exploration.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import os\n",
    "import preprocessing\n",
    "import json\n",
    "import datetime\n",
    "import dateutil\n",
    "import scipy.stats\n",
    "import pandas as pd\n",
    "import yfinance as yf\n",
    "from yahoofinancials import YahooFinancials\n",
    "\n",
    "from article import Article\n",
    "\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preparing the data for Bert"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "1 Failed download:\n",
      "- BRK.B: No data found, symbol may be delisted\n",
      "\n",
      "1 Failed download:\n",
      "- BF.B: No data found for this date range, symbol may be delisted\n"
     ]
    }
   ],
   "source": [
    "snp = yf.download('SNP',\n",
    "                      start='2020-01-01',\n",
    "                      end='2021-01-05',\n",
    "                      progress=False)\n",
    "snp['Delta'] = -snp['Adj Close'].diff(-2) / (snp['Adj Close'])\n",
    "\n",
    "table = pd.read_html('https://en.wikipedia.org/wiki/List_of_S%26P_500_companies')\n",
    "df = table[0]\n",
    "symbols = df['Symbol'].to_numpy()\n",
    "symbols\n",
    "deltas = []\n",
    "for stock in symbols:\n",
    "    returns = yf.download(stock,\n",
    "                      start='2020-01-01',\n",
    "                      end='2021-01-05',\n",
    "                      progress=False)\n",
    "    if returns[returns.isna().any(axis=1)].shape[0] > 0:\n",
    "        print('NA at: ' + stock)\n",
    "    returns['Delta'] = (-returns['Adj Close'].diff(-2) / returns['Adj Close']) - snp['Delta']\n",
    "    if returns[returns['Delta'] == -np.inf].shape[0] > 0:\n",
    "        print('-inf found at: ' + stock)\n",
    "    deltas.append(returns['Delta'].values[:-1])\n",
    "all_deltas = np.array([item for sublist in deltas for item in sublist])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(all_deltas, bins=100)\n",
    "all_deltas = np.nan_to_num(all_deltas)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.displot(all_deltas, bins=100)\n",
    "plt.grid()\n",
    "plt.xlabel('corrected relative returns')\n",
    "plt.savefig('plots/returns_dist.png')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DescribeResult(nobs=127861, minmax=(-0.5213978492917866, 0.6808071553099578), mean=0.002936975741777077, variance=0.002141514687782241, skewness=-0.06296098673018366, kurtosis=7.591148694343385)"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scipy.stats.describe(all_deltas)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.03226775858667886"
      ]
     },
     "execution_count": 195,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(np.abs(all_deltas))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "1 Failed download:\n",
      "- BF.B: No data found for this date range, symbol may be delisted\n",
      "\n",
      "1 Failed download:\n",
      "- BRK.B: No data found, symbol may be delisted\n",
      "Train DF of size 4954\n",
      "oldest article: 2020-01-14\n",
      "save DF to: /Users/marcglettig/MA5/Machinelearning_for_finance/Fin-418-NLFF/data/tabular_data/data_.tsv\n"
     ]
    }
   ],
   "source": [
    "data_df = preprocessing.format_to_bert()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>label</th>\n",
       "      <th>date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>cintas earns 2021 military friendly and milita...</td>\n",
       "      <td>decrease</td>\n",
       "      <td>2020-11-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>cintas announces evs technician winners for na...</td>\n",
       "      <td>not relevant</td>\n",
       "      <td>2020-09-29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>cintas celebrates 2020 custodian of the year w...</td>\n",
       "      <td>not relevant</td>\n",
       "      <td>2020-09-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>cintas names bancroft park's self-cleaning res...</td>\n",
       "      <td>not relevant</td>\n",
       "      <td>2020-11-19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>cintas corporation announces fiscal 2021 secon...</td>\n",
       "      <td>not relevant</td>\n",
       "      <td>2020-12-22</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                text         label       date\n",
       "0  cintas earns 2021 military friendly and milita...      decrease 2020-11-09\n",
       "1  cintas announces evs technician winners for na...  not relevant 2020-09-29\n",
       "2  cintas celebrates 2020 custodian of the year w...  not relevant 2020-09-30\n",
       "3  cintas names bancroft park's self-cleaning res...  not relevant 2020-11-19\n",
       "4  cintas corporation announces fiscal 2021 secon...  not relevant 2020-12-22"
      ]
     },
     "execution_count": 175,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(data_df['date'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.displot(data_df['date'], bins=12)\n",
    "plt.ylabel('Article Count')\n",
    "plt.xlabel('Release Date')\n",
    "plt.savefig('plots/dates_dist.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<function matplotlib.pyplot.show(close=None, block=None)>"
      ]
     },
     "execution_count": 187,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(np.vectorize(len)(data_df['text'].apply(lambda x: x.split()).values), bins=100)\n",
    "plt.show"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.displot(np.vectorize(len)(data_df['text'].apply(lambda x: x.split()).values), bins=30)\n",
    "plt.xlabel('Number of Words')\n",
    "plt.ylabel('Article count')\n",
    "plt.savefig('plots/text_length_dist.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "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>label</th>\n",
       "      <th>text</th>\n",
       "      <th>date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>decrease</td>\n",
       "      <td>425</td>\n",
       "      <td>425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>increase</td>\n",
       "      <td>529</td>\n",
       "      <td>529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>not relevant</td>\n",
       "      <td>4000</td>\n",
       "      <td>4000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          label  text  date\n",
       "0      decrease   425   425\n",
       "1      increase   529   529\n",
       "2  not relevant  4000  4000"
      ]
     },
     "execution_count": 245,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped = data_df.groupby('label').count().reset_index()\n",
    "grouped"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Seaborn Horizontal barplot\n",
    "sns.set_style(\"whitegrid\")\n",
    "bar,ax = plt.subplots(figsize=(10,6))\n",
    "ax = sns.barplot(x='text', y='label', data=grouped, palette=\"muted\",orient='h' )\n",
    "\n",
    "ax.set_xlabel (\"Count\")\n",
    "ax.set_ylabel (\"Label\")\n",
    "plt.show()\n",
    "bar.savefig(\"plots/label_dist.png\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "source": [
    "## Looking at the ARIMA and LSTM data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "def to_float(string: str):\n",
    "    if type(string)==str:\n",
    "        string = string.strip('[').strip(']')\n",
    "    return float(string)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "arima_data = pd.read_csv('data/tabular_data/prediction_arima_withdate.csv')\n",
    "arima_data['Date'] = arima_data['Date'].apply(lambda x: dateutil.parser.parse(x))\n",
    "arima_data.set_index('Date', inplace=True)\n",
    "for col in arima_data.columns:\n",
    "    arima_data[col] = arima_data[col].apply(lambda x: to_float(x))\n",
    "lstm_data = pd.read_csv('data/tabular_data/lstm_pred.csv')\n",
    "lstm_data['Date'] = lstm_data['Date'].apply(lambda x: dateutil.parser.parse(x))\n",
    "lstm_data.set_index('Date', inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "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>A</th>\n",
       "      <th>AAL</th>\n",
       "      <th>AAP</th>\n",
       "      <th>AAPL</th>\n",
       "      <th>ABBV</th>\n",
       "      <th>ABC</th>\n",
       "      <th>ABMD</th>\n",
       "      <th>ABT</th>\n",
       "      <th>ACN</th>\n",
       "      <th>ADBE</th>\n",
       "      <th>...</th>\n",
       "      <th>XLNX</th>\n",
       "      <th>XOM</th>\n",
       "      <th>XRAY</th>\n",
       "      <th>XRX</th>\n",
       "      <th>XYL</th>\n",
       "      <th>YUM</th>\n",
       "      <th>ZBH</th>\n",
       "      <th>ZBRA</th>\n",
       "      <th>ZION</th>\n",
       "      <th>ZTS</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-10-01</th>\n",
       "      <td>-0.000467</td>\n",
       "      <td>0.001045</td>\n",
       "      <td>-0.005775</td>\n",
       "      <td>-0.000457</td>\n",
       "      <td>-0.000949</td>\n",
       "      <td>-0.002937</td>\n",
       "      <td>0.002628</td>\n",
       "      <td>-0.005797</td>\n",
       "      <td>0.000515</td>\n",
       "      <td>0.002799</td>\n",
       "      <td>...</td>\n",
       "      <td>0.004087</td>\n",
       "      <td>-0.003514</td>\n",
       "      <td>-0.002804</td>\n",
       "      <td>-0.003742</td>\n",
       "      <td>0.003229</td>\n",
       "      <td>0.003044</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.008892</td>\n",
       "      <td>-0.011674</td>\n",
       "      <td>-0.002351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-02</th>\n",
       "      <td>0.000798</td>\n",
       "      <td>0.002571</td>\n",
       "      <td>0.002972</td>\n",
       "      <td>0.001425</td>\n",
       "      <td>0.002487</td>\n",
       "      <td>0.008576</td>\n",
       "      <td>0.005977</td>\n",
       "      <td>0.005118</td>\n",
       "      <td>0.000418</td>\n",
       "      <td>-0.004923</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.008521</td>\n",
       "      <td>0.001452</td>\n",
       "      <td>0.002567</td>\n",
       "      <td>0.000535</td>\n",
       "      <td>0.003431</td>\n",
       "      <td>0.001046</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.008651</td>\n",
       "      <td>-0.006812</td>\n",
       "      <td>0.003833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-05</th>\n",
       "      <td>0.004897</td>\n",
       "      <td>0.003448</td>\n",
       "      <td>0.001306</td>\n",
       "      <td>0.012548</td>\n",
       "      <td>0.001638</td>\n",
       "      <td>-0.000697</td>\n",
       "      <td>0.006376</td>\n",
       "      <td>0.002525</td>\n",
       "      <td>0.005720</td>\n",
       "      <td>0.023602</td>\n",
       "      <td>...</td>\n",
       "      <td>0.017500</td>\n",
       "      <td>-0.002871</td>\n",
       "      <td>-0.001480</td>\n",
       "      <td>-0.005741</td>\n",
       "      <td>-0.006484</td>\n",
       "      <td>-0.003212</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.005287</td>\n",
       "      <td>-0.002960</td>\n",
       "      <td>0.002550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-06</th>\n",
       "      <td>-0.007757</td>\n",
       "      <td>-0.001476</td>\n",
       "      <td>-0.000092</td>\n",
       "      <td>-0.004952</td>\n",
       "      <td>-0.004982</td>\n",
       "      <td>-0.000617</td>\n",
       "      <td>0.003250</td>\n",
       "      <td>-0.006410</td>\n",
       "      <td>-0.002637</td>\n",
       "      <td>-0.003778</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.010576</td>\n",
       "      <td>-0.006604</td>\n",
       "      <td>-0.000384</td>\n",
       "      <td>-0.003599</td>\n",
       "      <td>0.005156</td>\n",
       "      <td>0.003562</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.005182</td>\n",
       "      <td>0.003351</td>\n",
       "      <td>0.001978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-07</th>\n",
       "      <td>0.005789</td>\n",
       "      <td>-0.010353</td>\n",
       "      <td>-0.000342</td>\n",
       "      <td>0.011733</td>\n",
       "      <td>0.006782</td>\n",
       "      <td>0.004897</td>\n",
       "      <td>0.002135</td>\n",
       "      <td>0.009166</td>\n",
       "      <td>0.005879</td>\n",
       "      <td>0.011089</td>\n",
       "      <td>...</td>\n",
       "      <td>0.003339</td>\n",
       "      <td>-0.001932</td>\n",
       "      <td>0.003242</td>\n",
       "      <td>0.001542</td>\n",
       "      <td>-0.001411</td>\n",
       "      <td>-0.000989</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000187</td>\n",
       "      <td>0.006818</td>\n",
       "      <td>0.004590</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 501 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A       AAL       AAP      AAPL      ABBV       ABC  \\\n",
       "Date                                                                     \n",
       "2020-10-01 -0.000467  0.001045 -0.005775 -0.000457 -0.000949 -0.002937   \n",
       "2020-10-02  0.000798  0.002571  0.002972  0.001425  0.002487  0.008576   \n",
       "2020-10-05  0.004897  0.003448  0.001306  0.012548  0.001638 -0.000697   \n",
       "2020-10-06 -0.007757 -0.001476 -0.000092 -0.004952 -0.004982 -0.000617   \n",
       "2020-10-07  0.005789 -0.010353 -0.000342  0.011733  0.006782  0.004897   \n",
       "\n",
       "                ABMD       ABT       ACN      ADBE  ...      XLNX       XOM  \\\n",
       "Date                                                ...                       \n",
       "2020-10-01  0.002628 -0.005797  0.000515  0.002799  ...  0.004087 -0.003514   \n",
       "2020-10-02  0.005977  0.005118  0.000418 -0.004923  ... -0.008521  0.001452   \n",
       "2020-10-05  0.006376  0.002525  0.005720  0.023602  ...  0.017500 -0.002871   \n",
       "2020-10-06  0.003250 -0.006410 -0.002637 -0.003778  ... -0.010576 -0.006604   \n",
       "2020-10-07  0.002135  0.009166  0.005879  0.011089  ...  0.003339 -0.001932   \n",
       "\n",
       "                XRAY       XRX       XYL       YUM  ZBH      ZBRA      ZION  \\\n",
       "Date                                                                          \n",
       "2020-10-01 -0.002804 -0.003742  0.003229  0.003044  NaN  0.008892 -0.011674   \n",
       "2020-10-02  0.002567  0.000535  0.003431  0.001046  NaN -0.008651 -0.006812   \n",
       "2020-10-05 -0.001480 -0.005741 -0.006484 -0.003212  NaN  0.005287 -0.002960   \n",
       "2020-10-06 -0.000384 -0.003599  0.005156  0.003562  NaN -0.005182  0.003351   \n",
       "2020-10-07  0.003242  0.001542 -0.001411 -0.000989  NaN  0.000187  0.006818   \n",
       "\n",
       "                 ZTS  \n",
       "Date                  \n",
       "2020-10-01 -0.002351  \n",
       "2020-10-02  0.003833  \n",
       "2020-10-05  0.002550  \n",
       "2020-10-06  0.001978  \n",
       "2020-10-07  0.004590  \n",
       "\n",
       "[5 rows x 501 columns]"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arima_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "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>A</th>\n",
       "      <th>AAL</th>\n",
       "      <th>AAP</th>\n",
       "      <th>AAPL</th>\n",
       "      <th>ABBV</th>\n",
       "      <th>ABC</th>\n",
       "      <th>ABMD</th>\n",
       "      <th>ABT</th>\n",
       "      <th>ACN</th>\n",
       "      <th>ADBE</th>\n",
       "      <th>...</th>\n",
       "      <th>XLNX</th>\n",
       "      <th>XOM</th>\n",
       "      <th>XRAY</th>\n",
       "      <th>XRX</th>\n",
       "      <th>XYL</th>\n",
       "      <th>YUM</th>\n",
       "      <th>ZBH</th>\n",
       "      <th>ZBRA</th>\n",
       "      <th>ZION</th>\n",
       "      <th>ZTS</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-10-08</th>\n",
       "      <td>0.044484</td>\n",
       "      <td>-0.024163</td>\n",
       "      <td>-0.011208</td>\n",
       "      <td>-0.007579</td>\n",
       "      <td>-0.006428</td>\n",
       "      <td>-0.021936</td>\n",
       "      <td>-0.032452</td>\n",
       "      <td>-0.042213</td>\n",
       "      <td>-0.015041</td>\n",
       "      <td>-0.017350</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.009521</td>\n",
       "      <td>-0.022762</td>\n",
       "      <td>-0.003336</td>\n",
       "      <td>0.023142</td>\n",
       "      <td>-0.004740</td>\n",
       "      <td>-0.009547</td>\n",
       "      <td>-0.000442</td>\n",
       "      <td>0.000020</td>\n",
       "      <td>0.001468</td>\n",
       "      <td>0.000158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-09</th>\n",
       "      <td>0.008696</td>\n",
       "      <td>-0.039558</td>\n",
       "      <td>-0.061532</td>\n",
       "      <td>0.005215</td>\n",
       "      <td>0.004495</td>\n",
       "      <td>0.009758</td>\n",
       "      <td>0.010226</td>\n",
       "      <td>0.003644</td>\n",
       "      <td>0.003570</td>\n",
       "      <td>-0.005743</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.010558</td>\n",
       "      <td>-0.007619</td>\n",
       "      <td>-0.007586</td>\n",
       "      <td>0.000244</td>\n",
       "      <td>-0.005411</td>\n",
       "      <td>-0.002303</td>\n",
       "      <td>0.000318</td>\n",
       "      <td>0.000384</td>\n",
       "      <td>0.000276</td>\n",
       "      <td>-0.001689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-12</th>\n",
       "      <td>0.016926</td>\n",
       "      <td>0.058434</td>\n",
       "      <td>0.007241</td>\n",
       "      <td>0.007438</td>\n",
       "      <td>0.001088</td>\n",
       "      <td>0.033377</td>\n",
       "      <td>0.043008</td>\n",
       "      <td>0.024844</td>\n",
       "      <td>0.007527</td>\n",
       "      <td>0.016102</td>\n",
       "      <td>...</td>\n",
       "      <td>0.006034</td>\n",
       "      <td>0.014522</td>\n",
       "      <td>0.025416</td>\n",
       "      <td>0.048479</td>\n",
       "      <td>0.023668</td>\n",
       "      <td>0.017149</td>\n",
       "      <td>0.045866</td>\n",
       "      <td>0.013904</td>\n",
       "      <td>0.044396</td>\n",
       "      <td>0.013820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-13</th>\n",
       "      <td>0.000088</td>\n",
       "      <td>-0.048078</td>\n",
       "      <td>-0.017773</td>\n",
       "      <td>-0.016161</td>\n",
       "      <td>-0.007820</td>\n",
       "      <td>-0.012785</td>\n",
       "      <td>0.008188</td>\n",
       "      <td>-0.021457</td>\n",
       "      <td>-0.006942</td>\n",
       "      <td>-0.025947</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.010322</td>\n",
       "      <td>-0.016507</td>\n",
       "      <td>-0.001241</td>\n",
       "      <td>-0.004664</td>\n",
       "      <td>-0.004852</td>\n",
       "      <td>0.009293</td>\n",
       "      <td>0.002147</td>\n",
       "      <td>-0.004090</td>\n",
       "      <td>-0.011118</td>\n",
       "      <td>-0.003188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-14</th>\n",
       "      <td>0.004007</td>\n",
       "      <td>0.008282</td>\n",
       "      <td>0.010764</td>\n",
       "      <td>-0.004441</td>\n",
       "      <td>0.006890</td>\n",
       "      <td>0.020084</td>\n",
       "      <td>0.023160</td>\n",
       "      <td>0.025942</td>\n",
       "      <td>0.046504</td>\n",
       "      <td>0.022754</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.010469</td>\n",
       "      <td>-0.004577</td>\n",
       "      <td>0.007588</td>\n",
       "      <td>0.015359</td>\n",
       "      <td>0.015192</td>\n",
       "      <td>0.009761</td>\n",
       "      <td>0.018612</td>\n",
       "      <td>0.014536</td>\n",
       "      <td>0.011244</td>\n",
       "      <td>-0.001473</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 501 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A       AAL       AAP      AAPL      ABBV       ABC  \\\n",
       "Date                                                                     \n",
       "2020-10-08  0.044484 -0.024163 -0.011208 -0.007579 -0.006428 -0.021936   \n",
       "2020-10-09  0.008696 -0.039558 -0.061532  0.005215  0.004495  0.009758   \n",
       "2020-10-12  0.016926  0.058434  0.007241  0.007438  0.001088  0.033377   \n",
       "2020-10-13  0.000088 -0.048078 -0.017773 -0.016161 -0.007820 -0.012785   \n",
       "2020-10-14  0.004007  0.008282  0.010764 -0.004441  0.006890  0.020084   \n",
       "\n",
       "                ABMD       ABT       ACN      ADBE  ...      XLNX       XOM  \\\n",
       "Date                                                ...                       \n",
       "2020-10-08 -0.032452 -0.042213 -0.015041 -0.017350  ... -0.009521 -0.022762   \n",
       "2020-10-09  0.010226  0.003644  0.003570 -0.005743  ... -0.010558 -0.007619   \n",
       "2020-10-12  0.043008  0.024844  0.007527  0.016102  ...  0.006034  0.014522   \n",
       "2020-10-13  0.008188 -0.021457 -0.006942 -0.025947  ... -0.010322 -0.016507   \n",
       "2020-10-14  0.023160  0.025942  0.046504  0.022754  ... -0.010469 -0.004577   \n",
       "\n",
       "                XRAY       XRX       XYL       YUM       ZBH      ZBRA  \\\n",
       "Date                                                                     \n",
       "2020-10-08 -0.003336  0.023142 -0.004740 -0.009547 -0.000442  0.000020   \n",
       "2020-10-09 -0.007586  0.000244 -0.005411 -0.002303  0.000318  0.000384   \n",
       "2020-10-12  0.025416  0.048479  0.023668  0.017149  0.045866  0.013904   \n",
       "2020-10-13 -0.001241 -0.004664 -0.004852  0.009293  0.002147 -0.004090   \n",
       "2020-10-14  0.007588  0.015359  0.015192  0.009761  0.018612  0.014536   \n",
       "\n",
       "                ZION       ZTS  \n",
       "Date                            \n",
       "2020-10-08  0.001468  0.000158  \n",
       "2020-10-09  0.000276 -0.001689  \n",
       "2020-10-12  0.044396  0.013820  \n",
       "2020-10-13 -0.011118 -0.003188  \n",
       "2020-10-14  0.011244 -0.001473  \n",
       "\n",
       "[5 rows x 501 columns]"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lstm_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "aapl_returns = yf.download('AAPL', start='2020-10-08', end='2021-01-01', progress=False)\n",
    "aapl_returns = aapl_returns['Adj Close'].diff(-1) / aapl_returns['Adj Close']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaUAAAEGCAYAAADFWoruAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAACR6ElEQVR4nOydd3xb5b3/34+2vPeKk9jZe5AQwip7lr3poEALpdDCLbeD/trbedtyO+4tnZSWllHKLoVS9gg7hLBCQhJnJ3a8pyRr6/n98ZwjS7Jky7a84vN+vfSydHR0zmON83m+4/l+hZQSAwMDAwODiYBpvAdgYGBgYGCgY4iSgYGBgcGEwRAlAwMDA4MJgyFKBgYGBgYTBkOUDAwMDAwmDJbxHsBYUlJSImtqasZ7GAYGBgaTinfffbdNSlk6FueaUqJUU1PDxo0bx3sYBgYGBpMKIcS+sTqX4b4zMDAwMJgwGKJkYGBgYDBhMETJwMDAwGDCMKViSgYGBocGwWCQ+vp6fD7feA/lkMLhcFBdXY3Vah23MRiiZGBgMOmor68nNzeXmpoahBDjPZxDAikl7e3t1NfXU1tbO27jMNx3BgYGkw6fz0dxcbEhSBlECEFxcfG4W5+GKBkYGExKDEHKPBPhPTVEycDAwGAoSAm97RCJjPdIDkkMUTIwMDAYCiEfdO0Hfw+PPfYYQgi2bdsGwN69e3E6naxYsYJFixZxxRVXEAwGAVi3bh1nnXUWAHfddRdCCF588cXoYfVjPfLII9Ftra2tWK1W/vjHP47hPzi+GKJkYGBgMBRkJPr3/vvv55hjjuGBBx6IPj179mw++OADPvroI+rr63nooYeSHmbp0qXcf//90ccPPPAAy5cvj9vn4YcfZu3atXH7HeoYomRgYGAwFDRRcrtdvPHGG9x5551xoqRjNptZs2YNDQ0NSQ9z7LHHsmHDBoLBIG63m507d7JixYq4fe6//35++ctfUl9fn/I4hxpGSriBgcGk5gf/2sLHB3syesxFVXl87+zFyZ+UEoB//uvfnH766cybN4+ioiLee+89ioqKorv5fD7efvttbrvttqSHEUJw8skn8+yzz9Ld3c0555zDnj17os8fOHCApqYm1qxZwyWXXMKDDz7IzTffnLl/coJiWEoGBgYGQ0GzlO5/6DEuu+wyAC677LKoi23Xrl2sWLGC4uJiZsyYwbJly1Ie6rLLLuOBBx7ggQce4PLLL4977oEHHuCSSy7pd/xDnXG1lIQQpwO3AWbgz1LKWxOeF9rzZwK9wJVSyve05wqAPwNLAAlcLaV8a+xGb2BgMBFIadGMGpL2ji5eevV1Nm+rQwhBOBxGCMH1118fjSk1NjZy/PHH88QTT3DOOeckPdKaNWvYvHkzTqeTefPmxT13//3309zczH333QfAwYMH2bFjB3Pnzh31/3A8GTdLSQhhBn4HnAEsAi4XQixK2O0MYK52uxb4Q8xztwHPSCkXAMuBraM+aAMDAwMZ4ZF/v8AVl1/Evn372Lt3LwcOHKC2tpb6+vrobpWVldx666389Kc/HfBwP/3pT/nJT34St2379u14PB4aGhrYu3cve/fu5Vvf+lbS2NWhxni679YAO6WUu6WUAeAB4NyEfc4F7pGK9UCBEKJSCJEHfAK4E0BKGZBSdo3h2A0MDKYqUnL/489y/lmnx22+8MIL+4nLeeedR29vL6+99lrKw51xxhmccMIJcdvuv/9+zj///H7HnwouPCG1oN2Yn1iIi4DTpZRf0B5/FjhCSvnlmH2eBG6VUr6uPX4R+CYQAu4APkZZSe8CN0kpPUnOcy3KymLGjBmr9u0bs15VBgYGo8TWrVtZuHDh+Jzc3Qo99ZBdBvnTxmcMo0iy91YI8a6UcvVYnH88LaVk9SwSFTLVPhbgMOAPUsqVgAe4JdlJpJR3SClXSylXl5aOSTdfAwODQxq9ksP4TOgPdcZTlOqB6TGPq4GDae5TD9RLKd/Wtj+CEikDAwOD0UX3LkmjzNBoMJ6i9A4wVwhRK4SwAZcBTyTs8wRwhVCsBbqllI1SyibggBBivrbfSShXnoGBgcHoEq3oYFhKo8G4pYRLKUNCiC8Dz6JSwv8ipdwihLhOe/524ClUOvhOVEr4VTGH+ApwnyZouxOeMzAwMBgdopaSIUqjwbiuU5JSPoUSnthtt8fcl8ANKV77ATAmgTeD8ePWDbeytGQpn5z1yfEeioGBRiThr0EmMSo6GExontnzDG80vDHewzAw6MOwlEYVQ5QMJjTBSBBfeHw7YRoYxKGJUc6M/uWDtm/fzvHHH8+KFStYuHAh1157Lc8++ywrVqxgxYoV5OTkMH/+fFasWMEVV1zBunXrEEJw5513Ro/x/vvvI4TgF7/4xZj9SxMJoyCrwYQmGAnSG+od72EYGPQxQNbdjTfeyFe/+lXOPVfVAfjoo49YunQpp512GgDHH388v/jFL1i9WkUe1q1bx9KlS3nwwQf5/Oc/DyRvYTGVMCwlgwlNMBLEFzIsJYMJxABuu8bGRqqrq6OPly5dOujhZsyYgc/no7m5GSklzzzzDGeccUZGhjoZMSwlgwmLlJJQJGSIksHAPH0LNH2U2WNWLIUzbk3xZGpL6atf/SonnngiRx11FKeeeipXXXUVBQUFg57uoosu4uGHH2blypUcdthh2O324Y37EMCwlAwmLMGIaiNtiJLBhCJqKfW3mK666iq2bt3KxRdfzLp161i7di1+v3/QQ15yySU8/PDD3H///f1aWEw1DEvJYOi8+COYdRzUfmJUTxMVJSPRwWAgUlo0o8QglRyqqqq4+uqrufrqq1myZAmbN29m1apVA76moqICq9XK888/z2233cabb76ZyRFPKgxRMhgaUsLr/we+rtEXpbASJW/IO6rnMTAYEgPElJ555hlOOukkrFYrTU1NtLe3M21aekVbf/jDH9LS0oLZbM7USCclhigZDA2/C2QY/O5RP5VuKRmiZDCxUJZSr9cXl9Rw8803U19fz0033YTD4QDg5z//ORUVFWkd9aijjsr8UCchhigZDA1fl/rrd436qWJjSlJKVCPiYRAJgzDBcF9vYBCLZilFGt6Hyv6p2//7v/+b8qXr1q2Le3z88cdz/PHH99vv+9///khGOKkxEh0Mhoa3U/0NjJ0oSSSBSGB4B5ESblsBG/+SuYEZTG2Mig6jiiFKBkPD26X+joX7TospwQgy8IJe6N4PHbszNCqDKY+M6adkCFPGMUTJYGhELaWxiynBCOJK/h7tYEZVCINMIVPcN8gEhigZDI1oTGmSiJJPFyUjWcIgQ8gI0UunYSllHEOUDIaG7r7LoKX0961/58frf9xve6woDdt9p1tKAc/wXm9gEIsuQiZdlIz2FZnGyL4zGBq6+87vUj/QDGS0bWjawPaO7f22x4nScBfQ+rq1gxmWkkEG0EVImIGQYSmNAoalZDA0dPcdMmPWhy/kixMgndhEh5HHlAxRMsgAmgh1uTz8/q6HmCgxpbvuuosvf/nL4z2MjGCIksHQ0C0lyJgLzxdOIUqZcN9FLSUj0cEgEyhLqavHze/veTip+y4cDo/1oA4pDFEyGBp6TAkyluzgC/kIhPuvQ8qMKBnZdwYZRLOUbvnvX7JrXz0rVh/B17/+ddatW8cJJ5zApz71KZYuXcrevXtZsmRJ9GW/+MUvogtid+3axemnn86qVas49thj2bZtW9wpIpEINTU1dHV1RbfNmTOH5uZm/vWvf3HEEUewcuVKTj75ZJqbm/sN8corr+SRRx6JPs7JyYne//nPf87hhx/OsmXL+N73vpeJdyTjGDElg6Hh6wIEyn2XmQW0/rB/UFEyUsINUvE/G/6HbR3bBt9xCCwoWsA313yz/xOaZXTr925h88fb+OCdt8CWzbp169iwYQObN2+mtraWvXv3pjz2tddey+23387cuXN5++23uf7663nppZeiz5tMJs4991wee+wxrrrqKt5++21qamooLy/nmGOOYf369Qgh+POf/8zPfvYzfvnLX6b1Pz333HPs2LGDDRs2IKXknHPO4dVXX+UTnxjdGpZDxRAlg6Hh7YTcSnAdzFipIT2mlFhKKG7x7LATHYyYkkEG0RMbhDn+MbBmzRpqa2sHfLnb7ebNN9/k4osvjm5L1tri0ksv5Yc//CFXXXUVDzzwAJdeeikA9fX1XHrppTQ2NhIIBAY9XyzPPfcczz33HCtXroyOZceOHYYoxSKEOB24DTADf5ZS3prwvNCePxPoBa6UUr4X87wZ2Ag0SCnPGrOBT2W83VC2QBOlzMWUJJKQDGEV1uj2jFpKgVGylHo7YN2tcOqPwDJ1G7ONJ0ktmlFDF6X+KeHZ2dnR+xaLhUik7zmfT02qIpEIBQUFfPDBBwOe5cgjj2Tnzp20trbyz3/+k+985zsAfOUrX+Hmm2/mnHPOYd26dUlr5MWeW0pJIBCI3v/Wt77FF7/4xSH9x2PNuMWUNEH5HXAGsAi4XAixKGG3M4C52u1a4A8Jz98EbB3loRroRMLg74b86epxhhId/CE1U4y1jGAUYkqjkb675xXY8Edo2pz5YxtMPDQRys3Lw+X2kCr7rry8nJaWFtrb2/H7/Tz55JMA5OXlUVtby8MPP6wOJyUffvhhv9cLITj//PO5+eabWbhwIcXFxQB0d3dHW2HcfffdSc9dU1PDu+++C8Djjz9OMKh+R6eddhp/+ctfcLvV77ahoYGWlpbhvAujyngmOqwBdkopd0spA8ADwLkJ+5wL3CMV64ECIUQlgBCiGvgk8OexHPSURs9ky9fK9WfIfadbQYlxpVAk1HfqkWbfISE0eAfQIaOnxQeNxblTAm1iU1xSytGHr2DJqrV8/etf77eb1Wrlu9/9LkcccQRnnXUWCxYsiD533333ceedd7J8+XIWL17M448/nvRUl156KX/729+irjtQ1cMvvvhijj32WEpKSpK+7pprruGVV15hzZo1vP3221EL7tRTT+VTn/oURx55JEuXLuWiiy7C5Rr9wspDZTzdd9OAAzGP64Ej0thnGtAI/Ar4BpA7ekM0iENPBy/InKUUjoQJSSU+iWnh+uMca84I3HfdffeDvWB1DO84qdBFabTcgwYTC91dZzLz99/9BApmQlYRQL8WFDfeeCM33nhjv0PU1tbyzDPPDHqq1atXIxOs+3PPPZdzz02cu6uMuyuvvBJQVtr69eujz/30pz+N3r/pppu46aabBj33eDKellKyUgCJtnDSfYQQZwEtUsp3Bz2JENcKITYKITa2trYOZ5wGOno6eN40QGQkpuQP91kvie0pdHdeji1n5IkOMDoZeIalNMVISHSYIItnDyXGU5Tqgekxj6uBg2nuczRwjhBiL8rtd6IQ4m/JTiKlvENKuVpKubq0tDRTY5+a+DRLyVkI9tyMuO9iLaBAgnstGAliERayLFkjS3RwFGgHHIUMPMNSmlpEywwZte9Gi/EUpXeAuUKIWiGEDbgMeCJhnyeAK4RiLdAtpWyUUn5LSlktpazRXveSlPIzYzr6qYhuKTkLwZaTkXVKcZaSvyfuuWAkiNVsxWlxDi+mJKWylHIrtQOOpqVkiNJYk+jaGqOTqr+m/inhhwLj8p4mMG6iJKUMAV8GnkVl0D0kpdwihLhOCHGdtttTwG5gJ/An4PpxGayBQo8pOQrAnpMR912s2IRiSxihWUomCw6LY3juu2AvyDDkVqjHo2HN6HE1owr5mOJwOGhvbx/7i2hcQVYOKVGSUtLe3o7DkeG46xAZ13VKUsqnUMITu+32mPsSuGGQY6wD1o3C8AwS0YuxOgs0SykDohRzjIAvQZTCQawmKw6Lg25fd+JL0zi4ZnnpojQa1ox+TMNSGlOqq6upr69nzOPE/h7lMeiyQXcLOPzg6BjbMYwiDoeD6urqcR2DUdHBIH28XWDNUotEM2Upeduj9wMJwhOIBLCarDjNTprD/Wt8DX5w7Xg55ervqMaUDEtpLLFarUOqZpAxXvkZvPxj+K92+O9j4Jib4aT/GvtxHMIMKkpCiFLgGqAmdn8p5dWjNyyDCYm3qy9pwJYLnr0jPqSvty16P1GUgpE+S2lYiQ56jCoaUzJEyWCEhPzKdWe2gNkGSWo2GoyMdCylx4HXgBcAoyb7VMbbqZIcQGXfZSLRISaOFAwkJDqEg9jMtuEnOkTdd7qlNArCobsfDffd1CDkA4sWczHbDVEaBdIRpSwp5VgWlzKYqPi6VDwJMue+i4kjBRJSzGMtpWElOugLZ3P0mNJoWEq98X8NDm1CfrDY1H2LbXSqhExx0sm+e1IIceaoj8Rg4uPt6rOUMpXoEFNxIRhIIUpmB76Qb+iZVmOR6GAsnp1ahP2GpTTKpCNKN6GEySeEcGm3nkFfZXDo4e3siynZc9QPcoQzxVhRCiSIXOw6pbAMJ+1OOyB6TCm7RMUBRsVS0lPCDUtpShDyq1gSGJbSKDGo+05KadSWM1DEuu9s2tfC7x5RywZ/jAswmGBtxKaEg6r+YNMvCGmNt1uJkS1HZQ2OhnAYKeFTi1CipWSIUqZJKyVcCHEOoHeCWielfHL0hmQwIQn51YU3GlPSRCngguziYR/WF4xZp5RwYQ9FQjgtTpwWp9o35CPfnj+Eg/eocQoBVmfmhSMU6HPfGNl3U4OQv28SZrGp74BBRhnUfSeEuBXlwvtYu92kbTOYSuglhmLddzDiZAdfjHUUSEj7jk10gGF0n/X3gCNP3bc6M+++i7XsMil4O54HT/vg+xmMPeEYUTIspVEhnZjSmcApUsq/SCn/ApyubTOYSkSrOcQkOsCIkx38IS/ZWv5CMBgvOtGYkrnPUhoSvh7QLStbduaTEXTryJqdOdfg+3+D+y6CjXcO6WXd/m7ebR60aL7BSIl131nshqU0CqRb+64g5v4Q/CcGhwz6eqJE990ILSVvyIdTmLEgCIb7i5Je+07tO0RLZ7QtJV2IckqV4I20Dlr9u/DkV9X9noYhvfS+rffx+Wc/j8fIAhxdYhMdzDbDUhoF0hGlnwLvCyHuEkLcDbwL/GR0h2Uw4YitEA4xltLIFtD6w37swowVE4GEH3iyRIch4esBuy5KWaMgSpogZ5epQp0jycRyNcODn1Hp64U16vEQqHfVE5Zh9nbvHf4YDAbHsJRGnUFFSUp5P7AW+Id2O1JK+cBoD8xgghFbIRxiYkojFKVIAKewYhUm1Q49xtroF1MasvuuGxyaYT8aiQ66+y5b69M13OOHAvDQFeo9vuzvUDwH3E1DOkSjpxGA3d27hzeGych798KTN4/tOcMxi2cNS2lUSClKQogF2t/DgEpUw70DQJW2zWAqkRhTyoT7Tkq8kRB2sxWbyUIA4rLYdFGKZt8NOdGhO8Z9Nwop4fpYc0rjHw+VZ26BA+vh3N9CxVJVQNbdMqRDTElR2vQgbH50bM8ZV2bIWKc0GgyUEn4zcC3wyyTPSeDEURmRQVr0Bnt56+BbnDTzpLE5YTT7Tk8c0FPCRyBKQS9+JA6zHRsRggJl3WhW2IgSHaRUVtxouu/0+E12mfZ4GKL37l0qqeHom2DpRWpbTjm4myESAdPgHvZwJExzr3L37e6aQqLUuk1NlsIhVSB1LEhcPBse4oJug0FJ+Y2XUl6r3T1DSnlC7A0j+27ceWbvM/zHuv+gwT20gPiw8XaqTDa946bZomaMI3Hf+brwCYHD4sBmshIUoq/dBCOMKQXcKs4Tl+gwWpZSWfzjdDmwAf79NZh9Ipz0vb7tuRUQCYE3vT497b52QpEQAHt69gxtDJMVTxt4tF5KuhU/FhiLZ0eddBId3kxzm8EY0q71IWrztg2yZzzfeu1b3LX5rqGfMLaag85I6995O/GZBHZLFhazjUCMKEkp+8WUhiRKet073VKyZY1iTKlE/R3K8Xsa4cHPQv40uPDOPrGHvv5PrvTiSrrrbn7hfA70HBh6OabJSMvWvvu9Y7imK64gawYTHbob+r6zU5yBYkoVQohVgFMIsVIIcZh2Ox7IGqsBGiSnW6sZ15nQrXUgAuEAz+x5huf3Pz/0E3o7+4uSPWdklpK3C78QOKzZ2Mz2OFEKyzASidVkxW5WixV9vW1Q91x6x9br3sXGlIK9mW1fHZt9B+nHrEJ+eOiz6r277O+QVRT/vF5ANs1kB12Ujpp2FCEZ4kDPgfTGMZlp3dZ3v3eMOr9KmVCQNYOJDvecAy/9KDPHmuQMZCmdBvwCqEbFlfTbV4H/N/pDMxiILn8XMDRR2tW1i5AMsatr19ArbsdWCNex544s0cHbqdx31mxsFkdfTAmis32r2YpJmFRPpQNvw98vhoPvD37sqKUUk30HKlCdKQK9YLL0vS/prBGSEp76GtS/A+f9HsoX998naimllxbepInX0VVHA1Mk2aHl4777Y2UpRULKJWzWywxpVcIzMdFxNUP7zpEf5xBgoJjS3Vr86Eop5YkxMaVzpZT/GMMxGiQhain50xelbR1qdukJemjyDC3lGF9XXzq4ji13ZO47PaZkz8NmccZZSlFRMlkBcJgdeDUh5qX/TuPYWmwqmhKuGfeZTHYIeFSlCJt27HQspY1/gffugWP/Exafl3wfXZTc6YlSo6eRHGsOS0uWAlNFlLZBXrW6n2bsbcTomXbRMkOaG2+k7SukVL+jnsaRHecQIZ2Y0iohRIH+QAhRKIRI46pgMJrollLXEIK8Wzv6/PA7u4Y4K4vtOqszYvddJz5hwm7PxWpxECRGlMIJomRx4PO7QJhg5wuw762Bj53MfQeZLZwa8GgVyLPV48FiSvvegqe/AXNPhRO+nXo/W5aKhQ1BlCqyK8iyZlGRXcGe7kM82UFKZSnVKMtwzCylqCjFLJ6N3T5cgr2AhJ6DIzvOIUI6onSGlLJLfyCl7MTIvht3ugPq4t3hS3+WuK1jG7PzZwNDFCUpNfddQfz2ESY6RHo7CZgEDlseVoudgMkczaTqZylZHHiDHlh8gbIkXvzhwG4T3VKyx2TfQWYtpWCipTSA4Pl61ALZgplwwZ/iExuSkVOedqJDk6eJyuxKAGblzzr0LSV3s/qeTFulBGKsYkp6/Ci6eFYTpZFaSvr3xt9tVJsnPVEyCyGiDXOEEE5g+A10YhBCnC6E2C6E2CmEuCXJ80II8Wvt+U36ol0hxHQhxMtCiK1CiC1CiJsyMZ7JxFDddxEZYXvHdtZUrqHUWTo0UQp4IBLs774bYUt0v5ZB6LA6sZlsBE3mpDElAIfZji8SgJJ58Imvw/43YdeLAxw8haWUyQy8gEcd16IL3gDHbt4CnhY49b/7i3sycivStpQSRWlP9x4iMpLWayclejypbCE4i8ZOlPR4ZNRS0sRppJZS7MTOcOGlJUp/A14UQnxeCHE18Dxw90hPLIQwA78DzgAWAZcLIRYl7HYGMFe7XQv8QdseAv5TSrkQVQLphiSvPWSRUkZFKV333QHXAXpDvSwoWsCcgjlDE6XEag46ttwRue98mpVnN9uxmq0ETKaUMSUnJnxCQMF0OOxzkD9DxZZSWUu+HtXgTxcj22jFlHLUAldr1sCzXL3AalFtesdO01Lyhrx0+jupzFGiVJtfizfkpdkztNp5k4oWFRt9zt/Cb/Kzxz6mZE60lEYoSrETuyEW4j0USaf23c+AHwMLgcXAj7RtI2UNsFNKuVtKGQAeAM5N2Odc4B6pWA8UCCEqpZSNUsr3tPG5gK3AtAyMaVLgCroIyzCQvvtOjyctKFrA7ILZ7O7anf5sOlqMtSB+uz1XubAiw5uV+7XjOi3KUopLdNBiSjaTugA4pVSilD9dzVCPv0Vl4W1L0W9SrxAuhHoctZRSC8furt1Du5gH3Mp9px9/IEtJjxfkVaV3bN1SGiSzS09YqchWaeS1+Ur0DmkXXsvHkFXCM01v8rA1OI4xJd1SypD7DsBlWEppta6QUj4tpfyalPI/pZTPZujc01C19HTq6S8sg+4jhKgBVgJvJzuJEOJaIcRGIcTG1tbWkY55QqBbSbm23LTdd9vat2ERFuYUzGFu4Vx8YR8NrjRnZdG2FUkSHWDYcSU9m85utmMz2+IqOugVCqLuu3AYr0lAvpZxtexSKJ6rrKVIOO6425p6eHPLbiKxXWrTiCl9dd1X+car30j/Hwj09omSbZDaej0NyrJ05KfeJ5acMiVyg1ii+hqlWPcdHOKi1LoNyhbS7m2nhwhyzEUpw5ZSrCgZyQ4DLp59XfvrEkL0xNxcQohMLD0WSbYlTgsH3EcIkQM8CvyHlDLpmKSUd0gpV0spV5eWlg57sBMJXZRq82rxBD2quvYgbOvYxuyC2djMNmYXqGSHHV070jthYoVwnRE2+vNrF1yH2YHVZCUIUVEKRNT/ZDGpmmaOUEBZSnnanMRsgRP+n7pAffRI3HEfe78Br6sTnym7b+MgKeERGaHeVc97Le+lf0EPePrcgtZBmgj2NKRvJQHk6AtoBy7MqltKuigVOYrIt+cfuhl4Uir3XdlCOv2dhAHPaLnv7jxVpfDrhBMsJXOmLKWYiYdhKQ24TukY7W+ulDIv5pYrpczLwLnrgekxj6uBxGlCyn2EEFaUIN031dZN6engNfk1QHoLaLd1bGN+0XyAaAberq5d6Z1Q/6HkVsZvH2GlcF9AzSMcFoeKKSFTp4QHfXjNlr5ZKsCi81RV7XU/iSuM+fbuDnJFL55kopQi7tPh64gK4T93/DO9f0CPKcHgllJ3gyoplC65+lqlgeNKjZ5GBILSLDXhEkIc2hl43fXqIl66IOq67gm4+lnLIyYUgANvQ0NMN99oTElfPJuhdUr6d9KWY1hKDGwpFQ10y8C53wHmCiFqhRA24DLgiYR9ngCu0LLw1gLdUspGIYQA7gS2Sin/NwNjmVRELSUtfqCLVCpae1tp97WzsGghADm2HKqyq9K3lLrr1Q9Rr/Gmo1+Qh5PsEIng02IwdrNdxZSIIH3doNW9g5hEh2AvPpHwdTWZ4MT/gs698P69AHj8IT5q6CaXXnqks2/fQdx3usWRa8vl8V2PD14/Tl/wOJSY0nAspUGSHRrdjZRmlUbfJ+jLwDsk0coLBUvmRX8HPSbiCvlmhF6tpmRsZl+/xbMZTnQonmOIEgPHlN4FNmp/W4E6YId2/90BXpcWUsoQ8GXgWVSiwkNSyi1CiOuEENdpuz0F7AZ2An8Crte2Hw18FjhRCPGBdpsya6eillJeDTB4skNskoPO7ILZ6VtK3fVqli8SvKm6pTSc7rP+HvyaJ9ZpcWLTXCEhGYaAp19KuNPnwpfMmTv3VJh+BLzycwj6eHdfJ+GIJFd46Qg5+vYbJCVcj818btHn6PB18OqBVwcefzgAMhyT3ZedOvsuFFBJC3oFgnSIWkoDJ17EpoPr1ObX0uHrGNLC6kmDlg7end/3P/eYTJlPC9fdpnGipKeEJ1hKI3bfaaJUMtdw3zGw+65WSjkLJRpnSylLpJTFwFmoDrQjRkr5lJRynpRytpTyx9q226WUt2v3pZTyBu35pVLKjdr216WUQkq5TEq5Qrs9lYkxTQa6/d0IBDPyZgCDu+/08kK6+w5gTuEc9nTviSYUDEhPQ18sJ5Zo99lhuO98XXi1XkG6pQREkx10UXrp43be3duGw9dDCNnfghFCWUuug7DxTjbs6cBsEhSYvLQFY5bTWWyqTl0KS6nRrS4GF827iDJnGY/uGKR5XKzLBQa2lNxNgByapeQoUDPxWEvJ74aXfwLBvvp9jZ7GpKIEh2iyQ8s2yK2knT53nRKlDCc7eHRLKea4upsu05ZSwKO+m4U1ahISTuM3eQiTTvbd4bEXfCnl08Bxozckg8Ho8neRa8ulxKncaYNl4G3r2EZ1TjW5emM+YE7BHIKRIPtd+wc/YXeDSsVOZCSJDt5O/JrlZbfYoxaRnhaux5RuX7eXx177AIf2Q03a6K/2WJh1PLz2Sz7cdYCllTlkyV4afbb4/QYQjkZPI06LkyJHEefOOZc3Dr4xcHq4/j+nk33XrWU5JhP2VAjR1+xP5+N/wiv/o2IdqOSMZJaSnoE34Vx4W/4JzR8PutuAtHwcF08CTZQynezg0Swlb39LqS3o4YSHTmCrR/tcM2Ep2XLUpEVG0l40faiSjii1CSG+I4SoEULMFEJ8GxjDBiYGiXT7u8m355Nvy0cg0rKUFhYvjNs2p2AOADs7B1lEGw4pKyRZkD6a6DAM951XFWMF5b7TYyKBBEupNyAQ3QdwaOt1UvZUOvG70NvOYY0PcuxMJyYkB/12fMGYAPgAjf70i7sQgvPnnE9ERnh81+Opx68LUDSmNED2nb4gciiJDqBceLEXqD2aS1Gz9vTkDH2Nkk5VThUOs2PiWUpPfAXe+t3wX9+xW0sHXxT3nR8dS0lbPuLt7FuHp4nPDk8Dbd42trm11SqZsJRsOZCrWdJT3IWXjihdDpQCj2m3Um2bwTjR7e+mwF6A2WQm354/YKKDK+DigOtAXDwJlItHIAaPK7mb1OwtP0k8ZISWki5K+jolgIDWviLqppNm7J4GnNpC35Qt0atX0V59Mp83PckxJUocXGRR3xkjQlbngIkOusUxPW86ayrW8I8d/0i9wDjqvkvDUhrqwlmdnPK+9hVSwu5X1H1NWBPTwXVMwkRNfs3EEqWARy1oHq54dO6Du89R1u7qq+ItJfMoxpRkpK+iiW4paTHUrpD2eWeizJAtG/K0zzETyQ4N78EHfx/5ccaBdCo6dEgpbwKOlVIeJqX8DynlGNX1MEhGl7+LfG1haKGjcMBEBz2elChKTouT6bnTB8/A665Xf5MF6a1OVcpnOJaSryvqvlPt0JPHlJBmsn2NOCODWErAU6WfJwcvh+1Ss3GXdLKvPVaUUicj6JW2dS6YewEN7gbeaXon+ckS3XfWbFUfMBzsv+9QF87q5Fb0pYS37ei7rwlrdOFsTmW/l9bm1U4s951u8Q1HlLrr4e6zlahd8U8omUuHrwOzMJNvy6fHbB09Swn6BE8Tn7ZAFwCdQe07MNKUcL9bxWd19+4wLSVfyNfXJ23jnfCv/0j+fZzgDCpKQoijhBAfAx9rj5cLIX4/6iMzSEmcKNkLB3TffdT2EQBLSpb0ey6tDDxdlJJZSkIMvyir1grdLMxYTda+mBLxMSUpLRQFm7FrWW6+cOomfU81F/GK/RNYd6nutC6y2N8xuKXkD/tp97XHidJJM04i15abOuEhmaUUuz2WoS6c1cmpUO6jkB/2vNK3XbOU9OSMREsJoLagloPug0NrIT+aRLPZhigeriZlIXk74bOPQeVyQLkuC+wFFDgK6LHaR1mUtGOH/WCy0OpTj7t0URqxpaRVm88qVgtyh1H/LhwJc8Y/zuD3H2qX5kCvGm9sh95JQjruu/9DdaFtB5BSfgh8YjQHdSjQ6Qlw1m9eY932gVfkD4cefw8F9gJAWUoDue82tW5ieu50ihz9l5bNKZjDvp59A1eEiIpSinjIcBv9ebvwmSzRVucDWUrTRBtWh0rqSOW+84fCvLe/k4/m3qCsNyBoyUlLlPSEhtiLu8Pi4KxZZ/Hivhej62Hi0GNTsdl3sdtjGerCWZ3YtPA9r6iLFsRZSk6Lkzxb/7Xss/JnIZHs69k39POOBnoW4VDEw92qBMnVBJ9+RLWq0OjwdVDkLCLPlke3xdZXdSRTuFv71op5Yywls502r8rM69QWf4988axb/Y6EUNbxMCqF73ftp83bxt8+/hvugLvve37wg5GNbRxIt/bdgYRNGV4+fejxi+e2s7mhh7d29f8RdvUG+P4TW+gNDD31MxQJ4Qq6BnTfdfUG2N/ei5SSD1s/ZFnpsqTHmlMwh7AMD+zm6WlQbid7bvLnh9voz9uJz+rAoZVsiVpKVif4uhJEqR1hVxULUs38N9V34w9FmLdwOaz8DADZ+SXsj3Xf2bKTikZi/TidC+ZeQCAS4MndSYq+6kIcu04JUlhKQ1w4q5NTThDo7dgDe16DOaeo7doFJzY5I5FoDbyuCRJX0i0lX1d6Kc+9HXDPudC1Hz79MMw4Iu7pTl8nRXYlSj1m8+hYSqXaEgr92CE/WPpEqcvfrZpOZiqmBCrZYRjuux2dyg3vDrp5pO6RvqSbxg9GNrZxIB1ROiCEOAqQQgibEOJrqMWuBinY3NDN3zeoVOv6zv4X0Ze3t3DXm3tZv3voPyR91h61lOyFdPu74wLyP3t2Oxf84U0a3I20edtYVpJClApVBt6ALrzuesirjs9ii2W4jf58XfgtNhxmJUpRS8me02cpScGcsjymiTZC1oEtpbe193JNbRGc8gM461dYSucksZTSF6UFRQtYVLyIf+z4R5+vXifRfZeqjNFwFs7q5JTztbISrlz/X+piPvtE5d7R3Xexa5T2vQn/+GI0hjAzbyYmYRow2SEUjvDZO9/m8Q/GoF1CbLmkwawab6cSpPadcPn9fR1mY+jwdVDk0ETJJDKb6BCJaKKkxWGjMSUfWOy09irXXpe/S30eGcm+075HeVXDct/VddZhEiZWlK7g3q33EtS/540fjmxs40A6onQdcAOqOnc9sEJ7bJCESETy3cc3U5xtY/n0gvjsL40DHUqotjcN/WKui1K+rc9SCsswrpiqCvWdXtrcfv69fT0Ay0uXJz1WbV4tFmEZuLdSdz0Nsoil33+WAx1JXFPDtpS68Jlt/S0lWxb4uvGHAkhp5uhpVvJEL70W5cpKZSm9vaeD+eW5FGXbVDXz1Vcxozib/R29RLQkCbVOqf/rdVEqzy7v99wFcy6grrOOj9sT1tekiiklit5wFs5qvNF7gJeys9gRUIVHqf2EaiioCfNB98G+ONjOF2DTA/DBfWo4ZhvTc6cPKErPbmnmtR1tPPXRGKQgx6a2D2TV+HrgbxeqWMhl98HsE5LuFnXf2fPoIZJZS8nXpap1FM0CU0wSRTgAFjvtWnNKJUr2ka9T0hMdQBOlxkFbliRS11lHTV4NX1z+RVp6W3gqorkWmzZPusW4A4qS1ojvV1LKT0spy6WUZVLKz0gpjXVKKXjs/Qbe29/FN09fwKLK3KSWkj5739E89Iu53gY9NqYE8aWG2t1q5rZu70bsZjvzCuclPZbVbGVG3owBRSnUVc8rzXaCYcnGfeoc2zq29cVZ7LnDTHTowmc294sp6aLU0esFaWZtsXqvulCCkSzRIRiO8O6+To6YFR83m1GUhT8UoVV7P1JZSs2eZkqcJdG09FjOmHUGDrODf+xIKGIS8KhZsiamWHX3XcLx9YWzQ4wpBSNBfrb5zwCEBDSXzlUpw9r/4A646fR3Mj13evx5190aFd7BMvD+/LoSrM0NmSj6PwiuNETJ74b7LlKz+4vvhrmnJN0tEA7gDroptBeSZ8vDJUPI2PVEI0V3NeaUqjhe1H3nw2e24Qq6sJlsdPu7CVtsI4spRcIQ8vbFJnMr1eMhloiq66xjXuE8jq46mrmFc7nL4kOabepYbXXDH984MKAoSSnDQKlWMNVgEFy+ID99ehsrZxRw4WHVVBdm0e4J9Isd6aJU1zJ0UdLrmeU7+rLvIL4oa7tb/Uh29mxhUfGiqBWSjIG60Ho9Liy+DjotZTitZjbVd7O/Zz+XP3k5X37xy8plONxEB18XfpM5ainpghC0KlHq8nqR0sJsu3L1NIdVTCmZ+25zQze9gbBy3cUwo1gJRdSFZ02+lihZqR6dPFsep8w8haf2PBVvpWkul0hE8u6+DmS04GuC+053xQylmgPw0PaH2N29m0/3qu9OfdVS7X9QyRoHXCrMGxWloEcleLgaYcOfAJWBt7dnb9JSUu/u6+T9/V3MKsmmoctLp2eEs/3BcDf3vQfJRCnQC/dfBvUb4cI7YUHqUpb6BExPdAgj8RCBmISU7U0umrpTZ2oOiF7NIbsUsor63I2hAG1W9VuaXTAbiaTHYh+ZKCVa3NG1Sulbr+6AmwZ3A3ML5yKE4MrFV7LTDK9PW6x2mGRxpXTcd3uBN4QQ/yWEuFm/jfK4JiW3vbCDdo+fH5yzGJNJUF2oLlQNCdbSgail5CYcGZqZrotPKktJSkm7x4/NEsEr9jO/YPGAx5tTMId6V31St9jvn1BpyCetPYwl0/LYVN/Nbz/4LWEZ5oPWD3ho+0OQXawC+a/9cmgBX28nXpPoH1OyOsDXTbfXB9JMhVT++33BMiC5KL29R/3v/USpSLnUomuVrFlq5pgwo05co5TI+XPPxx108/y+5/s2aqvw/75hPxf+4S3e3K+NK1H0ogtn0xelLl8Xv//g96ytXMunTerzrS+a3vc/BL3Uu1VWZJylVDADZp8Er/8v+LqZlT+LUCREg7t/jOLO13eT77RyyxkqbrLl4ChbS+4WKFuk7ieKUiQCD3wK9r4O5/8RFp834KH0JRBFDuW+g/gFtKFwhE/9aT3f+efm4Y1VTwfPLutnKbVZVH+vuYVz1Vis1pElOkTXu+nuO+17MoQFtPqkUveInFFzBuXhCP9r8/NKbgGhhveGP75xIB1ROgg8qe2bG3MziGFHs4u73tzLZYfPYFl1AQDVheqiGOvC84fCNPX4mFbgxB+KJI/TDECPloaab8uHSCSa6q3/UHt8IYJhyZr5PoQphDlYM+Dx5hTOQSLjYg+hcIR739rLex+pH/WCeQtZOq2ALW1beXrP01y95GrWVq7lV+/9iqYVn4L5Z8CLP4Tfr4Vt/1ZpqLteUs33NvwJ1v0PPP1NePQa+NtF8KcTIeDGj8CuFbfsy75TotTj9yEwk+trIoiVPd4c7GZ7UvHcsKeDWaXZlOU64rZPK3BiErGWkmbNxAiblJImT9OAorS6fDUzcmfwaF3MmqWgB2nL5q9vKPfYXRtbotvjiC6cTb8F2e8//D3uoJtvHP4NKrMqMUtJvUNPqFDuu/6WklfNtk/6rprZv/nblBl4Bzp6eWZzE586YgaH16jvz5aDGW79EEskrKyPMj1xIEGUWj6G3S+rBJVlFw96uKilpCU6QHyl8A17O2j3BHh7d7ua9O1+Be6/PH3xcOuipFlKeqJDOECrWS030Mt0dZmtaSc61DW7+OBAV/zGxMK+es8yV/qiVNep3HO6KFnNVr7V3kWHDPLlkjxObn2eX7zzC+pd9WkfczyxDLaDlPIHAEKIPPVQDiOqfWgjpeT7/9pCtt3C10/rq8Q9XbOUYpMdGjq9SAknLSzjnrf2sb3ZRU1Jdr9jpqKrczcWBNl/Phm6Gyj48ka1XbOg2rT4SWV5Cx82QkNj2YDH07vQ7uraRal1NvdvOMDfN+yjucfP18t80APkV7Os2sr9+58m25LL1UuvptvXzQVPXMBPNt/ObZf+DbHrRSU8D3wq+YnseSoBwVmofuhLL8Ev9+M0q/coWvvOYgNfN26/D4vJiug+QJe1lBZ3EGeRs58ohSOSd/Z0cNby/okENouJynwn+9sTXCRBbzQxodvfjTfkTem+A9U47/y553Pbe7ext3uvaq4Y8OAK29jV6mFNTRFv7a0HB0kspaGtUWr2NPPQ9oe4eN7FajY+71Qqt++jXluwGeu+K7QXkqNfzIIeZUVVrVDND9/6HbUr1Wexu3s3J9CXMPDXN/ZiEoLPHVlDYbaNaQVONo+mpdTbrsr1FMxUF9/ETDndKphxZFqHixUl3TUZW5T12c0q08/lD7G1sYclO1+A7U/Bh/fDqisHP4GnVblCnYXgLIq3lOwmkDGWksWcVqLDy9tb+NLf3sVhNfPud07BJGBj80ZWh82qvbY9QZSG4L6r66wjx5rT9x0OBznJ4+YT0y/m9c6P+WfTW9y95V4qrKv4zIphZIGOMelUdFgthPgI2AR8JIT4UAixarDXTSWe3tzEGzvb+dqp81T2l0ZJjh2bxcSBGEtJn7WfsECJRVrJDu274NVfwB+Ooev9u8kPhRBBH/h7cLjUAkr9h6rHk9qCddgoYsOucP905him50zHIqz84Y03OOrWl/i/F+pYUJHHn69YzZdW2gEBuVWYsnZjydnO2uKLyLPlMT1vOtevuJ6XD7zMC/tfgDknw5fewnPeXYQvuQ+uehqufxu+tgP+qw2+dQD+YxN88RW1Mv/CP+GV4ailFK19Z7aBDOMNeLGZrNB9ALejklaXH4fZ0S/RYWtjDy5/iLUJSQ46M4qy+ltKMdZMqnTwRM6dfS5mYeYfO7WEh4CHhl4TJTl2/nTFamxOXRySJDoMIfPuHzv/QViG+dziz6kNR3yR6oqVUXedXul8d+d+rLKEd/ZqF/hAb18G4InfgZCP3PV/pNRZGmcFd3uDPPjOfs5aVklFvrIsl0zLY0vDKFpK+sLZnHLN8kiwlHSrILGzcQpSW0rtRCKSZ7c0UznnEZzVd/Fi3Y6+87/+f+llonlaVENLk0m577wdysUY8tMqJCZh6muwaTIPain98/0Grrl7I06rma7eIB8c6GRj80aufvZqPmjdpHbSJ0wWm7LQhmAp7ejcwbzCeX3r1bTvoNWewwm1p3FbczMlO69mSXHyLNyJRjruu78A10spa6SUNah08L+O6qgmEb2BEP/95McsqszjU0fMjHvOZBJUFzjjLCVdoBZV5jGtwEldc4okgY496kf0x0/Abw6Dl34Etiy6q1aQX1gLF6rMLNwtFDmKogkQeubdfs82ZucuornHz/YkwufyBbn7zb2ccdub+L3F1Hv2cNXRNaz72vHcffUaTl5UjqmnAXLKkGYrD++6AxnKw+nt61ry2UWfZWHRQm7dcCsf7G/nmr9vYvEDNhbcZ+KUR4Nc96yHn7/RwT8+bObDA124fPF1uPxhf1/2nZ7ooDVOE8EeyiId0PAevXmz8AbD2MyOfjGl9bHrk5KgREmbFCTpPpuqqGkipVmlHFt9LE/sfIJgJIi/t4d6j4lPHzGD/CwrnzlqNj5ppbOrK/6FQ1g4G4qEeLTuUY6qOqrPLQdU51ZHXS/dYStN7V28e3AHDa1Z/OdDHxIKR9SFSF8rVTIXVn4aNt7JrOyqaAbee/s7ueD3b9AbDPOFY2dFj7+4Kp/dbZ5+n0/GiGazlUdjNP6wv8+d1NOImvykdqHG0uHrwGKykGPNiYpSt+a+29TQTVOPl4B9C5bcbdy1/yu86tqtUuk798KWxwY/gadNxZNAiajUkihCftqEpNhR3Oc2N5sGtJT++sYe/uPBDzi8pognbzwWs0nw8rbW6ALcHT1adqQtxluSW5l2TElKSV1nXdRyA/q+31YnVK4AYHmkiXnl6buQx5NB3XeAS0r5mv5ASvm6EMJw4Wn8Yd0uDnb7uO3ylZhN/VfWTyt0xsWUDnT0YreYKM2xM688h7pEwWj+GB6/AQ5qwclpq+G0n8CicyG/mu5nP09BJAQ52o/G3UyBvYAOv5o9tnkCCLOLFu9Bzlt8MRs2wCvbW1lQob6Qeuzrsfcb6A2EWV6dz/LyBbQFt/PtTy6KH0t3PeRXs7F5Ix+0fkCV/DQfH+wTBYvJwvm1V/CTd7/FhXffS3Z4IV86XnMHtrjZ0eLiha3NhGKSOcrz7MwqyWFOWQ69QS9OS7z7LmhWX8mlYju9QRMc/nl2F10Ju/ZiEbZ+orRhTwczirKozHeSjBnFWbS5/Xj8IbKTlALSLaWBYko6F869kHUH1vFa/Wssd/XgE9P59FrVaPFzR87E+7qDrfsbOUp/wRAXzr5W/xrNvc18a8234rZX51TT6e/kxe376NrRzSpTL8IKx9Scwgtv9vKvTQc5P+DpEyWA426BDx+ktqeFf0kXP/n3x/z59T1U5Dm45+o1LJnWVxx2yTT13dja6Eop7iNCX6OUq4tSGz9e/2Ne2PcCr1/+OibXQWUdDJAlqtPtDdLuVQtnhRB9iQ4WC/S28+yWJiwWH4GIlxmWE9nv2coN9ka+ULOYm7o9KiFnyYXKCkqgxeXj35sauaCtgazcYqzQV9qpt0MTpSxKnBU4LU4cZgddYZJaSlJK/u/5On790k5OW1zObZetxGE1s2pGIS9vb2FGjfrd7/No4hPT66xJFuHeVcczL+3g00fMpDDG+5JIo6cRd9Adv+xD/35bs6BkHgFh56isA2TZ0rncjz/pjHKDEOKPwP2ABC4F1gkhDgOQUk6u1I4MsrfNwx9f2c35K6dFA8aJVBdmseVg32r2/e29VBc6MZkE8ypyeWNnO8FwBKtZ+5Gs/z20bodTfqhiA4Xx1leXv4tpOdP6ZnKelrhSQ+1uPyanCoIfM2MV/65w8UpdK5cePp3/fb6Ov63fh9Vs4uzlVXx27UyWTy/gT5u28uv31+EOuPtiFKDiIaULoiVMjqg4jgfXdxEIRbBZ1Hj/vaEQKRysWLibu866gVxH/IUlGI6wv6OXXS1udrV62NXqZlerm0ffO4B5tp8D7Wp2bhImLMJCwJGPFCaayMdSVANn/pyCHa3AXkzEJzpEIpINezs4ZWH/Ra86egbegc5eFkRFKd5SsplsSWsDJnLMtGModZby0PZHWOJzU1JUGE2uKM6x02XLoqmtg+YeH+V5jiEvnH247mHKnGV8Ynp8acnqXCVq//fyeq62ZiGzASSfXLiUA7ty+d3Luzgv4kXYYkQpfxqsuYZZW+7GU1zIn976kMsOX8L/O3NB/GfUdYAlpWqWvuVg9yiJkvb917LZDnTU8cSuJwjLMK29rZS7mvpZSd5AmLpmF9ubXGxrcrG9uYftTS7a3AEWr6ynKEuNM8uShVmY6bFlI3s7eKauiRW1sB04supItjx3AhfN+AZ/zmrhyiO/Tv4TX1HxpYVnxZ3vrV3t3PjA+7S6/JxkO8iLLbn8+ucvc9sqwXJQohT20yaD0eaaBY4COr2+fpZSWFtAf9/b+7l09XR+fP4SLNrv+/gFpfzsme0c7FG/131ezYqMsZS29eawLNzOL56r47cv7+TiVdO5+phaapPEnhOTHIB4S8lsYaephuXmCVIDMQ3SEaUV2t/vJWw/CiVSJ2ZyQJOJHz35MVaz4FtnLEi5T3Whkw5PQM3U7Rb2d/RGL5TzynIJhCPsa/cwp0zNlLwNm+nIXkjJEV/GbjH3O16Xv4vFxYvB6gB7vnLf5RZFM6za3H6y8+oxCwsLixZy3Lw9/OWNPZzwi3V0e4N8Zu1M/uPk+NiXnkm0q3tXX/UHKVU8ZM4p0Qv34dUzuDfUQV2ziyXT8tnV6ub1ui5WrTqKhsDbWC1hIF6UrGYTs0tzmF2aE7d9f2cXn3wCntrUzjS2cfMp87GarQQcefzr3I/Y+dZ1rHSqWXBprnLxCWnDG+4TlLoWF129QY6YVZzy/Y9NC1+Qp5cCireUKnOS149LxGKycO6cc7nzo7/gNgeYMy3+Qpqdk4fD5+Nv6/fxn6fOj1s42+72c9Htb2ExCaoKnEwrdDKtwElVgYOqfCdmWyevN7zOtcuujVqNOroobWvbx4LpZTS0BwEnM/Kmc/0JFdz0wAeEctxYrX0XLV8wzO+9n2Sl/x4Avn1uEV84fGn8P+TthN+vpeyoGynNXTV6i2jdLSrRxZYFWcXcafERlkrMGz2NlPUcpM1Uyt+er9NEqId9Hb3RogYOq4l55bmcML+MHS1udnW3cESBSh4RQqhSQzYX7s5m9rR5OG5FhO0NcHTtHO6XeznV1cOzWaU0Tl9FfmGNspYWfBKEIBKR/OGVXfzyue3UlGTzpytWM+1uNx0VtXTUB3hxX1gTpXYI+WiVQRZmldLm9lNgL6DbezBunZI/FObmhz7k35saue642Xzz9Plx360TF5Txs2e2s7VZWY/79AQWTZQiEckH7nyOFy6e/9Iy/vROBw++c4C/vb2PkxeW84VjallTWxQ9pi5K+m8Y6Pt+W7PVwvLATC6Rr6u4WBILcaKRTj+lEwa4TVlBemlbMy9ua+Gmk+dSludIud907aLY0OVFSsmBWFEqV0IUjStFItC6jefairnkj+s52NU//VnvOgsoF57mvtNbore6PIjcd1ldsRqHxcGpi8sJhiN8oriH185o44eO+yna+re49Fi9Bl5cF1pvp0oIyJ/GQc9BKnMqWT69AICPtKD4vW/tw2oWXHvYRXiCHl6tfzXt9y8/S/2oVlaX8ruXd3HjA+9jM9kIhAMc6PSBCJNrV+9raY4SJRmxxrnv3t6tZptHDDC7n1ms3uv97b0xMaV4UUrHdadz4rSzkER4IddMaXH8ea3OHCqcEV7fqeIFsQtnX9/Zxp42DxX5Djo8AZ7Z3MTPn93OVx/8kEvvWM+nHvw1EQn3Pl/BRX94M666fHWOEqW8vB7mVZdxQKg6hNNzp3PWsipqipyYQr1IzRJ8d18HZ972Gr9e30lH4dkAZIU/6P/PfPSIWifTU8+SqrzRSwt3Nal4EnDQauPxLDtrylcD8OimTXQ37+e5AyZ+89IO6ppdLKzM46aT5nL7Zw7j5a8dz5YfnM4TXz6Gn1+8nO+evYiIcNPttkcPn2fPo8dipbujGSGgqkR9R5aW13BYoY+qkHq/Dnqb4ZivKtf47pfp9AS4+u53+Pmz2/nksiqe+PIxrCizYA55WbFgDsunF7CtR5sg9LYTDgXoiPixyDyO+MmLtHRZ6CQS/S15/CG+cPdG/r2pkW+fuZBbzljQb7IzvzyXynwHu9uVGNUHewhCNCV8d5ubd/0qnjg3soefXbScN245ka+cMIeNezu49I71nPu7N3j8gwaC4Qg7OncwLWdavIcj6r5zsq/dw6ZIDfZIr+rcOwmYHE7GCYYvGOYH//qY2aXZXHlU7YD7VsekhZfl2nH5Q1GhmlOWgxBq/cKZSyvZuXMrc6QXa+Uidja7OPs3r/Oby1dy1Jy+YqT+sD9GlMrB3UKh4xi8IS/ekJc9vteI2Lq5aslVAKzKd7Or5OuYWw/CS6haXpEgvPJzOOY/4LArmJYzDafFGV/ZIdrCu5rGvW9SkV3BjKIs8hwWNtV3c/byEI++W88nl1Zycu0ySt4t4d+7/82pNaem9x5q4nLRqlqOKp3H/71QR9VSM8FIkP3tvVjMYRxa0kNhlg2zSRAOWwjEiNKGPR1U5Tui73Ey8p1Wakuy+demg3xhcZVKv41x3zV6Gjmq6qiUr0/kr6+4kL0zeCnbz7WxMRwAazbC6eKjhja8gTDOmIWzb7+xj1yHhbuuWhONPXoDYQ52eznQ4eKWd26lyHIYC2vm8u6+Dq69513+9LnVHDevlK0NIWTYwaIZQSyOQuqtZhxmO6XOUoQQfPkTMzE/E6GuM8yDT37MX97YQ1W+k/u+cARHVR/LL+4/it3bH4e13+wbq5Tw7t3ah9HNkmn5vLqjDV8wjMPa30IfEe6WqCj9xaO+Y5eWf4YNzRt55P33+RE9rFm+mI/PO33Qcx82oxCz1cPuJqJub1WUtZ2Qu43DZhTiDm/HarJS5CjimPIQFftVxl2jpxGWXw7r/gfX8//DJzslbe4APzpvCZ85YoYSkA5tjVJOGbUl2Tx/QLtE9rbTKQNEgF5vFuGIpLnTjDk7gAwF6PQEuOqud9jc0M3PL1rGxaunJx2/EILj55fxr8ZORA6EkDQ4sqnRYqnv7O1kS6RG7dz4IdR+gtJcOzefOp/rT5jDo+/Vc+fre7jpgQ/4n6e3YZm5mSVlc+NPEuO+29HsZnOkFiksiI5dUDKHic7Et+UmIHe+vod97b18/5zF0dhKKvpEyRtNTdZFyWkzM6MoK5rs8PobytI499STefzLx1CYbeMzd77NH1/ZhZSyXzUH3VLSSw11+DpoMT1DDjM5slJb87HpIczug/DJX8J1r8O3m+CKJ6CoFp7+Bty2HNP6PzArryZelGI6zja5+1okLKsu4KOGLh57rx6XP8QVR9VgNpk5veZ0Xmt4LXnvoSToqd0Os4MbTpjN/PJc3D7wBv3s7+jFYolE3Vgmk6Akx0YwZImKmZSSt/e0c8Ss4gFdb0IIrv3ELDbVd/N2g/Zj1VLCg5Egrb2taVtK67a38I/3GphfWEWH2dy34FHHlsXXC7qwTruTjXtbVbaXPR8ceby9u53Da4rikmGcNjOzS3MI2D/CE+7ilmOu5peXLOefNxzNnLIcrr1nI2/uauPXL+7AFC4hO6sbrFkcsFiozq6K/t/nLC4A4IEP2rnz9T18+ogZPPvVT3D0nBKEI49ZWeXs9bbC7nV9Y238AJpVA0i8XSyuyicckWxrGloOk5SSdrefzQ3dPP9xM/e8tZc3dUtRx90MueU0e5r5R9cWzne5Wb+xDcLZnDhfCcbc2fPSEkNvyIsUfly9Dv69SSWp5Nny6ACyQj2cvriCJrdaDG0SJlYVeSmORLAKK43uRqTZxvrKT5HbtJ5lcjuPfOlIPrt2Zt93KFrNoZTakmwafVakyQLuZlq1uJDHk4VJwJoZ0+kRIVweD5f88S22NvZw+2dWpRQknRPmlxKmF7NQQrTP0ed23bi3E5Fdgsyr7lfh22E18+kjZvLCV4/jzs+tZnqxjY5AA69+ZOGH//q4byG+vuTBmkVds5s6puP7+n6Yd9qg7+9EYFxFSQhxuhBiuxBipxDiliTPCyHEr7XnN+nJFem8drTYVN/Fb1/ayemLKzh2bumg+5fm2LFbTNR3eqPVwasL7TR5mgiGg8wrz6Wu2U2nJ0DbbvUlzJ2+lDllOfzzhqM5fUkFP316Gzf8/T0aXcrk7xMl3VJSovRo3aOELS0syjq370dW9wxUHQaHfwEqloLZArOOg6uegiv/rcrzP/v/mN3wEbtaNvUVV9VEKZhbTqu3NZoyvbQ6n22NLv76xl6WVeezUnPpnTXrLIKRIC/seyGt91EXF4fFgcVs4ofnLiYcNrG1qYP9Hb2YTZG42Epprp1A0BKNKe1q9dDmDgzoutO54LBplOfZ+eOb2oJEbSbZ2tuKRA6aDg7g9of49mObmV2azWGVJWpdjC0+8ByxOGkREkv2Hn676VfQvBnKF9Pm9qtFtinG+nDdw1RmV3J0lWrRUJBl497Pr2FmcRZX/vUd3tzVzuzCGRz0NIDVyQGrheqsvkXRtoj6fwoL8vn7F47gv89bSo69zwlSW7WG3Xa7qrqhB2reuxcsDph+BPi6ohl4yVx4/lCYt3a1c+/6ffzsmW189cEPuOyOtzju5y8z/7+eYdV/v8BZv3mda+7ZyHcf38IX7tkYXZqg3rxmyCnnr1v+ikTy+e4e9h7YT761jDCaCOQN/hlAX+WS0qxibtcma/6AncZAkCLh4twVlXG1DOdluRFAjrmYA66DXH/fe1y1aREuUz6/qX4pWn0lSoIogSBkKwBXI21aNYf2Hgczi7M5snYGfnOESCRIc7ePe69ewymLUifd6Bw9pwST2Ue+WSUx7bX3uf/f3dfBqpmFiMrlKdtOmEyCkxaW870LixFCsqx8Afe8tZfjfv4yT3x4sM9SsmVR1+JiWmEOzqz0F+iPN2mJktYS/VNCiCv020hPrFUg/x1wBrAIuFwIkZCTzBnAXO12LfCHIbw24zyzuYlL/vgWRdk2vnt2eqcTQmhp4b1RS2mP9x1OeeQUdnTtYF55DnvbPNzz1j5ms59gTpVqqgfk2C387lOH8f/OXMAzm5u48aHXAWLcd6Xg76HIor5w93x8L5FAMUsLj1XPu5pVgcv5KYpb1hwDn3sCrn6WuVkVtIZ76b5tqQoEt24Dk5VmInEX7mXT8glFJLvbPFxxZE1U/BYVL2Jm3kye2vNUWu9LrKUEcMSsYnIdDna2ddPY7cVkCscVki3NseMPmPEGvYQiId7e0x593WDYLWauOXYWr+3VZpCaz72lV8VtymIu8Kn4xbPbOdjt5WcXLaPQ7MBjMhGyxscSXTY7UoA5ks/Hnn/zr546qFjKOylq8wHs79nP+sb1XDj3QsymPkuhOMfO375wBNUFTspy7Rw5Yx4N7gbCFgcNFgvVzpgxaxehr5y2IurqjWVW4VxaTQLXwfdg25MqEP7Rwyq7s2AmeLuYVuCkIMvK5oYeIhHJ5oZubn9lF5+9822W/+A5Lv/Tev7rn5u549XdbNjTQSgsWVZdwJVH1fC9sxdx+6eX8/zFWbyx9h1OC7/CHa9p8Qu/GwJumhxZPLz9Yc6edhzTQmGyQt3MzJ+m4jygmtulgZ5lesaiOWxrcvFfj29m/Q4vfnMEC2HKbMG4OGFesJ1eHPh8Bby6u47nPm7m5jNXknPcV7Dufh4aN8WfILqmqoxu+THC0kWvpQB6+kTpYLuF2aU5UQ9F0Cp58sZj0vouAmTbLTgdATyefAows1cr8trq8rO3vZfVNYWq7XvbjgEr8OtJDj/+5Km89s0TqCnO5r71+2Lcd1nsaHYxrzwn5TEmIoPGlIQQ9wKzgQ/o6zgrgXtGeO41wE4p5W7tPA8A5wKxjWvOBe6RqiTBeiFEgRCiEqhJ47UZQ0rJn1/bw0+e3sry6gL+dMXqaEZYOlQXZnGgw0u+00ZRto3aAvUDbPQ0Mq98AaGI5PZXdvG0owlrRXwBVeV+ms2Safnc8JjqrfjR/iCHVxD10xeE1cfiD/sItJ9B6TItxrLjWUDC/NMHHuCMtcw+8Yfw4vXsrFrMqhd/qLYXzKRRu2hU5miipFlGRdk2zlrWN7sVQnBm7Znc/uHtNHuak/YmisWvBYf1KuEA0/JzcbvDRCQgwv0spY/2TydkD/FK/Sts2FNEWa6dmuKsxEMn5fI1M/jdyzsJRSxYtB+tXkcwWTvxWN7d18Hdb+3lirUzWTWziK0Nalxuk5mCmP16NBFdkXMpm9qe4gcF+5mTX8bbezpwWs0sjVkbpPNI3SOYhZkL5l7Q77myXAdP3ngMHn+Ylw92EIwE2ervwGsyMd0eI3DR+mnJ34vaPBX33F06m+Uv/Tf4usHfA4d9Fj5+HHzdCCFYUpXPvzcd5JnNjXT2qlT9eeU5XL5mBsfMKWHJtHxKcux9LkhXM+x6UfVyev2laDXtn9izOfzNY7jm2FmU+NX353c92wD40tJr4LW7KTW7mVZWwz/q1iMBkaalFBWlhXP411tt/G39fmbNK6TdFCIChN3NcZY9rkbctjK63TnYcw/y4LVrWV1TBN5r4M1fqwnYJXf3ncCjuR6zSrj1vQtxls+hO5BLfk9DtO5dfZuF0+bnUOAoAKBbhJlTlN73UMds8dPVZWGe1cw+bS7yrtYaZtXMIvAtB6SytmesTXqMus467GY7M3JnYDaZOWNpBbe/shvvPBdOIGiys6fNw0kDLJmYiKRjKa0GjpZSXi+l/Ip2uzED554GxLZZr9e2pbNPOq8FQAhxrRBioxBiY2tr67AGeuvT2/jxU1s5Y0kFD1y7dkiCBCquVN/Zy4GOXqYXZVGVrYmSuzGagRcIBpgePgBlC5Me46jZJVx/svqh/eiJffz06a2Es5T7sDCoUlLzbUUEuw+jOFsb3/anIX86lC8ZdIz6ivCdqz8L165Ti3WXXdqvDE9VvoMl0/K45thZ/WIAZ81Saz8e3P7goOfTLSW9ogNAltXOjGKV3CDpL0pdbbMozyrnoe0P8fbujrjU2MHItlu46uhaeqWNjs4uANx+JUq55tQXFH8ozDcf/YiqfCdfP12l/ueh/u+ehFP3aBetpZXTWHlwGQWRCF9reIb1e9pZNbOwby2aRiAc4LGdj3HC9BMozUruCs6yWSjNtUfTwt/yqPUm07VZOhC/WDIJswq0wqyLP6ks4Ge+pRrYzTxaWeW+bohEOH1JBYXZNk5aWM6vLl3Bhv93Es999Ti+d/ZiTlpYTnm2GfP+N+CF78Ptx8Av58E/v6Ratc87Q7WcOOv/cEY8LAjXcceru8Hdwg6rlSc6P+LyBZdTWazew6VFYWbkV+OTITpsWaBd4AdDF6Wy7GJ+fN5SvnT8bD59+AIiSDxC0Nq1m4iMUJWjWV6uJrJLqlleUYM097B0uubGchYol/bHjyuLRMfTAo4CImYL7qAbS04dzeFscDXRZjaTbbITDFmYU9ZnKXWaTUNuXxGSvZhkFuW+CHtNalK5cW8nNotJuVIrtaUZA3SO3dG5gzkFc6IW9kkLywlHJPua1HVub3eEYFgyt2xyWUrpiNJmIP2c2fRJdjVJLNKWap90Xqs2SnmHlHK1lHJ1aengMaBkHDu3lOuPn81vLz9sWJlJ0wuz6OwNsq3JxfRCJwX2AhxmBwc9B5lVmo3ZJFiT3405Eugr758Mk7r4XHrYPP74ym6+95Ka1eX5XMzKn8Xp0z4L0kpprk2Z8LteVhW807hwl2eVk2PNUckOVSvhknvgxG/T6G6MPg/KInryK8dGKzfEMiNvBqfWnMrft/190ISH2JiSjs1soyTPzKNfOgpJKF6UcuyEIibOrDmPNw++SbO3IW13ic7njqzBh52tB9Ts3bXnZQBy96ROZf/dSzvZ2eLmx+cvicZp8rSfTY+Ib4HRo10cDqsqZ4Vs4kudLvZ7m9nRsSOp6+75fc/T5e/i4vmDV8aenqOC5291q2SU6ZaYQv36uhRb8rjBtJxpWE1W9mQXqM/W3wOHXaG+F44CQIK/h8+snckrXz+BX1y8nPNWTotf6vDyT+B/auGuT8Kbv1EJHCd9D774Gvzndjj/D7D0IuUSRPCFqn3c89ZeulsPcFtRAdkWJ9csu4a9nQG6ZRYL8wLRydnB/PK0vqPQJ0rFjmJOXlTON09fQJEmaD1mE43t24GYCh2uRrKLq/n0qhWAKngbZe31Kq72+v/1bXO3QHZpX0KNqZe3hVALZy1m8oV6j+eU5URju10m05DaVwTCAQIRPyumVVDmDdBKGE/Qwzv7OllRXaDWJ+ZWqMXGA4iS3thPZ3l1AcXZNhpaO8DiZEer+l7oE9/JQjqiVAJ8LIR4VgjxhH7LwLnrgdg0lWpUm4x09knntRnjmLklfOP0BZiSlBFKBz0Dr83tZ0ZRFkIIKnMqafI0YbeYue64WXzzMO0Cl8JSArVGyWlxcusFq/jG6fN5dp/SYeFp4fHzHmdRjnLTFWfbVbn+kFeJUhoIIZhdMLtfw79GTyNFjqI48RiIa5ddiyfo4d6P7x1wP79WmkWPKYEqNRQMB1k1s5BgJIjF1OddLtUqJ6wtPROBCWvBBtYOsfpAfpYVsyObts5OerwBXHvWAZCz762k+29t7OH363ZxwcppHD+/L4aTq82JeogXpW7t11Sdlc1qRwM1ITU+U9bOpKL0cN3DVOdUs7YyuXsmlooclU32Qc9uTFIyzRyTBh/NtkqeGm8xWZiZN1PVwDv9VqhYBis+rZ50Fqi/vkGyJjf+RdXUu/Q++MYeuOrfcOzNULksfkFmVhFUreR462aCYcl9H77KK1lOPr/gM+Tb83l5ewvtMo8ZDm/UmjmYVTDo/6/T6evEYXZEy1MBMUVZzTS2KQ9+ZXalSurQqkXolv5BT8xlIqcUVn0ONj0IXfvVNk8b5JTRG+pby/aOVQlUm9mEQyprdHZpdlSUOs3mIVlKroDKcDxp/kxmhZSbtK59D1saullVo1nAQihrKYUotXnb6PB1xImS2aTSzVs7OpFWJ3XNLoSg38L1iU46ovR94DzgJ8AvY24j5R1grhCiVutsexmQKHZPAFdoWXhrgW4pZWOar50wxK6j0RfOVmVXcdCtfiBfP20BK+xaUcrS+ckOAahqDnqSw2mLK+ggD4mIBmf1thXFOTZVSsWWCzOPSXucehfa2KriTZ6m6Iw2HeYVzuPkGSdz39b7ojGbZOjlgmLFzmq2EowECUfChGVCooPmMo0E8yg1HYa9cGPU1TcUnFk5OKSfLa8/jqu3FauU2He/0peVphGOSG55dBP5Tiv/dVa89ZqnNQnsicQXMO0R6hh5JgsLxT4avdPJMVVizdnFCi0Wp7OraxfvNr/LxfMvxiQG/xlaTVYqsysJyhAVoTDW2ItgTGA7FbPyZ6lq4TPWwnWv9dVO1JJqBmy/3duhstKWXKDK8wzWG2r2CTia3uOypXn8I7SRslCYTy/9PAAvb2/Fa8knO9wdjVM2OtKPx3T4Oih0FMa5baP17/IradKKz1ZkV6gYV9gPuZV953IntIQ46iuAgDd+rR5rFcJjy1ntylIi0mo2Yw5lUZHnINdh7YspDdFS0kWpMreQxVK97u/vbyQUkayeGeOWrVwOLVsh2L+xpZ7kEFeIFdUSxxTyERB2djS7mVGUhdOW4XVno8yAvwYhhAn4nZTylcTbSE8spQwBXwaeBbYCD0kptwghrhNCXKft9hSwG9gJ/Am4fqDXjnRMo4Xe7A/6RKkiuyIarwFUo7OiWSlnu6AsJX12VlOcjcViVZlBWsHLdncAm8VEjs2kUsHnnKRK4afJnII5dPm7aNdLn9BXhmcofHH5F3EH3dy39b6U++iWUmxMyWayEYgECEm1diUxpgTQ6vbhbjsczB5e3P/ikMYF4MzKJc8SJP/9P+K255BryUK4m6E5/uvz1zf28GF9N987Z3G/gph9ohQ/O+7RPMh53m4Kgi18GJqBu6sWa/ZuzOZ4q+qRukewmCycN+e8tMeuV3aYHgrFLQDu11I7CbX5tdS766PvexQ9luPtSn3ids16LpmXep9YZp0AMsxJVe/Q7PDwma4QLr+Z3kCI9bvbseaWQm87edZcciMRGizpr+Fv97X3q1MYtZQKqmnsbaHQXqgsKZf2+8qtoCKrAoGI/80B5FfD8svgvXtU4oanFbLLoqI0K3cR3fZemsxm2sxmwn4nc7QYjd1sJ8tk02JKQxelXFsus4O9IOFfHyuLaFWiKMkwtPS/tOn1KBNF6di5JWSb/LgjNuqaXcwtm1yuOxhElKSUEeBDIcSM0Ti5lPIpKeU8KeVsKeWPtW23Sylv1+5LKeUN2vNLpZQbB3rtRKUkx4ZdW2SrL5ytyqmiw9fRVzanZeuArjvQLCWbmtmaTYK55Tm0i4Lo2oo2d4DSHDvi4AdKqFKlgqcgWm5Ic+FJKYdchgdgQdECjp9+PPd+fC/uQPKU1qj7LiGmFAgHCIaVBRIrSiU5Shg+PNBNc/N08q0VaSVUJCJsThabD7Co9x16imeTo63xYtdL0X32tXv4xXPbOXlhGWcv6y/IeSGtsVyiKMkQtojE0fA+AB/LmXh7aokIP5vb+lpze0NeHt/1OKfMOCWtQrA6erJDdaIoDZLoAMpSisgI+3oSCnNG3XddqU/cpmbllMxNvU8s09eANQtP2ztqvL1ZXPnXDTy3pZlAKEJ+SYWyvrydVAZDNCbE5gai09eZWpRyy2gMuqnQMz+jolSJ1Wyl1Fka9U7EccxXVZWTN25T1lVOnyidplUoeTonC5/JhMfjiIoSQKElS/VUSqPRn44uSnlmJ85QgGLhBGsbc8tyKMiKmQANkOxQ11lHqbO033uR67BSmSVpC5jZ0+aZdOngkJ77rhLYIoR4McMxpSmDEILqQidmk6BSa6ym+7gbPY3KPG/fNXCSA9AdiKl7B8wvz+NgKK/PUvL4leuu7mnVOXPuKUMaZ7Qwa9cudb40urKm4rpl1+EKuLh/2/1Jn/eGvAgENlPfj1CPKQUj/UUpx27BYTXx748aARNn1VzAey3vxdfrSwdrNnmhDrzSxgFLIbmOQrWAWBMlKSXf+sdHWEwmfnTekqTZfY6gD6uU9ATjqx/0yCB5kQjiwNvqcf58Qp7ZCATrD66P7vfs3mdxBVxpJTjEEhWlYKKlNHCiA8Rk4MU0/FP/TIH66+smGA5GEwniaKsDs02taUoHix1qjqFTi+/MLa9me5OLrz/yIdk2MyVllarAac9BqkIhGiL9azymQnffxaL/JnqyCmm0mKjUk0D05n5aunlFTkW0f1YcxbNh8fmw4Q71OMZ9d3jVMqzBXP6Roy7uwUA2s2NEqcCaPWRLqSeoZX0K5Vab6yzGkdXOcfMSErEKZqjPJ4ko6Y39klHhlHSHrIQikrmHqCj9ADgL+CGZjSlNKWYUZVFd6IyWsI+KkrsR2ncoM30QSynWfQewoCKXhlAuYVef+6442wYHNqjqDVlDSwQodhRTYC+IugbS7cqajMUlizl22rHc/fHdeGI6ver4Q34cFkfcRd9mthGMxIhSTExJCEFprp1Wl598p5UvrLgEq8nKw3UPD21gmnv0nxxHo8ejClnOPhH2vQlBLw9tPMCbu9q55YwFKXs0iWAveRFJjz8+ZtYTDijX3oG3IbeS+bNmYSabBUULWd/YJ0oP1z1MbX4tq7WipOkSdd8FgwmWkgfMdjCljh3U5NUgENGGf1G0mFKwt50vvvBFLn7iYiIywXJp2wHFcwY8fj9mnUCXrwOzlMwsr+FnFy0jGJYcM7cEc3YJhHzQvpOqUJjGQM+A3ZF1PEEPHd4Oih3xWZdOixOLsNBjy6bJYqFS6uuoNEspR1n6VdlV8YkOsRxzs7KWALLL6NWsz2xrNsWR+ey1qe9iJJTH7NI+8S+w5tI1SKO/RKLuO6muBbVZZTicHdyS2G0gRbJDKBJiZ9fOlKJUbA/jk2qyd8i57wCSxZMyEVOaanzzjAX8/KK+dsR65lGjp1G57mBAS0lKGV8hHJhfkUurLFAxESlpc/uVKDV9pLKihoiegadbSiMRJVCxpW5/d1I3my/si8u8A2UZBSKBpJYS9FULP7ymiJKsYk6tOZV/7fpX9AKSFrYsQLB1xmfo9PWQa82F2SdB2E/X1lf4739vZU1tEZ9aM4DHOuAhD1O/RI6eiE+JkrsZKpby1VPmcefnVnNk1Vo2tW6iN9jL9o7tbGrdxMXzLk57jZXO6orVHD3taFYHZXzb9aB3wFgkKDdpVU4Ve7oSRMmeixRmftK0jnea3qHF28LWjq3x+7TVpe+605l9Ah0mMwXhCCK3ggsOq+aBa9fyw3OX9DXOa95MVSiEJ+wbMCkGlLv3ppduIizDHDf9uLjn9GZ/DREfbpOJyoBmtbiawFmo2rygvsdNnqb+ogtQsUSttQLILo1aSk6LkxnONdHdQqH8ePedLZfONFqix6KLUo4mxDOzyvGEPHQFkliplctVvDPcl1Szr2cfwUiwXzxJxx7xYbJnYxLEjXWyMKgoCSFcQoge7eYTQoSFEKPUeOXQZUFFXlxacFlWGSZhUjO3lo9V9e7i/mt/dFxBF2EZjhOlBRW5tMl8TGE/0tdNuztAja0bvB0q7XcYxGbgRUVpiIkOOstLl3NU1VHcveXufsLhC/mwW+IXIQ8UU4K+ZAe93t0l8y7BFXTxzN5n0h/Uqqvh7F+xYuVqwvQSCNph5lGqUOfzD+EPRbj1gqXJ0/8DvbDpYdj/FrnJRCnkiyZBULGUaQVOjp9fxtrKtYRkiI3NG3m47mHsZjvnzD4n/TFrlDhLuP3k2yk22/u77wZw3enU5tf2d98Jwd+LSnjEs5sL514IwFsHY1LkQwHo2JN+koNO6QK67FkURsLRyiNrZxWr5oe6KDUpUQKSx3o0wpEwt7x6C283vc2Pjv4Rq8pX9dsnz5bH9i5l4Vd4VGUJlQ7e992tzKkkGAnS7m3v93oATvovmHU8lC2ME6VF5cfg0D5Xu7k4OjkCKLDlaZZS+qLkDrixCAvOsPrfazQLeG/33v47Vy5X6eat26Kbkjb2iyXYS1VpMWctq8p8xfcxIB1LKVdKmafdHMCFwG9Hf2iHNhaThbKsMuW+a9mqfvQDtIPWF6PGuu9Kc+302tQP3NNxkEA4wlypzYQrliYeIi3mFMzBHXTT3NtMo7sRu9keXbk+HK5bfh0dvo5+bjZ/2J/UUgpGggS0BIKUojRLidLKspXMKZjDQ9sfSn9A1atg1ZWcuKAMYfbR2g1uaWOLZRE13Rv47lmLmBW7rkNK2L8enrgRfjkf/vEFQJJXUNPffRfykB/uEyWdlWUrsZlsvHzgZZ7c/SSn1ZwWN7kYMtasBEvJM2CSg86s/Fns7dlLOBKObnuj4Q1+lmvnBHMB3z3yu8wrnBcX/6Jzj3ItD1WUhKAzq4CCcCQqSlFiLSWrci+lEiUpJT9a/yNe2P8C3zj8G5w9++yk++XZ8tjbsxeAyi6t7UrPwbiOttFKKokZeDrli+GKx8GRFydKc8uLWeENY4tIphdVxFm4hfZ8PCYTgWAacTFfD7ia6An0kGvLRWhJQDV5NQD9k1AAKleovzEuvLrOOizCQm1+irY5QS+zKkv49eUrBx/TBGTIVcKllP9kCnebzSRV2VWa++5jKEvdvRaSi5IQAmehmgm62tQPbbpfC/yXx9fQSxc92WFn185oteWhupliWVm2kiMqjuCvm/8a16DPF/L1W5BrMys/uH5BSBSlldMLmVeew6JKlW0lhODieRezpX0LW9qGtiIgx2FCmALsbYlw+R3r+bdnIQtMB/jMIi3xousAvPpz+M0q+MtpqoDpgrPgc0/CjR+SV7Kgv6UU9PRZSuV9ouSwOFhZvpJH6x7FE/Rw8byhJTj0w+pMYimlJ0r+sD8aV+n2d/ONV7/BHGnmVlmCSZg4svJI3mt5r2+dzlAz72LostoojERU2nUsuih1H6BKKyybKtbzm/d/w6M7HuWapdfw2UWfTXmuXHtu1C1X2bFfvScJlpKeRZoyrhSD/v9nWbKYVZLNxR0mvtvewYyy+DhtgTZh69KabKYk5FfVMO49H1fARa4tN5rKX5E3HZvJFhXVOIpmqRYpCaJUk18T/b30I+hNa5IyUUnHfXdBzO0iIcStpCjpYzA0KnMqafQchBWf0cqzpEbvpZQ4wy4sUz94T4eaHZZ66tQX2T68AGdsBl6Tp2nI6eDJ+OLyL9Lua+fRHY9Gt/nCvrg1SkA0E09PjLAmWI4Xrqrmua8eF00WATh79tk4LU4eqhuCtRRzjk63mR0tLk745KXqiZd/DPecC79aCi/9t7qonft7+FqdKqVTeyyYVGM5PTYAysXkCrqVKFmzVb+qGNZWrkUimVs4t6/l/HCxZvVPCbcO7r7TM/D0ZIe/bP4LroCLn5gqyfIpgT2y6kiCkSDvNb+nXtSqyvZQPAxRkmEKZ5+qShvFEpOAU5CrGkwms5Tu/fhe/vTRn7ho3kV8ZeVXBjyXnhZuEWaKw2HlfXA3x4mSHsdtcifJwEvAG/JiERasZtUkMj+Qy7luDzXl8V6DQi0lu3OwPmLPfxeaNkHLVlz+rjhRMtvzqMiuoLm3uf/rTCblio8RpYEy7wDNch44xjiRScdSOjvmdhrgQlXkNhghldmVNHtaCH/ia7Bo4BhDKlGqmFYDQGujKpOS171t2K47gAJHASXOEnZ07ojrSzMSDq84nFXlq/jLR3+Jrk9KZinpIqTHnxItpWTk2nI5s/ZMnt7z9KDB8lh0QVlaWc59XziCNUccp9xM79+r2kYffwvc9KEqp7Py0/1EPs+uREmfnUfXnkiUlZqQqab3Srps/mUjsjwBdcEJxVpKnrQsJb1a+J7uPTR7mrlv6318ctYnmZ9VHi0zdFj5YVhNVt48+KZ6UdsOyJsG9qEFzCMyQlegm4KyRf3r2jkKQKtiIfIq46qb6Pxr17/42Ts/45SZp/CdI74z6Humi1KFs0Rd1Pa+qtyOMe67XFsuOdactC0lvZRRtt2C15xHUJqZXR7/+yvQRKk7MIAobXsK3r5dS2SSuDytmihpkxqbqjjelWqtWOVylbwUCdMT6FHdBVKJUjgIkVBa34eJSjqi9Gcp5VXa7RptoerQp00G/ajMriQkQ7R6B69ensx9B1A7fRoBaaat6QA59GLv2TciUQKYXTCbrR1bVQuAYSY5JHLd8uto8bbw2I7HgNQxJeizYlK6JxK4eP7FeENentz1ZNrjcQeVP/+G45eqVgEmE3zqIbjyKbjxQyVKhTUpX59ny0Mio8eJa4NRfXi//RcWL+Shsx7ionkXpT3GlFgcCZbS4Nl3oCYcRY4idnfv5vZNtxOWYW5YcYNWKbwLUDGUw8oO461GLdlhOJl3QI+/h4iMJI9Hmkzg1Kyl3CrNY9AX53m1/lW++8Z3OaLiCG499ta4PlOp0EWpMne6So/f9bJ2/Pjvb+K5UhErSgBhZxEBLMwpjZ+cFGap/lWdqSZE3Q3w+PXK2rlArYNy+TrjLCVs2RTaC6MTz35ULlfWcPvO6HKNgZIcgEPbfQf8Js1tBkMkLi18EHRRSuz9M78yn3byCXY3sUBoRSWHmXmnM7dgbjTDJxOWEsARFUewonQFf/7ozwTCAbwhb3/3nSZCejHMdCwlgMXFi1lcvJiH6x5Oa70LxKwVscZcZKpWQM3R8QVGUxCtIqAlO+iilH/yD+DEbyd9zcLihWnVuRuUpIkO6XUWrc2vZf3B9Ty24zEumXeJWpDrKFBlhrT3bm3VWnZ07qCtt1VZSkNNcgA6tRhLQaqWFHpcKa+SaTnTaHAr9/P7Le/zn+v+k3lF87jtxNvSnpjoHoTKnCo13v2aqCaKUnZl//p3SfAGvThjhH5P1dncIS9gWmG8+Bc41f/RlaxySTgEj35BWS8X36UWaZttuILuGFESYM2iwF4Qfc/6EVPZIVXNuyj6QupD0X0nhDhSCPGfQKkQ4uaY2/eByZdnOAGJVi4eIB1Wp0vzQyfOGnPsFrpNhRRGOlnlUC3MR2op6XGl2DGOFCEE1y2/jubeZh7f9biylBITHRJjSmmKEsCl8y9lZ9dO3mt5L639o2tFbMNbxxEVJU2MdHHKK5yVVnr2iBhmogOoZIeDnoPYzDauWXaN2ugsUAtHtWMeWXUkAG/tfka5mIYhSvqsP2Xmpi5KuVVUZlfSE+jhg5YPuOHFG6jIruAPJ/+B7DSFFmLcd9kVahG6XrA2Nz4mWpldmbb7LsvS956ec+5FHP+Fn/Y1ONTI17oWdwaTiNKrP4f9b8In/1ct9zBboXgurohfTYb8bvVdMZkodBSmdt+VzFPWsSZKeba8aCuZfkQtpcnT/jyRgaZtNiAH1Z02N+bWA2TAB2EQV2poEBKrOcTid5RQIrpZbtmvfuy5IxOS2QV966UyJUoAR1UdxdKSpfx505/xBD393HdRS2kIMSWd02pOI9eam3Z6uO52y7UNLyEkWpk6EG8pDdbFNiMkTXRIX5QArlh0BSVOrXV6QqXwhUULKbAXsP7AOrV9GO67Tt9glpLmvtMsJYBrn78Wp8XJH0/545BqAkLf51GZXRlTGUX0VUPXqMyuxBVwpazJqNMb6o1z35XlOlg5o7/AWq3Z5IYjdMW0ugBg7+vw6s9g+eWw/NLo5mDpPLxIzVJyRycwBfYCfGFfXHXyKGaLatSpidK8wnmpY2zRivGHoKWkVW74AbBW+/sLKeUPpJT/K6Xckep1BumTpZnt6bgTBhIlkVNOqehmPlo8aYSB9FhLabC25kNBt5YOeg7SE+jpn+iQEFMaiihlWbM4e/bZPL/v+eT12xJI6r4bArr46MdJ5V4dFazOvhlxJKLup2mdnTjjRC6YewFXLr6yb2NCpXCTMLG2ci1vdWxRabYjsJSK7CnEJWop9bWVsJqs/PHkP/Z1jR0CuuUwM29mX2WU7NJ+a//SdZknxpRSYrZQEInQGStKnnZ49BoorIUzfxG3u7tETfhyTTZNlJSlrtfzGyjZIdL4YRqZd4O3MZnopOPgrhJCfIxqEYEQYrkQ4vejO6ypQ7ruhNheSonYCyspoZsZob0jdt2BcmlVZFdQ4izpF/cZKcdOO5aFRWomm3jsaPadHlMaYDFxMi6edzHBSJB/7vznoPvqYpI9TFdbqpiSPmMfVWLdd/rarzQvQlU5VfzgqB/Euy2TVAo/supIWkMedmXlD8vyHtRSKl0AedXgLGR+4XxOrzmdP5z8h2il+qGyuHgxfz/z76qeoL7mL6//uNP1TqQtSsCMUJgPA21qUbKUKrGhtw0u/mu/rEVXgepNmutzaVmTfZYSMGBcqSHswRvyDp4ODoempRTDr1Cp4O0AUsoPgU+M4pimFHo9rsEYSJQKSqdhERGsMjjiJAedFaUrmFuQ+SRL3VoC+v3oE2NKsZ1n02FO4RxWla/i4e0PJ69vFoM74MZpcQ7JGoslmfvOYXZkXMSTorvvpMxMtlVMpXCdIytVXOnN4sphWd7JOsTGccR18JV3QQgcFgc/P+7nLCsd/ndXCMHS0qXKrZU/Q8VUkohpXCHkARiKKJ3vDdEQ7uWNg2+o1O+6Z+CUH/UlKMTg0oQyz9OuRElbaqBbSrqY9x/4cups6veRlqV0iKeEI6U8kLApnHRHgyFTlaPWaAyWNdbj70npviuuiOkMnwFLCeBHR/+IX53wq4wcK5ETpp/Azatu5vSa0+O2jySmpHPJvEuod9fHl8pJgivoGrbrDtRKf7Mwx4nSmLjuQCswKlWVgGha8UhESZvsxDT6q8yppCYsecs+9O6+oGb8Ka0kUBmOVkfq50eCyQSf+E8Vz0mgNKsUi8mSUUvpxJCJEmHjoQ//pBbJzjsDjvhi0n17HMpyyu1pAr8rfUupbCF1dgeC+JhvP6ZISvgBIcRRgBRC2IQQX0Nz5RmMnIrsCnpDvQMu/AxFQriCrpSWkiVPZRhJs31YK++T4bA4yBqlL7YQgquWXMX0vOlx23UR0pMQhiNKJ888mUJ74aAVHlwB17Az70CrTG3L63Pf+XvGxnUHfRecYG9mLkJOLYAfG8/wu1nrcfNuxE0gtvV6mnT5u0ZUM3HEHPufsPi8fptNwkR5VvmgLvOhiJLVbOcCUyGvtn7AwZwSOO/3Ka1Ll+YFyOk6EOe+09+rlDEli50duUVMxzrw7/JQTnSI4TrgBmAaUA+s0B4bZIB0Aq96EF3vOtsPreClKF+kMnUmKf3WKQ0xpqQf47y557HuwLrUrhCU+264mXc6efa86GSiO9A9hpaSdsEJetNq8Dco0ey7mKoE7Ts50uvDK0N82Nq/ydxgdPm6Ulr2401VTtWALvOIjChRSvfCbrFx8Z73EEgeWXHWgH3MopU/2ndroqRZTrZcTMKU2lIC6qwW5vm80fVkSQkc4paSEMIM/EpK+WkpZbmUskxK+RkpZYra7wZDJZ21SqmqOUTR014z5LobL/SYku6+s4jhCezK0pWEZXjA2fBILSVQmXux65TG3lLyZsZSMpnBnhfnvqNtB4d7fZiFKb6VRZp0+jv7dYidKFRmVw74e9MLB6drKWG2UxEO84msah5t2RBtvZKMaNZnTxN4WqOiZDaZybflp7SUeoO97I94med1Q09D6rEc6u47KWUYtXh2eI5lg0FJJxtIr6uVUpRsObD2BliZuoryZEC3jDxBD1aTddg14oq0EjYd3tSp4e6ge0QxJeirfwdjHVPSLaUY991IA9sxpYYAaKsjF8HS4iXDEqUuX9eEFqVWb2u0mWQisW0r0qJwJtQex6VHfosOXwcv7n8x5a6uoAsTgiwp1YLlGAu3wJG6qsPu7t1IYF4gmLQ9ehTdfWcZpXjdGJDOVHQv8IYQ4gkg2tdaSvm/ozWoqUSRowiH2THgzE2fPeU7UrjvhIDTfzIKoxtbYtcpDTcrDoguvBxovZLe02Yk5NnyouVxxkeUvH2JDiOdGeulhnTa6qCwhiOnHc3tH97er+vxQATDQVxB14R131VmVxKREVp6W6ILd2OJbVuRFpf9HYCjhKA6p5oHtz/I6bWnJ93VFXCRa80hOt2KSRkfqP5dtLFfMKREacEnk48l2AsWZ1qlsiYq6Yz8IPCktm9sZQeDDCCEoMBREHXRJSNaITxVTOkQQY8pBSPBYcWTdIodamHmQKLkDrhH7L7TEx2CkSCeoGfs3XehDLnvQK1Vio0paTXvjqw6EolkfePA2YyxDFpiaJzRF+umSgvXY5ppW0omM5jMmISJi+dfzMbmjezq2pV0V1fARa49P+q2I+Y7WGAvSBkHreusw2lxMq1g9iCWUu+kTnKA9DrP/iDZbSQnFUIUCSGeF0Ls0P4m/fYKIU4XQmwXQuwUQtwSs/3nQohtQohNQojHhBAFIxnPeJNlyYr+EJIxaEzpECG2+OZILCWnxYnD7EjZ9tof9hOIBDLivusJ9PQFrydrogPEu+8iYWjfCSVzWVKyhBxrzpBceIMWYx1nBnOZD9l9F8N5c87DarKmLHcVbfBXOl9tiBGlQsfAltLcgrmYKlcM7r4b7dqLo8x42Xi3AC9KKecCL2qP49CSLH4HnAEsAi4XQmj1Q3geWCKlXAbUAd8ak1GPEoOKUqAbi7AMqUDlZCRWiEYiSkIIihxFKS2laLA5A+67sAxHM7lG1OJ8KFhiY0qj4L7r2g9hP5TMw2qycnjF4axvXJ92BXbd3TxhLaVBkotGIkpFjiJOrTmVJ3Y9EU3YicUVcKnJi16fL0ZA9KKsie+zlFKJUuFctSDX1QiuJA0BYWpYSqPEucDd2v27gfOS7LMG2Cml3C2lDAAPaK9DSvmclDKk7bceqE7y+klDljULb2yBzQT0ag4jbg43wTEJUzTjbiSiBAwoSnoxzky47wDqXfVxj0edWEsp6EW1PxjhhSjWfdemlbbUat4dWXUkDe4GDrgS19AnR7eUJmqig8PioMhRlNpS0n6LaaeEJ3Dp/EtxB908vefpfs/1BHrU9660vygV2AsIyVB0nZ5OS28L3f5uVclBrxLRtCn5ydPsrTWRSacd+tDK9aZHuZSyEUD7W5Zkn2lA7K+gXtuWyNVA/09fQwhxrRBioxBiY2vr4M30xoMs6+DuuzGbhY8zeixpxKLkHECU9ArhI3Tf6ZZWvXusRSkmJTzgUY9HOmFxFCirKxxUSQ7QJ0payaF0XXhRS2mCihJofZVGwX0HWomuwrk8uP3BflZP1H03/QhAQMGM6HOpirLu6Ipp7Kcv+2j8IPnJ9e/DJCYdS+ltIcTDQogzxRCm6kKIF4QQm5Pc0m2lnuxccZ+wEOLbQAi4L9VBpJR3SClXSylXl5aWpjv8MSXLkhWt95aMgSqEH2rocaWRJDqAspTafcljSvraokwsnoUYS2nMEh0SUsIzMTOOLTXUVqeqeGuLQGfmzaQqu4ondj8xaE1BgA6/mgxM5IlUVU7VoKKUdvZdAkIILp13KVs7trK5bXPcc269wd/0w+Ebu+PagqQqNRTX2M+RB0UDJDsEvVNClOYBdwCfBXYKIX4ihBi0lr2U8mQp5ZIkt8eBZiFEJYD2tyXJIeqB2Do01ahMQLTXfQ44C/i0TNfZPUHJsmYl9T/rDFSM9VBDt5Ay5b5L9tU4pNx3Q2jwNyCxlcITus0KIbh+xfVsat3E3z7+26CH6vJ1kWvNHfFnOJpUZFfQ6G5M+v0YqaUEcNbss8iyZPHg9gej20KREJ6gp28ylFD5IVpqKCHZoa6zjvKs8r5rQOXyQUTpEHffScXzUsrLgS8AnwM2CCFeEUIcOczzPqEdB+3v40n2eQeYK4So1RbvXqa9DiHE6cA3gXOklKmv5pOEwRIdppIo6VUdMiFKes3ARDLlvouKkua+G7OUfZMZzPa+RIdMJMDEVgpvq+vX2O+c2edw/PTj+fX7v2Z39+4BDzVoMdYJQFV2Fb6wL2m2WyZEKduazVmzzuKZvc9Es2d1b0iqyYv+niWmheuN/aJULlfJKL1J3NPBKeC+E0IUCyFuEkJsBL4GfAUoAf4T+Pswz3srcIoQYgdwivYYIUSVEOIpAC2R4cvAs6gCsA9JKbdor/8taq3U80KID4QQtw9zHBOCLGsW3pA3pWtkSrrvMiBKkLyqQ8ay7zR3XaO7UbXBGKHLcUhYnRD0aSnAGbgI6e67jj2qF1BCYz8hBN878ns4LA6+8/p3CEVCSQ6i6PKNczHWNIhm4CUpReUNeTEL84i/g5fMvwR/2B/t7zWY2ziZpRQMB9nTtae/KEHyZIepYCkBbwF5wHlSyk9KKf8hpQxJKTcCwxIDKWW7lPIkKeVc7W+Htv2glPLMmP2eklLOk1LOllL+OGb7HCnldCnlCu123XDGMVHItqiZrl5zKxZfyIc/7J8ylpJ+YbeMsLBssTP1AlpXwIVAjLgKeo41B4EgJENj57rTsWYpSymQfiv0AdHdd/Ub1N8k3WZLnCV854jv8FHbR9y15a6Uh+ryT9wSQzr6Atomd//CrN6QlyxL1oizXecXzWd56XIerlP9vQbrdpxtzcZissRZSru7dxOSoeSilMyFN0XWKc2XUv5ISlmf+ISU8n9GYUxTDv3imCzZQZ81TRVLSZ+d6m684TJQVQd3UFVzMImRrYgwCVN01jtmSQ46VoeWEu7JzEVId7fVv6P+JrjvdE6vPZ3Tak7jdx/8ju0d25Pu0+nvnPDf14Espd5Q74hcd7FcOv9S9vXs4+3Gt6OilCqWKYToV2ooLvNOJ6tINTJMKkqH8DolIcS/tHp3jwshnki8jeEYD3n0H0CyuFK0bcUUsZQyGVOC1JbSSONJOroojXkJKL37bCDD2XdNH4HZBgUzU+767SO+TZ4tj++88Z1+FbGllHT6Jm6FcJ0CewFOizNpBp43OIS2FYNwas2pFNgLeGj7Q2lV/ihwxJcaquusw2qyMjM/4fOoXNZflEIBiIQmvSgN5CP5xZiNYoqjW0rJMvCmmqWUqZRwPWicLC08E20rdPJseTTQMA7uO2dMSngGLCWrQ1WWDvmgbJFKpkhBoaOQ7x/5fW58+Ubu+OgObljR117NG/LiD/sn/PdVCKHWKiWpfzeUBn+DYTfbOW/Oedz78b0sKFoADBzLTLSU6jrrmJU/q/8krXIFbHsSfD0qTRwOibYVMIClJKV8ZaDbWA7yUEdfD2FYSplbPGs1Wcm35ydNdIiuFckAuttu7N13zr5+SplIdIA+F14K110sJ8w4gXNmn8OfNv2JLW1botujxVgnuKUEqRfQZlKUAC6edzFhGeb+bfcDA4tSgT2+fcWOjh3xrjsdPa7UHLMOKtp19hAVJSHEQ9rfj7TCp3G3sRvioY9hKfWRqXVKkLrUUCbdd7qFNOkTHaAv2SFJkkMyvrnmmxQ7i/n269/GH/YDMcVYJ8H3tTJnbERpRt4Mjqo6inZfOwIxYA1Lvf4dqNTwFm/LwKIU68I71C0l4Cbt71nA2UluBhlCz77zhPonOkw1SylTMSVIXdUh0+672L9jhtWpSsqE/ZnLtopaSumJUp4tjx8e9UN2de/id+//DugrkaPH9CYyldmVdPg6ouuSdDKZ6KBzyfxLAAZNsCmwF9Ad6CYcCbOjM0mSg05uOeRUpBClQzSmFFObbt/YDWdqoltKyYqydvm7cFqc2M32sR7WuJCpdUqgLow7u3b22z4a7rsxnzRYs9R6Iv1+JtCTHdJw3+kcPe1oLpp3EXdtuYsTZ5w4uSwlLQOvydNEbX5tdHumLSWA46qPozyrHLNIHasDZSnp6ePRzLuiFJOExMoOh7r7TkcIsVYI8Y4Qwi2ECAghwkKInrEY3FRhsJjSVLGSIHOJDpDcfSelVA3+rJPcUrI4+qp6Z2pmrLvvitMXJYCvrf4aVTlVfPv1b0cTByZLTAn6N/sbDVGymCx8+4hv8/mlnx9wv9j6d3WddRQ5iqLLG/pRuRxat/X11NItpUzFGMeJdBZq/Ba4HNgBOFGlhn4zmoOaagwUU+r2dx/yHWdjyWRMqdhRTLe/m2CkL23ZG/ISluHMWUq2cUx00MmU+276Gph9YlyL7nTItmbzo6N/xH7Xfu7cfGfc+q2JTFVOFdC/2d9oiBKo5BDdjZeK2KoOdR2qsV/KRbyVy0FGoOVj9ThqKU1u911aqwellDsBs5QyLKX8K3DC6A5ramE1WbEIS1JLqcvfNSlcIZki04kOEN8KIFMVwnWi7rvxWKeU7P5IOPwL8NnHhvfSisP59MJP4wl6KLAXjHhh8lhQmlWKSZjiFtBGZARfyDcqopQO0aUM3nZ2du1UlcFTEU12+ED9DWSo4eM4k843p1criPqBEOJnQoivApO7jsUEQwiRslJ4d2CKuu8yIUrO/gtoM1UhXOfYacdy48obWVi8MCPHS5s4S2liXIRuOuwmZubNpCwrWXu0iYfVZKUsqyzaORhUWS+JHHEJquGiW0qb2jbhC/uSJzno5FeDs7AvrnSIWErpFBj7LEq8vgx8FdVO4sLRHNRUJMuavKfSVCrGCpmPKUH8AtpMVQjXybZmc82yazJyrCERZylNjDmi0+Lkr6f9tV8220SmMrsyri16JiqEjwTdUtrQqGoQpkxyANXYMTbZISpKE+P7MFwGFSUp5T4hRKl2/wejP6SpSbL2FVLKKZfoMBruu1hLKdPuu3FjAlpKoFxik4nK7Eo+bO3LYBtvUXJanDjMDrZ2bMUkTMzOnz3wCyqXw1u/VyWGDpGU8IEWzwohxPeFEG3ANqBOCNEqhPju2A1v6pBMlFxBF2EZnlKilOl1ShDfviLT7rtxYzRiSlOQqpwqmj3NhCNhYPxFCZS1FJERZuTOwGFxDLxz5XKIBKF1a58oDfaaCc5AMaX/AI4GDpdSFkspC4EjgKO1uJJBBsmyZvVbp6QvnJ1K7rto6wrTyFpXgMqMs5gs8TGlDLvvxo3Y2bAhSsOmMruSkAzR5lVrviaCKOlxpQHjSTqVK9Tfxg+VKFmcYJr4SSYDMdDorwAul1Lu0TdIKXcDn9GeM8ggWdb+ltJUFKVMxpSEEP3WKh067ruY2fAEct9NNqJrlbS08IkgSvrvPS1RKqwFW64mShlq+DjODCRKVillW+JGKWUrMIYtNqcGWZb+iQ563bup5L7LZEwJ1FqlxOw7i8ky+StkTMBEh8lIKlHSF7SPB3qyQ1qiZDL1tbHIZB3EcWQgUQoM8zmDYZAsJXwqilImY0rQv6pDh6+DPFveiLuKjju6+85kAcvIGiJOZfQOtHoG3kSwlPRY6ICZd7FULoemzeDvmfRJDjBw9t3yFOWEBDC5I2kTkGSJDlPafZdBUdrbsxdQa1Be3P8iR1YdmZFjjyv6jPgQmBmPJ9nWbPJseRPKfbe6fDU7u3ZSlV2V3gsql0PIC02bwDnxC+EOxkAFWQeuHGiQUbKsWXhDXiIyEl0Nr4vSpI9/DIHREKV2bztSSp7Z+ww9gR4unX9pRo49rugzYkOURkxVTtWEEqWTZ57MyTNPTv8FemWHrv2QVz06gxpDJneaxiGE7sP2hXzRbV3+LnJtuRnJRJsszCucx5LiJXFVm0dCkbMIX9iHN+Tloe0PMSt/FqvLV2fk2OOKLkaHQGB7vKnIroi673QX+nhVdBgWxXNV1h0cEu47Q5QmCHrjr9hkh6lW9w7UBeL+s+6n2JmiMvIQ0f3zrzW8xkdtH3HJ/EsmfzwJYiwlI8lhpFRlV0VLDXlDXszCnDFLfUwwW6BiibpviNLwEEIUCSGeF0Ls0P4mrXMvhDhdCLFdCLFTCHFLkue/JoSQQoiS0R/16KK7C2LjSj3+niknSplGF6XbP7wdp8XJObPPGecRZQh9ZmxYSiOmKqcKd9BNT6AnWiF80k1cdBdepirGjyPjZSndArwopZwLvKg9jkMIYQZ+B5wBLAIuF0Isinl+OnAKsH9MRjzKJGtf0eXvGvuWCIcYei+anV07ObP2zEMnPmcygdluxJQyQEV2BaD6Ko1W24pRRxclw1IaNucCd2v37wbOS7LPGmCnlHK3lDIAPKC9Tuf/gG8AchTHOWYka/Q3Fd13mSa2LfchkeAQi9VpiFIG0LPcGj2No9IKfUyIitLk/z6MlyiVx7RbbwSS1bqfBhyIeVyvbUMIcQ7QIKX8MMnr4hBCXCuE2CiE2Nja2jrykY8SySwlw303cvQOqMtKlo19e4nRxppluO8ygL5WqdEziS2l0oXgKIDcivEeyYgZtbQuIcQLQLJ36NvpHiLJNimEyNKOcWo6B5FS3gHcAbB69eoJa1VlW7REh5BKdAhFQriCrim1cHY0cFgcfG7R5zh++vHjPZTMc+J3oGjWeI9i0lPkKMJmsk1u953FBl9+BxyT/3oxaqIkpUyZaC+EaBZCVEopG4UQlUBLkt3qUb2bdKqBg8BsoBb4UAtGVgPvCSHWSCmb+h1lkqBbSnpRVn2N0lRqhT5afO3wr433EEaHlZ8e7xEcEpiEiYrsiqilpE8QJx05k6O54mCMl/vuCeBz2v3PAY8n2ecdYK4QolbrfHsZ8ISU8iMpZZmUskZKWYMSr8MmsyBB/5jSVKzmYGAwXlTmVHLQc3DyWkqHEOMlSrcCpwghdqAy6G4FEEJUCSGeApBShlDdbp8FtgIPSSm3jNN4R53EmFJ3wBAlA4OxojK7kiZ3E96gF+chkME2mRmXUgFSynbgpCTbDwJnxjx+CnhqkGPVZHp844HVZMUiLFFLqcvXBUytYqwGBuNFVXYVrd5Wcmw541oh3MCo6DBhEELgtDqjFR2mYoVwA4PxojKnEonEFXAZ7rtxxhClCUS2NbvPfWfElAwMxgy9rxKMbzFWA0OUJhSx7Su6A91YhCVaE8/AwGD0iG0TYYjS+GKI0gQiVpT0EkOTrgaXgcEkpDy7PHrfEKXxxRClCUSWNStunZLhujMwGBtsZhulzlLAEKXxxhClCUSWJSsu0cEQJQODsUOPKxkp4eOLIUoTiCxrVtziWSPzzsBg7NBr4Bkp4eOLIUoTiCxrVjT7rsvfZYiSgcEYErWUDPfduGKI0gQiLvvOiCkZGIwphihNDAxRmkBkWbPwhrz0Bnvxh/2GpWRgMIYcVn4Y1TnVVOdUj/dQpjTjUmbIIDm6L7upV9WWNUTJwGDsWFC0gKcvfHq8hzHlMSylCYS+UPag+yBgVHMwMDCYehiiNIHQfdmGKBkYGExVDFGaQOjtK5o8hvvOwMBgamKI0gRCjykd9ChLyeg6a2BgMNUwRGkCoVtKje5GAAocBeM4GgMDA4OxxxClCUSspeS0OLGb7eM8IgMDA4OxxRClCYSefdfS22LEkwwMDKYkhihNIHRLKSIjRjzJwMBgSmKI0gRCjymBkQ5uYGAwNTFEaQJhNVmxCFVkw3DfGRgYTEXGRZSEEEVCiOeFEDu0v4Up9jtdCLFdCLFTCHFLwnNf0Z7bIoT42diMfHQRQkR7uRiiZGBgMBUZL0vpFuBFKeVc4EXtcRxCCDPwO+AMYBFwuRBikfbcCcC5wDIp5WLgF2M18NFGT3Yw3HcGBgZTkfESpXOBu7X7dwPnJdlnDbBTSrlbShkAHtBeB/Al4FYppR9AStkyusMdO/RkB8NSMjAwmIqMlyiVSykbAbS/ZUn2mQYciHlcr20DmAccK4R4WwjxihDi8FQnEkJcK4TYKITY2NramqHhjx66KBmWkoGBwVRk1FpXCCFeACqSPPXtdA+RZJvU/lqAQmAtcDjwkBBilpRS9nuBlHcAdwCsXr263/MTDT0Dz7CUDAwMpiKjJkpSypNTPSeEaBZCVEopG4UQlUAy91s9MD3mcTVwMOa5f2gitEEIEQFKgIlvCg2CYSkZGBhMZcbLffcE8Dnt/ueAx5Ps8w4wVwhRK4SwAZdprwP4J3AigBBiHmAD2kZzwGOFYSkZGBhMZcZLlG4FThFC7ABO0R4jhKgSQjwFIKUMAV8GngW2Ag9JKbdor/8LMEsIsRmVAPG5ZK67yYguSoalZGBgMBUZl3boUsp24KQk2w8CZ8Y8fgp4Ksl+AeAzoznG8UJ33+Xacsd5JAYGBgZjz7iIkkFqzp59NuVZ5VhMxkdjYGAw9TCufBOMBUULWFC0YLyHYWBgYDAuGLXvDAwMDAwmDIYoGRgYGBhMGAxRMjAwMDCYMBiiZGBgYGAwYTBEycDAwMBgwmCIkoGBgYHBhMEQJQMDAwODCYMhSgYGBgYGEwZxiJSMSwshRCuwb5gvL2FyF32d7ONPh8n8P07msQ/EZP+/Jvv40yGd/3GmlLJ0LAYzpURpJAghNkopV4/3OIbLZB9/Okzm/3Eyj30gJvv/NdnHnw4T7X803HcGBgYGBhMGQ5QMDAwMDCYMhiilzx3jPYARMtnHnw6T+X+czGMfiMn+f0328afDhPofjZiSgYGBgcGEwbCUDAwMDAwmDIYoGRgYGBhMHKSUk/IGTAdeBrYCW4CbtO1FwPPADu1vobb9FOBd4CPt74kxx1qlbd8J/BrNrZnknD8GDgDuhO124EHt9W8DNSler++3F+jR9t8C3AQ8A3QDrRN07J8A3gNCwEUJz4WBD7TbExP0M0o5/knwGaU79nQ+oy2oNSkT4TO5GfgY2AS8iFoLM1k+k3THPt6/m6TjTzLOpO/HEL57Ka8j2mfUBTw50Bii+6ez00S8AZXAYdr9XKAOWAT8DLhF234L8D/a/ZVAlXZ/CdAQc6wNwJGAAJ4GzkhxzrXaeRO/oNcDt2v3LwMeTPH664HbtWPcon2I+tivBh4Ftk7QsdcAy4B7kvy4kn7hJ9hnlHL8k+AzSnfsg35GE+wzOQHI0u5/aYDv3kT8TNId+6CfySj/bpKOP8k5k74fQ/jupbyOACcBZ3Ooi1KSN+Vx1IxhO1AZ8+FuT7KvANpR6l4JbIt57nLgj4OcK/EL+ixwpHbfgpqJ9pt5pdovZuz7gecn4thj9r8r3R/XRPqMBhr/RP+M0h37cD6jifCZaM+tBN6YbJ/JYGMfz9/NEMY/6Psx3N9NzPPHk6YoHRIxJSFEDeqL8TZQLqVsBND+liV5yYXA+1JKPzANqI95rl7bNhSmoUxkpJQhlDuhOM39VsSMvQjwT9CxD4RDCLFRCLFeCHFesh0mwGeULhPxM8oE/T6jCfaZfB41Q0/GRP9MBhr7QIz27yZdJsp1BFCKNqkRQuSgzPf/kFL2CCEG238x8D/AqfqmJLvJoQ4jzWMk7ieAvzI5xj4QM6SUB4UQs4CXhBAfSSl3RU8wMT6jdJmIn1EmSPyMdqFmv+P+fwkhPgOsBo5LtUuSxxPiM0lj7AMx2r+bdJko1xFgkmffCSGsqA/tPinlP7TNzUKISu35SqAlZv9q4DHgipgPvx6ojjlsNXBQCGEWQnyg3X44yFDqUcFJhBAWIB/oEEL8WD9Gkv0c2rnuiRl7B8q9MBHHnhIp5UHt725gHWpmp497onxGSZkkn1G6Y09Jwmf0CnAvE+AzEUKcDHwbOEef3U+WzyTNsadkDH43qcad+H8mfT8GOcZAn1H0OjLQMVKSjo9vIt5QynwP8KuE7T8nPhj4M+1+AfAhcGGSY72DCgjqQb4zBzl3YlzmBuKDfA+leN0NqICtAF4F6hKef4D4gO2EGXvM/ncR41sGCtH816hqwzuARRPtM0o1/snwGaU79iF8Rj2oi/p4/25WAruAuYO8bsJ9JumOfbx/N4N9p9J9Pwb77jHIdYQhxJQG3WGi3oBjUObhJvrSKs9E+TFf1D7kF4Eibf/vAJ6YfT8AyrTnVgObtS/Zb0mdHvoz1Iwgov39vrbdATyMSofcAMxK8Xp9v3pt7NtixrIZFRwMA0Hg/Qk29sO113lQgdQt2vajUKmkH2p/Pz9BP6Ok458kn1G6Y0/nM9o1gT6TF4BmkqRFT4LPJN2xj/fvJun4k4wz6fuRavwDfEb9riPAa6iUfa92rNMGurYbZYYMDAwMDCYMkzqmZGBgYGBwaGGIkoGBgYHBhMEQJQMDAwODCYMhSgYGBgYGEwZDlAwMDAwMJgyGKBkYjBJCiLC2wHCLEOJDIcTNQogBf3NCiBohxKfGaowGBhMNQ5QMDEYPr5RyhZRyMaqg5pnA9wZ5TQ1giJLBlMVYp2RgMEoIIdxSypyYx7NQK+dLgJmoUj/Z2tNfllK+KYRYDywE9gB3o3rb3IpaEW8Hfiel/OOY/RMGBmOMIUoGBqNEoihp2zqBBYALiEgpfUKIucD9UsrVQojjga9JKc/S9r8WtTr/v4UQduAN4GIp5Z6x/F8MDMaKSV8l3MBgkqFXU7YCvxVCrECVyJmXYv9TgWVCiIu0x/nAXJQlZWBwyGGIkoHBGKG578Ko6s7fQ9VOW46K7fpSvQz4ipTy2TEZpIHBOGMkOhgYjAFCiFJUpevfSuUzzwcapZQR4LOAWdvVhWqBrfMs8CWtlQFCiHlCiGwMDA5RDEvJwGD0cGr9ZqxACJXY8L/ac78HHhVCXAy8jKrCDKoydEgI8SGqXcBtqIy894Tq8tYKnDc2wzcwGHuMRAcDAwMDgwmD4b4zMDAwMPj/7dWxAAAAAMAgf+tp7CiJNqQEwIaUANiQEgAbUgJgQ0oAbEgJgI0AJWNGyOazqPoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.dates as mdates\n",
    "months = mdates.MonthLocator()\n",
    "days = mdates.WeekdayLocator()\n",
    "months_fmt = mdates.DateFormatter('%M')\n",
    "plt.plot(arima_data.index, arima_data['AAPL'], label='ARIMA')\n",
    "plt.plot(lstm_data.index, lstm_data['AAPL'], label='LSTM')\n",
    "plt.plot(aapl_returns, label='true value')\n",
    "plt.xlabel('Date')\n",
    "plt.ylabel('Daily return prediction')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}