GwN+7tiH5tmerXMfURYCGWLRuVuFHz3DzDpbohNcuX6fTkqXa0D3PIxfNgb3uvkfr8AVOvou3uKiIj4UK4KkYh4lTHQfAjs/h62zD1n98CWCRw6cZpPV/qmBHleTFi6jXGLtzG4bWU61vFTG8KN37j1Sk3v9s/9JfBVvRYKloQV7/s3juP74cM+bp1k7w8gPB+mPCe0hgFT4ORBmPsCRMdCmTq+v6+IiIgPKGkT/6rbCwqVhgWvnbOr2RUx1CwXzXvzt+Jm3QaG1dsP8+SUtbSqUpKHO3ipiMLFWDwCipRzPblEshIWAfV7u/5kx/zUWiUt1ZXiP77X9QEsUjb/7h2XBLd94dph1O2pRvAiIhK0lLSJf4VHQZNBrmT93nWZdhljGNgynp/2HGXBpgN+CjCzA8dSGDx2OaUKR/Jqn0RCQzxvAk/9ClvmwaI3XAl+X9uX7Ka8Jd2hnlNyYQ0HQEYarBmf//e2FqY9DNsWwI3DXSuJ/Fa2Ljy4LsfFWERERAKRkjbxv8Z3QFgBWHjuaNsN9ctTolAE7wVAs+209Azu+3AlR46fYHSnCIr/+D58ei8MbwrPV3TV6qY/5tbQnNV/zuuWjHTNrxvd5tv7SPArVd2te1wxxiVR+WnJW7BiNLR60DVm95eQUPclIiISpFSIRPyvYIwr1b1yDLT7GxQp8/uuqPBQbmlakVdnb2Tr/uPEl7xwVTuvs9atG9u5ghXzZ/DQrhU0iPiZ0MmpnthLuilYdbq7UYTIaBjbAz7oBXd8lblPlLecOuIqAtbpAYXVmF5yoGF/+Ow+2LE0//rtbZ7jPsSo1gnaPZk/9xQREblEKWmTwNB8CCx7F5a+dU5Z7n7NKvHG3E2MWrCVf9xQ27dxHNvnSoPvXP7H18lDANSxkewtUoPQeoNctcbYRlCs4rnrZG4eA2NvgokDXAU7b09fXDkWTh9XmX/Judrd4MvHXEGS/EjaDm2FibdCyWrQ/S0I0aQOERGRvFDSJoGhRGWocR0sfRta/Rki/hhRKx0dRZd65flo2XYe7FCNaF/0Q1s1Dub8649pjSYESteCmtezu0gd7pkN4WVr8cGglhCWzRvQK9q4kuZThsDUP8MNr3qvAEJGupsaWaEZlD+nfaJI1iKLuMRt7WTo+DxEFvbdvTLSXWl/a6HPh+7eIiIikif6+FMCR/OhblRr1bhzdt3eMoHjqel8tGyH9+97aKtLrgrEQIdnYOCX8PgOGDyfI+1foc/y6uyMrMxr/RoTkV3C9pvEftD6YTflc/5/vBfrhq9dvCrzL7nVsL9rxP7jp769z3fDYPsiuO4liEnw7b1EREQuE0raJHBUbOYa7y563X1af4a6cUVJqlScUQu2kJ7hxWIK1sKXj4IJdf2jWtwHlVpARCEyMiwPTVzF9oMneP2WhpSJzmVvqav+6ta6ffMPN8LhDYtHQJHyUPN671xPLh8VmkKJqq4gia/sXOFGrOt0dyX2RURExCuUtEngMAZaDHWFP36ads7u21slsP3gSb5Zt8d79/xpmuth1fYxKBqXaddrszfyzbq9PNmlFo3jY3J/7ZAQuPF1N5Vx0t2wfUneYt273hV3aKwy/3IRjHGjbdsXuZYR3pZ6AiYNcj3RrntZPdFERES8SEmbBJYa17viHlk02+5QqwyxxQp4r/x/6nE3yla6FjQbnGnX7PV7GfZNMjclxjKgeaWLv0d4lGsoHF0ePuwDB/MQ+5I3ITQSGg28+GvI5a1+HwgJc9N2vW3Gk3BgA3R9HQoU9/71RURELmNK2iSwhIZBsyFuNGD70ky7wkJDGNC8Eos2H2TtL0fyfq95L8KR7XDdK5lGrn4+cJz7x6+kZtlonu1WF5PXEYNCJVwVyYw0GNfr92qUuXLyEKweD/V6uuuJXIzCpaFaR1j9IaSf9t51N8xwRYSaD4Ur2nrvuiIiIgIoaZNAlNjP9Tdb+Oo5u3o3rkiB8FBGzd+at3vsXQ8LXoUGt0Cl5r9vPpGaxt1jlmOM4c3+jSgQ4aWGvCWruBG3g1tgQn9IS83d+SvHwukT0EQFSCSPEvvD8X2Q/JV3rnd8v6uUWrq2+rGJiIj4iJI2CTyRhd0UwHWfu0qJZyhaMJzujWKZsuoX9h9LubjrWwvTHoaIwtD+qTM2Wx775Ht+2nOU//VJpEJMwTz8EFmIbwk3Doet38Ln97s4cuK3Mv+VWkK5et6NSS4/Va6BwmXdBwF5Za17lk8egptGuunAIiIi4nVK2iQwNb3bVXRc9MY5u25rkUBqegbjFm+7uGuvmegSp2v+DoVK/r753flb+Wz1LzzcoTptqpW62MgvrP7N0PZxWD0O5r2Us3OSp7v+cSrzL94QGgYN+rj2EUd35+1aK8fC+qlw9d+gbB3vxCciIiLnUNImgSm6PNTt4cqTn7UGrErpwrSpVooxi34mNS0jd9c9eRi+/ivENoKGtwGw72gK45ds47lp67i2dhnubVvZSz/EebR5FOrdDLOfgTUfZX/84hEQHQfVr/NtXHL5SOwPNj3Lnog5dnCzK+QT39qtQxURERGfUdImgav5EDh9HJa9d86ugS3j2Xc0hS++/yV315z1DPbEARbU+Ct///xHOgybS+Nnv+GxSd9TtXRhXupZP++FR7JjDNzwqpvuOOVe+Hnh+Y/d8yNsmQdN7nQjJCLeUKKye/5Wjs35NN0zpae5NhYhYdBthGtvISIiIj6jf2klcJWt6yrRLRl5TuGOK6uWonKpQrz73VZsNm86j546zaz1e3h7wiQylr7N6NPt6fvFKSYs206Z6Cj+0rE6nw5pydT7WlEkKp/6n4VFws1jXXuD8X3hwKasj1s8AsKioOGt+ROXXD4S+8PBTfDzgtyf+90w2LEEurxyTn9DERER8T59dC+BrcV9MLY7/PCJW4fjERJiuK1lAk9++gPLfz5E0hnNr0+mprPs54Ms2HSAhZsO8P3OI9iMdKZEPM2vYcU42epxPqpRifpxxYgI8+PnFgVj4JaP4O1r4IOecOc3bttvThx06+/q9cq8XcQbat0I0x5xPdviW+b8vJ3LYc6/oG5PN4VZREREfE5JmwS2yle75tcLXoX6vd3UQo/uDWN5cfp6Rs7bTFqGZcGmAyzadICV2w9xOt0SFmKoX6EYg9tUplv6dCov2Qxd32ZwvUQ//kBnibnCtQIYfQOMvwUGfOpG4QBWvA9pJ1XmX3wjoiDU7Q6rJ0Cnf0NUdPbnpB6HSYOgSDnonMNCOiIiIpJnStoksBnj1rZNGQKbZ0Pldr/vKhgRRp8mFXlz3ma+/nEPIQbqxBbl9pYJNK9cgsbxMRSKDINje+HVlyHhysAcGajYDLq+Dp/cAVOGutLpGemuWXF8a1XlE99JHADLR7mR7KSB2R//9f+5qby3fgYFivk8PBEREXGUtEngq9sTZj4FC17LlLQBDG5bmYIRYdQqH02ThBiKFshiTdqMv7nG1J1fzjRSF1Dq9nA96WY97UbfytSGI9uhMjataAAACM5JREFU47/8HZlcymIbupHslWOyT9qSv4Jl77opywlX5k98IiIiAihpk2AQFglNBrmEZs9al9B4FCsYwf3XVD3/uVu/g9UfQuuHoFS1fAg2D1o/BAe3wNzn3fSzohWhWid/RyWXMmNcQZKvHneVSsvUyvq4Y/vcaHeZOtDuyfyNUURERFQ9UoJE0u0QXhAWDs/5OWmp8MVDrkJj64d9F5u3GANdhrlRjKO7VOZf8ke9myEk3I22ZcVa+PxPcOqIm7r725pLERERyTdK2iQ4FIyBxH6umuLR3Tk7Z9HrsG89dHrRFV0IBmER0GsMdHwBGt/l72jkclCoBNS4DlaPh7SUc/eveB9+mgbX/CPTKLeIiIjkHyVtEjyaDYaMNFj8ZvbHHt4Oc1+A6tdB9Y6+j82bChSDZvcET6IpwS+xP5w8CD99mXn7gU0w/XFIaANNB/snNhEREVHSJkEk5gqo2cUVQ0g9fuFjpz/mpnV1ej5/YhMJZpWvgui4zFMk09Ncef/QMOj6BoTonwsRERF/0b/CElxa/AlOHYaVH5z/mOSvYP1UaPMXt55NRC4sJBQa9IWNM+HIDrft25dh5zK3zrJorH/jExERucwpaZPgUqEJxDWBRcNdL7OzpZ6AaY9AyerQfGj+xycSrBJvASysGgc7lrnpxXV7QZ3u/o5MRETksqekTYJPi6Gup9n6qefu++4VOPwzXPeyK+ohIjlTPN6tXVsxxk2LjC4PnV/0d1QiIiKCkjYJRjW6uDeYC17LvH3/Bpj/X1fCPKG1X0ITCWoNB8CRbXBwM3Qb4YriiIiIiN8paZPgExIKzYbAjiWwbbHbZi1MexjCCkD7p/0bn0iwqtHFTS1u+zjEt/J3NCIiIuKhpE2CU+ItEFUMFr7qXq+dBJvnwNVPQpEyfg1NJGiFR8GQxdD2UX9HIiIiImdQ0ibBKaIQJN0O66bCrtUw/Qko18BtE5GLZ4y/IxAREZGzKGmT4NX0bggJg9E3wLE90OUVN3VSREREROQSoqRNgleRslCvl+vblnQ7xDbyd0QiIiIiIl4X5u8ARPKkzV9cEZKrn/R3JCIiIiIiPpHtSJsxJsoYs8QYs9oYs9YY80/P9gRjzGJjzAZjzARjTIRne6Tn9UbP/njf/ghyWSseD93egALF/R2JiIiIiIhP5GR6ZArQzlpbH2gAdDTGNANeAIZZa6sCh4A7PMffARyy1lYBhnmOExERERERkYuQbdJmnWOel+GeLwu0Az72bB8NdPV8f6PnNZ79VxujcmQiIiIiIiIXI0eFSIwxocaYVcBeYAawCThsrU3zHLIDiPV8HwtsB/DsPwKU8GbQIiIiIiIil4scJW3W2nRrbQMgDmgC1MzqMM+fWY2q2bM3GGMGGWOWGWOW7du3L6fxioiIiIiIXFZyVfLfWnsYmAM0A4oZY36rPhkH/OL5fgdQAcCzvyhwMItrjbTWJllrk0qVKnVx0YuIiIiIiFziclI9spQxppjn+wLANcA6YDbQw3PYrcAUz/efeV7j2T/LWnvOSJuIiIiIiIhkLyd92soBo40xobgkb6K1dqox5kdgvDHmGWAl8I7n+HeAMcaYjbgRtt4+iFtEREREROSykG3SZq1dAyRmsX0zbn3b2dtPAT29Ep2IiIiIiMhlLldr2kRERERERCR/KWkTEREREREJYEraREREREREApiSNhERERERkQCmpE1ERERERCSAmUBooWaM2Qf87O84slAS2O/vIETyQM+wBDs9wxLs9AxLsNMznH8qWWtLZbUjIJK2QGWMWWatTfJ3HCIXS8+wBDs9wxLs9AxLsNMzHBg0PVJERERERCSAKWkTEREREREJYEraLmykvwMQySM9wxLs9AxLsNMzLMFOz3AA0Jo2ERERERGRAKaRNhERERERkQCmpO08jDEdjTE/GWM2GmMe83c8ItkxxrxrjNlrjPnhjG0xxpgZxpgNnj+L+zNGkQsxxlQwxsw2xqwzxqw1xtzv2a7nWIKCMSbKGLPEGLPa8wz/07M9wRiz2PMMTzDGRPg7VpHzMcaEGmNWGmOmel7r+Q0AStqyYIwJBYYDnYBaQB9jTC3/RiWSrVFAx7O2PQbMtNZWBWZ6XosEqjTgIWttTaAZMMTzu1fPsQSLFKCdtbY+0ADoaIxpBrwADPM8w4eAO/wYo0h27gfWnfFaz28AUNKWtSbARmvtZmttKjAeuNHPMYlckLV2HnDwrM03AqM9348GuuZrUCK5YK3dZa1d4fn+KO5NQyx6jiVIWOeY52W458sC7YCPPdv1DEvAMsbEAdcBb3teG/T8BgQlbVmLBbaf8XqHZ5tIsCljrd0F7g0xUNrP8YjkiDEmHkgEFqPnWIKIZ2rZKmAvMAPYBBy21qZ5DtF7Cglk/wH+AmR4XpdAz29AUNKWNZPFNpXZFBHJB8aYwsAnwAPW2l/9HY9Iblhr0621DYA43Mydmlkdlr9RiWTPGNMF2GutXX7m5iwO1fPrB2H+DiBA7QAqnPE6DvjFT7GI5MUeY0w5a+0uY0w53Ce/IgHLGBOOS9g+sNZO8mzWcyxBx1p72BgzB7c+s5gxJswzWqH3FBKoWgI3GGM6A1FANG7kTc9vANBIW9aWAlU91XIigN7AZ36OSeRifAbc6vn+VmCKH2MRuSDP2ol3gHXW2lfO2KXnWIKCMaaUMaaY5/sCwDW4tZmzgR6ew/QMS0Cy1j5urY2z1sbj3vvOstbegp7fgKDm2ufh+ZThP0Ao8K619lk/hyRyQcaYD4G2QElgD/B34FNgIlAR2Ab0tNaeXaxEJCAYY1oB3wLf88d6iidw69r0HEvAM8bUwxVqCMV9MD7RWvuUMeYKXFGzGGAl0M9am+K/SEUuzBjTFnjYWttFz29gUNImIiIiIiISwDQ9UkREREREJIApaRMREREREQlgStpEREREREQCmJI2ERERERGRAKakTUREREREJIApaRMREREREQlgStpEREREREQCmJI2ERERERGRAPb/pG1O+OrXFfkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Get our predictions\n",
    "dense_forecast = model_forecast(\n",
    "    model,\n",
    "    np_ts[split_time - window_size:-1],\n",
    "    window_size)[:, 0]\n",
    "\n",
    "plt.plot(dense_forecast)\n",
    "plt.plot(x_valid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "23.682121"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# MAE for validation set using a simple Dense model (already much better than baseline)\n",
    "keras.metrics.mean_absolute_error(x_valid, dense_forecast).numpy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# RNNS using Pandas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd_ts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>#Passengers</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Month</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1949-01-01</th>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-02-01</th>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-03-01</th>\n",
       "      <td>132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-04-01</th>\n",
       "      <td>129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-05-01</th>\n",
       "      <td>121</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            #Passengers\n",
       "Month                  \n",
       "1949-01-01          112\n",
       "1949-02-01          118\n",
       "1949-03-01          132\n",
       "1949-04-01          129\n",
       "1949-05-01          121"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = df.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Mean Is: 218.36\n",
      "Training Std Is: 73.84842855470927\n"
     ]
    }
   ],
   "source": [
    "SPLIT_TIME = 100\n",
    "train_mean = data[:SPLIT_TIME].mean()\n",
    "train_std = data[:SPLIT_TIME].std()\n",
    "print(\"Training Mean Is:\", train_mean)\n",
    "print(\"Training Std Is:\", train_std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "standardized_data = (data - train_mean) / train_std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1da40ce8fc8>]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(standardized_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "def windowed_data(data, start_index, end_index, history_size, target_size):\n",
    "    data = []\n",
    "    labels = []\n",
    "    \n",
    "    start_index = start_index + history_size\n",
    "    \n",
    "    if end_index is None:\n",
    "        end_index = len(dataset) - target_size\n",
    "        \n",
    "    for i in range(start_index, end_index):\n",
    "        indices = range(i-history_size, i)\n",
    "        data.append(np.reshape(data[indices], (history_size, 1)))\n",
    "        labels.append(data[i+target_size])\n",
    "        \n",
    "    return np.array(data), np.array(labels)\n",
    "\n",
    "def univariate_data(dataset, start_index, end_index, history_size, target_size):\n",
    "    data = []\n",
    "    labels = []\n",
    "\n",
    "    start_index = start_index + history_size\n",
    "    if end_index is None:\n",
    "        end_index = len(dataset) - target_size\n",
    "\n",
    "    for i in range(start_index, end_index):\n",
    "        indices = range(i-history_size, i)\n",
    "        # Reshape data from (history_size,) to (history_size, 1)\n",
    "        data.append(np.reshape(dataset[indices], (history_size, 1)))\n",
    "        labels.append(dataset[i+target_size])\n",
    "        \n",
    "    return np.array(data), np.array(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "HISTORY_SIZE = 10\n",
    "TARGET_SIZE = 0\n",
    "\n",
    "x_train, y_train = univariate_data(standardized_data[:SPLIT_TIME],\n",
    "                        0,\n",
    "                        SPLIT_TIME,\n",
    "                        HISTORY_SIZE,\n",
    "                        TARGET_SIZE)\n",
    "\n",
    "x_val, y_val = univariate_data(standardized_data[SPLIT_TIME:],\n",
    "                        SPLIT_TIME,\n",
    "                        44,\n",
    "                        HISTORY_SIZE,\n",
    "                        TARGET_SIZE)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(90, 10, 1)\n",
      "(0,)\n",
      "(44, 1)\n"
     ]
    }
   ],
   "source": []
  },
  {
   "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": "3.7.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}