crypto_predict.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'columns': ['Close', 'Volume', 'Low', 'High'],\n",
       " 'csv_src_file': 'BTC_ETH',\n",
       " 'filename': 'BTC_ETH_lstm_i30_o12_Close_Volume_Low_High',\n",
       " 'folder': {'data': 'data/', 'weights': 'weights/'},\n",
       " 'input_size': 30,\n",
       " 'lstm_hidden_size': 50,\n",
       " 'name': 'lstm',\n",
       " 'output_size': 12,\n",
       " 'pair': 'BTC_ETH',\n",
       " 'period': 300}"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from config import CONFIG\n",
    "from utils import series_to_supervised\n",
    "\n",
    "CONFIG"
   ]
  },
  {
   "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>Close</th>\n",
       "      <th>Timestamp</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Open</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>269507</th>\n",
       "      <td>0.082828</td>\n",
       "      <td>1519862700</td>\n",
       "      <td>0.082856</td>\n",
       "      <td>0.082729</td>\n",
       "      <td>0.082729</td>\n",
       "      <td>4.151247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269508</th>\n",
       "      <td>0.082609</td>\n",
       "      <td>1519863000</td>\n",
       "      <td>0.082828</td>\n",
       "      <td>0.082606</td>\n",
       "      <td>0.082828</td>\n",
       "      <td>5.551513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269509</th>\n",
       "      <td>0.082552</td>\n",
       "      <td>1519863300</td>\n",
       "      <td>0.082673</td>\n",
       "      <td>0.082547</td>\n",
       "      <td>0.082609</td>\n",
       "      <td>2.327443</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269510</th>\n",
       "      <td>0.082460</td>\n",
       "      <td>1519863600</td>\n",
       "      <td>0.082625</td>\n",
       "      <td>0.082419</td>\n",
       "      <td>0.082552</td>\n",
       "      <td>1.519736</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269511</th>\n",
       "      <td>0.082455</td>\n",
       "      <td>1519863900</td>\n",
       "      <td>0.082460</td>\n",
       "      <td>0.082418</td>\n",
       "      <td>0.082455</td>\n",
       "      <td>0.552411</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Close   Timestamp      High       Low      Open    Volume\n",
       "269507  0.082828  1519862700  0.082856  0.082729  0.082729  4.151247\n",
       "269508  0.082609  1519863000  0.082828  0.082606  0.082828  5.551513\n",
       "269509  0.082552  1519863300  0.082673  0.082547  0.082609  2.327443\n",
       "269510  0.082460  1519863600  0.082625  0.082419  0.082552  1.519736\n",
       "269511  0.082455  1519863900  0.082460  0.082418  0.082455  0.552411"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from pandas import read_csv\n",
    "from pandas import DataFrame\n",
    "from pandas import concat\n",
    "from matplotlib import pyplot\n",
    " \n",
    "#data file path\n",
    "dfp = ''.join([CONFIG['folder']['data'], CONFIG['csv_src_file'], '.csv'])\n",
    "\n",
    "#Columns of price data to use\n",
    "columns = CONFIG['columns']\n",
    "# df = pd.read_csv(dfp).dropna().tail(1000000)\n",
    "dataset = pd.read_csv(dfp)\n",
    "\n",
    "# to drop values before 2018 1514764800, March 2018 1519862400, July 2017 1498867200\n",
    "dataset = dataset[dataset.Timestamp > 1519862400]\n",
    "dataset.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "values = dataset.loc[:,columns].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(30, 4, 12)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# parameters to prepare the dataset for learning \n",
    "n_lag = CONFIG['input_size']\n",
    "n_out = CONFIG['output_size']\n",
    "n_features = len(columns)\n",
    "n_lag,n_features,n_out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# scale dataset\n",
    "scaler = MinMaxScaler(feature_range=(0, 1))\n",
    "scaled = scaler.fit_transform(values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>var1(t-30)</th>\n",
       "      <th>var2(t-30)</th>\n",
       "      <th>var3(t-30)</th>\n",
       "      <th>var4(t-30)</th>\n",
       "      <th>var1(t-29)</th>\n",
       "      <th>var2(t-29)</th>\n",
       "      <th>var3(t-29)</th>\n",
       "      <th>var4(t-29)</th>\n",
       "      <th>var1(t-28)</th>\n",
       "      <th>var2(t-28)</th>\n",
       "      <th>...</th>\n",
       "      <th>var3(t+9)</th>\n",
       "      <th>var4(t+9)</th>\n",
       "      <th>var1(t+10)</th>\n",
       "      <th>var2(t+10)</th>\n",
       "      <th>var3(t+10)</th>\n",
       "      <th>var4(t+10)</th>\n",
       "      <th>var1(t+11)</th>\n",
       "      <th>var2(t+11)</th>\n",
       "      <th>var3(t+11)</th>\n",
       "      <th>var4(t+11)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>0.911499</td>\n",
       "      <td>0.005383</td>\n",
       "      <td>0.909966</td>\n",
       "      <td>0.888220</td>\n",
       "      <td>0.904896</td>\n",
       "      <td>0.007199</td>\n",
       "      <td>0.906295</td>\n",
       "      <td>0.887362</td>\n",
       "      <td>0.903185</td>\n",
       "      <td>0.003018</td>\n",
       "      <td>...</td>\n",
       "      <td>0.902730</td>\n",
       "      <td>0.882924</td>\n",
       "      <td>0.903218</td>\n",
       "      <td>0.000487</td>\n",
       "      <td>0.902156</td>\n",
       "      <td>0.882912</td>\n",
       "      <td>0.903810</td>\n",
       "      <td>0.001179</td>\n",
       "      <td>0.902319</td>\n",
       "      <td>0.882907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>0.904896</td>\n",
       "      <td>0.007199</td>\n",
       "      <td>0.906295</td>\n",
       "      <td>0.887362</td>\n",
       "      <td>0.903185</td>\n",
       "      <td>0.003018</td>\n",
       "      <td>0.904533</td>\n",
       "      <td>0.882607</td>\n",
       "      <td>0.900404</td>\n",
       "      <td>0.001971</td>\n",
       "      <td>...</td>\n",
       "      <td>0.902156</td>\n",
       "      <td>0.882912</td>\n",
       "      <td>0.903810</td>\n",
       "      <td>0.001179</td>\n",
       "      <td>0.902319</td>\n",
       "      <td>0.882907</td>\n",
       "      <td>0.909454</td>\n",
       "      <td>0.001097</td>\n",
       "      <td>0.905317</td>\n",
       "      <td>0.885319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>0.903185</td>\n",
       "      <td>0.003018</td>\n",
       "      <td>0.904533</td>\n",
       "      <td>0.882607</td>\n",
       "      <td>0.900404</td>\n",
       "      <td>0.001971</td>\n",
       "      <td>0.900683</td>\n",
       "      <td>0.881134</td>\n",
       "      <td>0.900251</td>\n",
       "      <td>0.000716</td>\n",
       "      <td>...</td>\n",
       "      <td>0.902319</td>\n",
       "      <td>0.882907</td>\n",
       "      <td>0.909454</td>\n",
       "      <td>0.001097</td>\n",
       "      <td>0.905317</td>\n",
       "      <td>0.885319</td>\n",
       "      <td>0.907357</td>\n",
       "      <td>0.000555</td>\n",
       "      <td>0.908805</td>\n",
       "      <td>0.885254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>0.900404</td>\n",
       "      <td>0.001971</td>\n",
       "      <td>0.900683</td>\n",
       "      <td>0.881134</td>\n",
       "      <td>0.900251</td>\n",
       "      <td>0.000716</td>\n",
       "      <td>0.900682</td>\n",
       "      <td>0.876073</td>\n",
       "      <td>0.899950</td>\n",
       "      <td>0.001567</td>\n",
       "      <td>...</td>\n",
       "      <td>0.905317</td>\n",
       "      <td>0.885319</td>\n",
       "      <td>0.907357</td>\n",
       "      <td>0.000555</td>\n",
       "      <td>0.908805</td>\n",
       "      <td>0.885254</td>\n",
       "      <td>0.911044</td>\n",
       "      <td>0.007564</td>\n",
       "      <td>0.906050</td>\n",
       "      <td>0.887050</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>0.900251</td>\n",
       "      <td>0.000716</td>\n",
       "      <td>0.900682</td>\n",
       "      <td>0.876073</td>\n",
       "      <td>0.899950</td>\n",
       "      <td>0.001567</td>\n",
       "      <td>0.899785</td>\n",
       "      <td>0.879122</td>\n",
       "      <td>0.900536</td>\n",
       "      <td>0.001585</td>\n",
       "      <td>...</td>\n",
       "      <td>0.908805</td>\n",
       "      <td>0.885254</td>\n",
       "      <td>0.911044</td>\n",
       "      <td>0.007564</td>\n",
       "      <td>0.906050</td>\n",
       "      <td>0.887050</td>\n",
       "      <td>0.910657</td>\n",
       "      <td>0.003413</td>\n",
       "      <td>0.911403</td>\n",
       "      <td>0.887051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>0.899950</td>\n",
       "      <td>0.001567</td>\n",
       "      <td>0.899785</td>\n",
       "      <td>0.879122</td>\n",
       "      <td>0.900536</td>\n",
       "      <td>0.001585</td>\n",
       "      <td>0.899590</td>\n",
       "      <td>0.876207</td>\n",
       "      <td>0.898062</td>\n",
       "      <td>0.002247</td>\n",
       "      <td>...</td>\n",
       "      <td>0.906050</td>\n",
       "      <td>0.887050</td>\n",
       "      <td>0.910657</td>\n",
       "      <td>0.003413</td>\n",
       "      <td>0.911403</td>\n",
       "      <td>0.887051</td>\n",
       "      <td>0.907522</td>\n",
       "      <td>0.008177</td>\n",
       "      <td>0.907607</td>\n",
       "      <td>0.887051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>0.900536</td>\n",
       "      <td>0.001585</td>\n",
       "      <td>0.899590</td>\n",
       "      <td>0.876207</td>\n",
       "      <td>0.898062</td>\n",
       "      <td>0.002247</td>\n",
       "      <td>0.899590</td>\n",
       "      <td>0.877975</td>\n",
       "      <td>0.900530</td>\n",
       "      <td>0.001073</td>\n",
       "      <td>...</td>\n",
       "      <td>0.911403</td>\n",
       "      <td>0.887051</td>\n",
       "      <td>0.907522</td>\n",
       "      <td>0.008177</td>\n",
       "      <td>0.907607</td>\n",
       "      <td>0.887051</td>\n",
       "      <td>0.900615</td>\n",
       "      <td>0.003681</td>\n",
       "      <td>0.902132</td>\n",
       "      <td>0.887033</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>0.898062</td>\n",
       "      <td>0.002247</td>\n",
       "      <td>0.899590</td>\n",
       "      <td>0.877975</td>\n",
       "      <td>0.900530</td>\n",
       "      <td>0.001073</td>\n",
       "      <td>0.899590</td>\n",
       "      <td>0.876218</td>\n",
       "      <td>0.898052</td>\n",
       "      <td>0.000990</td>\n",
       "      <td>...</td>\n",
       "      <td>0.907607</td>\n",
       "      <td>0.887051</td>\n",
       "      <td>0.900615</td>\n",
       "      <td>0.003681</td>\n",
       "      <td>0.902132</td>\n",
       "      <td>0.887033</td>\n",
       "      <td>0.906061</td>\n",
       "      <td>0.001641</td>\n",
       "      <td>0.901759</td>\n",
       "      <td>0.881837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>0.900530</td>\n",
       "      <td>0.001073</td>\n",
       "      <td>0.899590</td>\n",
       "      <td>0.876218</td>\n",
       "      <td>0.898052</td>\n",
       "      <td>0.000990</td>\n",
       "      <td>0.899590</td>\n",
       "      <td>0.877294</td>\n",
       "      <td>0.900246</td>\n",
       "      <td>0.001661</td>\n",
       "      <td>...</td>\n",
       "      <td>0.902132</td>\n",
       "      <td>0.887033</td>\n",
       "      <td>0.906061</td>\n",
       "      <td>0.001641</td>\n",
       "      <td>0.901759</td>\n",
       "      <td>0.881837</td>\n",
       "      <td>0.899258</td>\n",
       "      <td>0.016586</td>\n",
       "      <td>0.895942</td>\n",
       "      <td>0.881826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>0.898052</td>\n",
       "      <td>0.000990</td>\n",
       "      <td>0.899590</td>\n",
       "      <td>0.877294</td>\n",
       "      <td>0.900246</td>\n",
       "      <td>0.001661</td>\n",
       "      <td>0.899590</td>\n",
       "      <td>0.876158</td>\n",
       "      <td>0.900536</td>\n",
       "      <td>0.000313</td>\n",
       "      <td>...</td>\n",
       "      <td>0.901759</td>\n",
       "      <td>0.881837</td>\n",
       "      <td>0.899258</td>\n",
       "      <td>0.016586</td>\n",
       "      <td>0.895942</td>\n",
       "      <td>0.881826</td>\n",
       "      <td>0.903952</td>\n",
       "      <td>0.012982</td>\n",
       "      <td>0.899591</td>\n",
       "      <td>0.883312</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 168 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    var1(t-30)  var2(t-30)  var3(t-30)  var4(t-30)  var1(t-29)  var2(t-29)  \\\n",
       "30    0.911499    0.005383    0.909966    0.888220    0.904896    0.007199   \n",
       "31    0.904896    0.007199    0.906295    0.887362    0.903185    0.003018   \n",
       "32    0.903185    0.003018    0.904533    0.882607    0.900404    0.001971   \n",
       "33    0.900404    0.001971    0.900683    0.881134    0.900251    0.000716   \n",
       "34    0.900251    0.000716    0.900682    0.876073    0.899950    0.001567   \n",
       "35    0.899950    0.001567    0.899785    0.879122    0.900536    0.001585   \n",
       "36    0.900536    0.001585    0.899590    0.876207    0.898062    0.002247   \n",
       "37    0.898062    0.002247    0.899590    0.877975    0.900530    0.001073   \n",
       "38    0.900530    0.001073    0.899590    0.876218    0.898052    0.000990   \n",
       "39    0.898052    0.000990    0.899590    0.877294    0.900246    0.001661   \n",
       "\n",
       "    var3(t-29)  var4(t-29)  var1(t-28)  var2(t-28)     ...      var3(t+9)  \\\n",
       "30    0.906295    0.887362    0.903185    0.003018     ...       0.902730   \n",
       "31    0.904533    0.882607    0.900404    0.001971     ...       0.902156   \n",
       "32    0.900683    0.881134    0.900251    0.000716     ...       0.902319   \n",
       "33    0.900682    0.876073    0.899950    0.001567     ...       0.905317   \n",
       "34    0.899785    0.879122    0.900536    0.001585     ...       0.908805   \n",
       "35    0.899590    0.876207    0.898062    0.002247     ...       0.906050   \n",
       "36    0.899590    0.877975    0.900530    0.001073     ...       0.911403   \n",
       "37    0.899590    0.876218    0.898052    0.000990     ...       0.907607   \n",
       "38    0.899590    0.877294    0.900246    0.001661     ...       0.902132   \n",
       "39    0.899590    0.876158    0.900536    0.000313     ...       0.901759   \n",
       "\n",
       "    var4(t+9)  var1(t+10)  var2(t+10)  var3(t+10)  var4(t+10)  var1(t+11)  \\\n",
       "30   0.882924    0.903218    0.000487    0.902156    0.882912    0.903810   \n",
       "31   0.882912    0.903810    0.001179    0.902319    0.882907    0.909454   \n",
       "32   0.882907    0.909454    0.001097    0.905317    0.885319    0.907357   \n",
       "33   0.885319    0.907357    0.000555    0.908805    0.885254    0.911044   \n",
       "34   0.885254    0.911044    0.007564    0.906050    0.887050    0.910657   \n",
       "35   0.887050    0.910657    0.003413    0.911403    0.887051    0.907522   \n",
       "36   0.887051    0.907522    0.008177    0.907607    0.887051    0.900615   \n",
       "37   0.887051    0.900615    0.003681    0.902132    0.887033    0.906061   \n",
       "38   0.887033    0.906061    0.001641    0.901759    0.881837    0.899258   \n",
       "39   0.881837    0.899258    0.016586    0.895942    0.881826    0.903952   \n",
       "\n",
       "    var2(t+11)  var3(t+11)  var4(t+11)  \n",
       "30    0.001179    0.902319    0.882907  \n",
       "31    0.001097    0.905317    0.885319  \n",
       "32    0.000555    0.908805    0.885254  \n",
       "33    0.007564    0.906050    0.887050  \n",
       "34    0.003413    0.911403    0.887051  \n",
       "35    0.008177    0.907607    0.887051  \n",
       "36    0.003681    0.902132    0.887033  \n",
       "37    0.001641    0.901759    0.881837  \n",
       "38    0.016586    0.895942    0.881826  \n",
       "39    0.012982    0.899591    0.883312  \n",
       "\n",
       "[10 rows x 168 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# frame as supervised learning\n",
    "reframed = series_to_supervised(scaled, n_lag, n_out)\n",
    "reframed.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>var1(t-30)</th>\n",
       "      <th>var2(t-30)</th>\n",
       "      <th>var3(t-30)</th>\n",
       "      <th>var4(t-30)</th>\n",
       "      <th>var1(t-29)</th>\n",
       "      <th>var2(t-29)</th>\n",
       "      <th>var3(t-29)</th>\n",
       "      <th>var4(t-29)</th>\n",
       "      <th>var1(t-28)</th>\n",
       "      <th>var2(t-28)</th>\n",
       "      <th>...</th>\n",
       "      <th>var1(t+2)</th>\n",
       "      <th>var1(t+3)</th>\n",
       "      <th>var1(t+4)</th>\n",
       "      <th>var1(t+5)</th>\n",
       "      <th>var1(t+6)</th>\n",
       "      <th>var1(t+7)</th>\n",
       "      <th>var1(t+8)</th>\n",
       "      <th>var1(t+9)</th>\n",
       "      <th>var1(t+10)</th>\n",
       "      <th>var1(t+11)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>0.911499</td>\n",
       "      <td>0.005383</td>\n",
       "      <td>0.909966</td>\n",
       "      <td>0.888220</td>\n",
       "      <td>0.904896</td>\n",
       "      <td>0.007199</td>\n",
       "      <td>0.906295</td>\n",
       "      <td>0.887362</td>\n",
       "      <td>0.903185</td>\n",
       "      <td>0.003018</td>\n",
       "      <td>...</td>\n",
       "      <td>0.908729</td>\n",
       "      <td>0.903118</td>\n",
       "      <td>0.904913</td>\n",
       "      <td>0.909531</td>\n",
       "      <td>0.907403</td>\n",
       "      <td>0.907393</td>\n",
       "      <td>0.907137</td>\n",
       "      <td>0.903187</td>\n",
       "      <td>0.903218</td>\n",
       "      <td>0.903810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>0.904896</td>\n",
       "      <td>0.007199</td>\n",
       "      <td>0.906295</td>\n",
       "      <td>0.887362</td>\n",
       "      <td>0.903185</td>\n",
       "      <td>0.003018</td>\n",
       "      <td>0.904533</td>\n",
       "      <td>0.882607</td>\n",
       "      <td>0.900404</td>\n",
       "      <td>0.001971</td>\n",
       "      <td>...</td>\n",
       "      <td>0.903118</td>\n",
       "      <td>0.904913</td>\n",
       "      <td>0.909531</td>\n",
       "      <td>0.907403</td>\n",
       "      <td>0.907393</td>\n",
       "      <td>0.907137</td>\n",
       "      <td>0.903187</td>\n",
       "      <td>0.903218</td>\n",
       "      <td>0.903810</td>\n",
       "      <td>0.909454</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>0.903185</td>\n",
       "      <td>0.003018</td>\n",
       "      <td>0.904533</td>\n",
       "      <td>0.882607</td>\n",
       "      <td>0.900404</td>\n",
       "      <td>0.001971</td>\n",
       "      <td>0.900683</td>\n",
       "      <td>0.881134</td>\n",
       "      <td>0.900251</td>\n",
       "      <td>0.000716</td>\n",
       "      <td>...</td>\n",
       "      <td>0.904913</td>\n",
       "      <td>0.909531</td>\n",
       "      <td>0.907403</td>\n",
       "      <td>0.907393</td>\n",
       "      <td>0.907137</td>\n",
       "      <td>0.903187</td>\n",
       "      <td>0.903218</td>\n",
       "      <td>0.903810</td>\n",
       "      <td>0.909454</td>\n",
       "      <td>0.907357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>0.900404</td>\n",
       "      <td>0.001971</td>\n",
       "      <td>0.900683</td>\n",
       "      <td>0.881134</td>\n",
       "      <td>0.900251</td>\n",
       "      <td>0.000716</td>\n",
       "      <td>0.900682</td>\n",
       "      <td>0.876073</td>\n",
       "      <td>0.899950</td>\n",
       "      <td>0.001567</td>\n",
       "      <td>...</td>\n",
       "      <td>0.909531</td>\n",
       "      <td>0.907403</td>\n",
       "      <td>0.907393</td>\n",
       "      <td>0.907137</td>\n",
       "      <td>0.903187</td>\n",
       "      <td>0.903218</td>\n",
       "      <td>0.903810</td>\n",
       "      <td>0.909454</td>\n",
       "      <td>0.907357</td>\n",
       "      <td>0.911044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>0.900251</td>\n",
       "      <td>0.000716</td>\n",
       "      <td>0.900682</td>\n",
       "      <td>0.876073</td>\n",
       "      <td>0.899950</td>\n",
       "      <td>0.001567</td>\n",
       "      <td>0.899785</td>\n",
       "      <td>0.879122</td>\n",
       "      <td>0.900536</td>\n",
       "      <td>0.001585</td>\n",
       "      <td>...</td>\n",
       "      <td>0.907403</td>\n",
       "      <td>0.907393</td>\n",
       "      <td>0.907137</td>\n",
       "      <td>0.903187</td>\n",
       "      <td>0.903218</td>\n",
       "      <td>0.903810</td>\n",
       "      <td>0.909454</td>\n",
       "      <td>0.907357</td>\n",
       "      <td>0.911044</td>\n",
       "      <td>0.910657</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 132 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    var1(t-30)  var2(t-30)  var3(t-30)  var4(t-30)  var1(t-29)  var2(t-29)  \\\n",
       "30    0.911499    0.005383    0.909966    0.888220    0.904896    0.007199   \n",
       "31    0.904896    0.007199    0.906295    0.887362    0.903185    0.003018   \n",
       "32    0.903185    0.003018    0.904533    0.882607    0.900404    0.001971   \n",
       "33    0.900404    0.001971    0.900683    0.881134    0.900251    0.000716   \n",
       "34    0.900251    0.000716    0.900682    0.876073    0.899950    0.001567   \n",
       "\n",
       "    var3(t-29)  var4(t-29)  var1(t-28)  var2(t-28)     ...      var1(t+2)  \\\n",
       "30    0.906295    0.887362    0.903185    0.003018     ...       0.908729   \n",
       "31    0.904533    0.882607    0.900404    0.001971     ...       0.903118   \n",
       "32    0.900683    0.881134    0.900251    0.000716     ...       0.904913   \n",
       "33    0.900682    0.876073    0.899950    0.001567     ...       0.909531   \n",
       "34    0.899785    0.879122    0.900536    0.001585     ...       0.907403   \n",
       "\n",
       "    var1(t+3)  var1(t+4)  var1(t+5)  var1(t+6)  var1(t+7)  var1(t+8)  \\\n",
       "30   0.903118   0.904913   0.909531   0.907403   0.907393   0.907137   \n",
       "31   0.904913   0.909531   0.907403   0.907393   0.907137   0.903187   \n",
       "32   0.909531   0.907403   0.907393   0.907137   0.903187   0.903218   \n",
       "33   0.907403   0.907393   0.907137   0.903187   0.903218   0.903810   \n",
       "34   0.907393   0.907137   0.903187   0.903218   0.903810   0.909454   \n",
       "\n",
       "    var1(t+9)  var1(t+10)  var1(t+11)  \n",
       "30   0.903187    0.903218    0.903810  \n",
       "31   0.903218    0.903810    0.909454  \n",
       "32   0.903810    0.909454    0.907357  \n",
       "33   0.909454    0.907357    0.911044  \n",
       "34   0.907357    0.911044    0.910657  \n",
       "\n",
       "[5 rows x 132 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# drop columns we don't want to predict\n",
    "# We're only concerned with the estimating the close value,\n",
    "# Close should be first in the list of column in the config file\n",
    "\n",
    "cols_to_drop = []\n",
    "\n",
    "for i in range (n_out):\n",
    "    for j in range(1, n_features):\n",
    "        cols_to_drop.append(reframed.shape[1]-(i*n_features+j))\n",
    "\n",
    "reframed.drop(reframed.columns[cols_to_drop], axis=1, inplace=True)\n",
    "\n",
    "reframed.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "reframed_values = reframed.values\n",
    "# split into train and test sets\n",
    "training_size = int(0.8* reframed_values.shape[0])\n",
    "train = reframed_values[:training_size, :]\n",
    "test = reframed_values[training_size:, :]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "((23632, 30, 4), (23632, 12, 1), (5909, 30, 4), (5909, 12, 1))\n"
     ]
    }
   ],
   "source": [
    "# split into input and outputs\n",
    "n_obs = n_lag * n_features\n",
    "\n",
    "# We're only concerned with the estimating the close value,\n",
    "# Close should be first in the list of column in the config file\n",
    "\n",
    "n_outputs = n_out * n_features\n",
    "train_x, train_y = train[:, :n_obs], train[:, -n_out:]\n",
    "test_x, test_y = test[:, :n_obs], test[:, -n_out:]\n",
    "\n",
    "# reshape input to be 3D [samples, timesteps, features]\n",
    "train_x = train_x.reshape((train_x.shape[0], n_lag, n_features))\n",
    "test_x = test_x.reshape((test_x.shape[0], n_lag, n_features))\n",
    "\n",
    "# reshape output to be 3D [samples, timesteps, features]\n",
    "train_y = train_y.reshape(-1, n_out, 1)\n",
    "test_y = test_y.reshape(-1, n_out, 1)\n",
    "\n",
    "print(train_x.shape, train_y.shape, test_x.shape, test_y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_name=''.join([CONFIG['folder']['weights'], CONFIG['filename'], '_model', '.json'])\n",
    "model_weights_name=''.join([CONFIG['folder']['weights'], CONFIG['filename'], '_model_weights', '.h5'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python2.7/dist-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n",
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "bidirectional_1 (Bidirection (None, 100)               22000     \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 100)               0         \n",
      "_________________________________________________________________\n",
      "repeat_vector_1 (RepeatVecto (None, 12, 100)           0         \n",
      "_________________________________________________________________\n",
      "lstm_2 (LSTM)                (None, 12, 25)            12600     \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 12, 25)            0         \n",
      "_________________________________________________________________\n",
      "time_distributed_1 (TimeDist (None, 12, 1)             26        \n",
      "=================================================================\n",
      "Total params: 34,626\n",
      "Trainable params: 34,626\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import Input\n",
    "from keras.layers import LSTM, CuDNNLSTM, GRU,CuDNNGRU\n",
    "from keras.layers import Conv1D, AveragePooling1D, MaxPooling1D\n",
    "from keras.layers import Dropout, Flatten\n",
    "from keras.layers import Activation, BatchNormalization\n",
    "from keras.layers import TimeDistributed\n",
    "from keras.layers import Bidirectional\n",
    "from keras.layers import RepeatVector\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "\n",
    "units= CONFIG['lstm_hidden_size']\n",
    "dropout = .1\n",
    "\n",
    "# design network\n",
    "model = Sequential()\n",
    "model.add(Bidirectional(LSTM(units), input_shape=(train_x.shape[1], train_x.shape[2])))\n",
    "model.add(Dropout(dropout))\n",
    "\n",
    "model.add(RepeatVector(n_out))\n",
    "\n",
    "model.add(LSTM(units/2, return_sequences=True))\n",
    "model.add(Dropout(dropout))\n",
    "\n",
    "# We're only concerned with the estimating the close value,\n",
    "# otherwise use n_outputs instead of 1\n",
    "# Dense(n_outputs, ...\n",
    "model.add(TimeDistributed(Dense(1, activation='relu')))\n",
    "\n",
    "model.compile(loss='mse', optimizer='adam')\n",
    "\n",
    "# store model\n",
    "# serialize model to JSON\n",
    "model_json = model.to_json()\n",
    "with open(model_name, \"w\") as json_file:\n",
    "    json_file.write(model_json)\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "epochs=25\n",
    "batch_size=1024"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 23632 samples, validate on 5909 samples\n",
      "Epoch 1/25\n",
      "23632/23632 [==============================] - 3s 109us/step - loss: 0.1261 - val_loss: 0.0294\n",
      "\n",
      "Epoch 00001: val_loss improved from inf to 0.02943, saving model to weights/BTC_ETH_lstm_i30_o12_Close_Volume_Low_High_model_weights.h5\n",
      "Epoch 2/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0244 - val_loss: 0.0490\n",
      "\n",
      "Epoch 00002: val_loss did not improve from 0.02943\n",
      "Epoch 3/25\n",
      "23632/23632 [==============================] - 1s 50us/step - loss: 0.0460 - val_loss: 0.0171\n",
      "\n",
      "Epoch 00003: val_loss improved from 0.02943 to 0.01709, saving model to weights/BTC_ETH_lstm_i30_o12_Close_Volume_Low_High_model_weights.h5\n",
      "Epoch 4/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0376 - val_loss: 0.0115\n",
      "\n",
      "Epoch 00004: val_loss improved from 0.01709 to 0.01148, saving model to weights/BTC_ETH_lstm_i30_o12_Close_Volume_Low_High_model_weights.h5\n",
      "Epoch 5/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0152 - val_loss: 0.0101\n",
      "\n",
      "Epoch 00005: val_loss improved from 0.01148 to 0.01012, saving model to weights/BTC_ETH_lstm_i30_o12_Close_Volume_Low_High_model_weights.h5\n",
      "Epoch 6/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0095 - val_loss: 0.0074\n",
      "\n",
      "Epoch 00006: val_loss improved from 0.01012 to 0.00737, saving model to weights/BTC_ETH_lstm_i30_o12_Close_Volume_Low_High_model_weights.h5\n",
      "Epoch 7/25\n",
      "23632/23632 [==============================] - 1s 50us/step - loss: 0.0076 - val_loss: 0.0054\n",
      "\n",
      "Epoch 00007: val_loss improved from 0.00737 to 0.00537, saving model to weights/BTC_ETH_lstm_i30_o12_Close_Volume_Low_High_model_weights.h5\n",
      "Epoch 8/25\n",
      "23632/23632 [==============================] - 1s 50us/step - loss: 0.0064 - val_loss: 0.0065\n",
      "\n",
      "Epoch 00008: val_loss did not improve from 0.00537\n",
      "Epoch 9/25\n",
      "23632/23632 [==============================] - 1s 50us/step - loss: 0.0073 - val_loss: 0.0123\n",
      "\n",
      "Epoch 00009: val_loss did not improve from 0.00537\n",
      "Epoch 10/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0116 - val_loss: 0.0130\n",
      "\n",
      "Epoch 00010: val_loss did not improve from 0.00537\n",
      "Epoch 11/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0138 - val_loss: 0.0053\n",
      "\n",
      "Epoch 00011: val_loss improved from 0.00537 to 0.00531, saving model to weights/BTC_ETH_lstm_i30_o12_Close_Volume_Low_High_model_weights.h5\n",
      "Epoch 12/25\n",
      "23632/23632 [==============================] - 1s 50us/step - loss: 0.0087 - val_loss: 0.0030\n",
      "\n",
      "Epoch 00012: val_loss improved from 0.00531 to 0.00302, saving model to weights/BTC_ETH_lstm_i30_o12_Close_Volume_Low_High_model_weights.h5\n",
      "Epoch 13/25\n",
      "23632/23632 [==============================] - 1s 50us/step - loss: 0.0053 - val_loss: 0.0023\n",
      "\n",
      "Epoch 00013: val_loss improved from 0.00302 to 0.00225, saving model to weights/BTC_ETH_lstm_i30_o12_Close_Volume_Low_High_model_weights.h5\n",
      "Epoch 14/25\n",
      "23632/23632 [==============================] - 1s 50us/step - loss: 0.0043 - val_loss: 0.0023\n",
      "\n",
      "Epoch 00014: val_loss did not improve from 0.00225\n",
      "Epoch 15/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0042 - val_loss: 0.0034\n",
      "\n",
      "Epoch 00015: val_loss did not improve from 0.00225\n",
      "Epoch 16/25\n",
      "23632/23632 [==============================] - 1s 50us/step - loss: 0.0048 - val_loss: 0.0057\n",
      "\n",
      "Epoch 00016: val_loss did not improve from 0.00225\n",
      "Epoch 17/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0064 - val_loss: 0.0080\n",
      "\n",
      "Epoch 00017: val_loss did not improve from 0.00225\n",
      "Epoch 18/25\n",
      "23632/23632 [==============================] - 1s 50us/step - loss: 0.0082 - val_loss: 0.0068\n",
      "\n",
      "Epoch 00018: val_loss did not improve from 0.00225\n",
      "Epoch 19/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0076 - val_loss: 0.0043\n",
      "\n",
      "Epoch 00019: val_loss did not improve from 0.00225\n",
      "Epoch 20/25\n",
      "23632/23632 [==============================] - 1s 50us/step - loss: 0.0057 - val_loss: 0.0027\n",
      "\n",
      "Epoch 00020: val_loss did not improve from 0.00225\n",
      "Epoch 21/25\n",
      "23632/23632 [==============================] - 1s 50us/step - loss: 0.0043 - val_loss: 0.0022\n",
      "\n",
      "Epoch 00021: val_loss improved from 0.00225 to 0.00220, saving model to weights/BTC_ETH_lstm_i30_o12_Close_Volume_Low_High_model_weights.h5\n",
      "Epoch 22/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0039 - val_loss: 0.0025\n",
      "\n",
      "Epoch 00022: val_loss did not improve from 0.00220\n",
      "Epoch 23/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0040 - val_loss: 0.0029\n",
      "\n",
      "Epoch 00023: val_loss did not improve from 0.00220\n",
      "Epoch 24/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0042 - val_loss: 0.0037\n",
      "\n",
      "Epoch 00024: val_loss did not improve from 0.00220\n",
      "Epoch 25/25\n",
      "23632/23632 [==============================] - 1s 51us/step - loss: 0.0047 - val_loss: 0.0041\n",
      "\n",
      "Epoch 00025: val_loss did not improve from 0.00220\n"
     ]
    }
   ],
   "source": [
    "# fit network\n",
    "history = model.fit(train_x, train_y, epochs=epochs, batch_size=batch_size,\n",
    "                    validation_data=(test_x, test_y), verbose=1, shuffle=False,\n",
    "                    callbacks=[ModelCheckpoint(model_weights_name, monitor='val_loss', verbose=1,save_best_only='true',\n",
    "                                              save_weights_only=True)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5909/5909 [==============================] - 2s 371us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.002201281368157488"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load the best weights\n",
    "model.load_weights(model_weights_name)\n",
    "model.compile(loss='mse', optimizer='adam')\n",
    "model.evaluate(test_x, test_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib import pyplot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAD8CAYAAABw1c+bAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzt3Xl8XNWV4PHfKalKUpW2KkleJNmWjBdsbDAgbBLsNGk2sxoCIRBoIAnjzEwzQ3c6mTgzHdJhOjMhk5B0ZshCGqdZkhDaQHCCE2MCxAQItmyM90W2ZUu2bEvWYu1SVZ3545VkWZbkklRaqDrfz6c+9eptdZ/LOu++++47V1QVY4wxicM11gUwxhgzuizwG2NMgrHAb4wxCcYCvzHGJBgL/MYYk2As8BtjTIKxwG+MMQnGAr8xxiQYC/zGGJNgkse6AL3l5uZqUVHRWBfDGGM+UjZt2lSjqnnRrDvuAn9RURGlpaVjXQxjjPlIEZFD0a5rTT3GGJNgLPAbY0yCscBvjDEJZty18RtjzFB0dnZSWVlJW1vbWBdlRKWmplJYWIjb7R7yPizwG2PiQmVlJRkZGRQVFSEiY12cEaGqnDx5ksrKSoqLi4e8H2vqMcbEhba2NnJycuI26AOICDk5OcO+qrHAb4yJG/Ec9LvE4hjjJvCfauvkB6/vZUtF/VgXxRhjxrW4Cfwahh+8vo/S8tqxLooxJgHV19fzox/9aNDb3XDDDdTXj26FNW4Cf0ZqMkkuoa6lY6yLYoxJQP0F/mAwOOB2a9asITs7e6SK1ae46dXjcgl+r5va5s6xLooxJgGtWLGC/fv3s2DBAtxuN6mpqfj9fnbv3s3evXu59dZbqaiooK2tjYcffpjly5cDp9PUNDU1cf3117N48WLeffddCgoKeOWVV0hLS4t5WeMm8AP4vR7qmq3Gb0yi++Zvd7Dz6KmY7nNufibfuPmCfpd/+9vfZvv27WzZsoW33nqLG2+8ke3bt3d3u1y5ciWBQIDW1lYuu+wybr/9dnJycs7Yx759+/jVr37Fz372M+68805efPFF7r333pgeB8Rb4Pd5qLWmHmPMOLBw4cIz+tr/8Ic/5OWXXwagoqKCffv2nRX4i4uLWbBgAQCXXnop5eXlI1K2uAr8Aa+H/dVNY10MY8wYG6hmPlp8Pl/39FtvvcXrr7/Oe++9h9fr5corr+yzL35KSkr3dFJSEq2trSNStqhu7orIUhHZIyJlIrKij+WfEJHNIhIUkTt6zF8gIu+JyA4R2Soin4ll4Xvz+zx2c9cYMyYyMjJobGzsc1lDQwN+vx+v18vu3bv5y1/+MsqlO9M5a/wikgQ8AVwDVAIbRWS1qu7ssdph4AHgy702bwHuU9V9IpIPbBKRtao6In2XAj43dS2dhMOKyxX/D3IYY8aPnJwcrrjiCubNm0daWhoTJ07sXrZ06VJ+8pOfMGfOHGbPns3ll18+hiWNrqlnIVCmqgcAROR5YBnQHfhVtTyyLNxzQ1Xd22P6qIicAPKAEQn8fq+HUFhpbAuS5R16AiNjjBmKX/7yl33OT0lJ4fe//32fy7ra8XNzc9m+fXv3/C9/uXc9OnaiaeopACp6fK6MzBsUEVkIeID9g902WgGfB8Bu8BpjzABG5QEuEZkMPAt8TlXDfSxfLiKlIlJaXV095O/xdwV+69JpjDH9iibwHwGm9PhcGJkXFRHJBF4F/oeq9nlHQ1WfVNUSVS3Jy4tqrOA+BbxO4Le+/MYY079oAv9GYKaIFIuIB7gLWB3NziPrvww8o6qrhl7M6FhTjzHGnNs5A7+qBoGHgLXALuAFVd0hIo+KyC0AInKZiFQCnwZ+KiI7IpvfCXwCeEBEtkReC0bkSDjd1GM1fmOM6V9UD3Cp6hpgTa95j/SY3ojTBNR7u+eA54ZZxqj5PEl4klxW4zfGmAHETXZOcAYo8PvcVuM3xoy6oaZlBvjBD35AS0tLjEvUv7gK/OD05bcMncaY0fZRCvxxlasHnBu8lrbBGDPaeqZlvuaaa5gwYQIvvPAC7e3t3HbbbXzzm9+kubmZO++8k8rKSkKhEF//+tc5fvw4R48e5ZOf/CS5ubm8+eabI17WuAv8fp+HXTFOx2qM+Yj5/Qo4ti22+5w0H67/dr+Le6Zlfu2111i1ahUbNmxAVbnllltYv3491dXV5Ofn8+qrrwJODp+srCwef/xx3nzzTXJzc2Nb5n7EXVNPwGupmY0xY+u1117jtdde4+KLL+aSSy5h9+7d7Nu3j/nz57Nu3Tq++tWv8vbbb5OVlTUm5YvLGn9DayfBUJjkpLg7rxljojFAzXw0qCpf+9rX+OIXv3jWss2bN7NmzRr+8R//kauuuopHHnmkjz2MrLiLjAGvG1VoaLUbvMaY0dMzLfN1113HypUraWpyxgc5cuQIJ06c4OjRo3i9Xu69916+8pWvsHnz5rO2HQ1xWeMHqGvpICc95RxrG2NMbPRMy3z99dfz2c9+lo997GMApKen89xzz1FWVsZXvvIVXC4XbrebH//4xwAsX76cpUuXkp+fPyo3d0VVR/xLBqOkpERLS0uHvP3b+6r5m6c28MIXP8bC4kAMS2aMGc927drFnDlzxroYo6KvYxWRTapaEs32cdfU4/dahk5jjBlI3AX+QI+mHmOMMWeLu8BvNX5jEtd4a7oeCbE4xrgL/GmeJNLcSZavx5gEk5qaysmTJ+M6+KsqJ0+eJDU1dVj7ibtePeA099hDXMYklsLCQiorKxnOKH4fBampqRQWnpUMeVDiNvBbjd+YxOJ2uykuLh7rYnwkxF1TDzh9+Wtb7AEuY4zpS1wG/oDXcvIbY0x/4jLw+62pxxhj+hWXgT/g9dDYHqQjGB7rohhjzLgTl4G/K19PvfXsMcaYs8Rl4O96ete6dBpjzNniMvDb07vGGNO/uAz83fl6bNB1Y4w5S1SBX0SWisgeESkTkRV9LP+EiGwWkaCI3NFr2f0isi/yuj9WBR+I3+cGrKnHGGP6cs7ALyJJwBPA9cBc4G4RmdtrtcPAA8Ave20bAL4BLAIWAt8QEf/wiz2wrqYe69JpjDFni6bGvxAoU9UDqtoBPA8s67mCqpar6lagd//J64B1qlqrqnXAOmBpDMo9IHeSi4zUZGvjN8aYPkQT+AuAih6fKyPzojGcbYcl4PNYTn5jjOnDuLi5KyLLRaRUREpjlVnP7/VYjd8YY/oQTeA/Akzp8bkwMi8aUW2rqk+qaomqluTl5UW564FZjd8YY/oWTeDfCMwUkWIR8QB3Aauj3P9a4FoR8Udu6l4bmTfi/F6Pdec0xpg+nDPwq2oQeAgnYO8CXlDVHSLyqIjcAiAil4lIJfBp4KcisiOybS3wP3FOHhuBRyPzRlzA57amHmOM6UNUA7Go6hpgTa95j/SY3ojTjNPXtiuBlcMo45D4fR5aO0O0doRI8ySN9tcbY8y4NS5u7o6EQFdffmvnN8aYM8Rt4O/K0GnNPcYYc6a4Dfzd+Xqsxm+MMWeI28BvGTqNMaZvcRv4T2fotMBvjDE9xW3gz0pzIwK1LdaX3xhjeorbwJ/kErLT3FbjN8aYXuI28IPTs8dy8htjzJniOvAHvB6r8RtjTC9xHfj9PsvQaYwxvcV14A94LUOnMcb0FteB3+9zMnSq6lgXxRhjxo24DvwBn5uOUJjmjtBYF8UYY8aNuA78Nui6McacLa4Df8AStRljzFniOvB3Z+i0G7zGGNMtrgN/wJp6jDHmLHEd+C0nvzHGnC2uA39majJJLrG+/MYY00NcB34Rwe/1UNtsGTqNMaZLXAd+cPryWxu/McacFveB3++1DJ3GGNNT3Af+gM8ydBpjTE9RBX4RWSoie0SkTERW9LE8RUR+HVn+vogURea7ReRpEdkmIrtE5GuxLf65+X2WqM0YY3o6Z+AXkSTgCeB6YC5wt4jM7bXaF4A6VZ0BfB94LDL/00CKqs4HLgW+2HVSGC1Ohs5OwmFL1GaMMRBdjX8hUKaqB1S1A3geWNZrnWXA05HpVcBVIiKAAj4RSQbSgA7gVExKHiW/z0MorDS2BUfza40xZtyKJvAXABU9PldG5vW5jqoGgQYgB+ck0AxUAYeB76pq7TDLPCgBnxuwtA3GGNNlpG/uLgRCQD5QDPyDiEzvvZKILBeRUhEpra6ujmkBujJ02tO7xhjjiCbwHwGm9PhcGJnX5zqRZp0s4CTwWeAPqtqpqieAd4CS3l+gqk+qaomqluTl5Q3+KAbQlaHTevYYY4wjmsC/EZgpIsUi4gHuAlb3Wmc1cH9k+g7gDXWGvToM/DWAiPiAy4HdsSh4tLpr/NbUY4wxQBSBP9Jm/xCwFtgFvKCqO0TkURG5JbLaU0COiJQBXwK6unw+AaSLyA6cE8jPVXVrrA9iIFbjN8aYMyVHs5KqrgHW9Jr3SI/pNpyum723a+pr/mjyepLwJLusxm+MMRFx/+SuiDh9+a3Gb4wxQAIEfnD68luGTmOMcSRE4A/43Ja2wRhjIhIi8PutqccYY7olROAP+Cw1szHGdEmIwO/3emho7SQYCo91UYwxZswlROAP+DyoQkOr3eA1xpiECPz+roe4rLnHGGMSI/AHuhO1WY3fGGMSIvD7u1IzW88eY4xJjMAfsKYeY4zplhCB33LyG2PMaQkR+FPdSXg9SfYQlzHGkCCBH5xavz3EZYwxCRT4Az5L22CMMZBAgd/v81DbYt05jTEmYQJ/wOu2Gr8xxpBAgd9vTT3GGAMkUOAPeD00tgfpCFqiNmNMYkuYwN+Vr6feevYYYxJcwgT+rqd3rUunMSbRJUzgt6d3jTHGkTCBvztfj2XoNMYkuKgCv4gsFZE9IlImIiv6WJ4iIr+OLH9fRIp6LLtQRN4TkR0isk1EUmNX/Oh1Z+i0ph5jTII7Z+AXkSTgCeB6YC5wt4jM7bXaF4A6VZ0BfB94LLJtMvAc8B9V9QLgSmBMqtxdTT3WpdMYk+iiqfEvBMpU9YCqdgDPA8t6rbMMeDoyvQq4SkQEuBbYqqofAqjqSVUNxabog+NOcpGRmmxt/MaYhBdN4C8AKnp8rozM63MdVQ0CDUAOMAtQEVkrIptF5L8Nv8hDF/B5LCe/MSbhJY/C/hcDlwEtwB9FZJOq/rHnSiKyHFgOMHXq1BErjN/rsRq/MSbhRVPjPwJM6fG5MDKvz3Ui7fpZwEmcq4P1qlqjqi3AGuCS3l+gqk+qaomqluTl5Q3+KKJkNX5jjIku8G8EZopIsYh4gLuA1b3WWQ3cH5m+A3hDVRVYC8wXEW/khPBXwM7YFH3w/F6Pdec0xiS8czb1qGpQRB7CCeJJwEpV3SEijwKlqroaeAp4VkTKgFqckwOqWicij+OcPBRYo6qvjtCxnFPA57amHmNMwouqjV9V1+A00/Sc90iP6Tbg0/1s+xxOl84x5/d5aO0M0doRIs2TNNbFMcaYMZEwT+6Ck6ETsHZ+Y0xCS6jA35Wh05p7jDGJLKECf3e+HqvxG2MSWEIFfsvQaYwxCRb4T2fotMBvjElcCRX4s9LciEBti/XlN8YkroQK/EkuITvNbTV+Y0xCS6jAD07PHsvJb4xJZAkX+ANej9X4jTEJLeECv99nGTqNMYkt4QJ/wGsZOo0xiS3hAr/f52ToVFVoqoYNPwPVsS6WMcaMmoQL/AGfm45QmOaOEPzp27Dmy1C1ZayLZYwxoybhAn/X07v1dXXw4a+dmeXvjGGJjDFmdCVc4O96ele3r4KORnD74JAFfmNM4hjpMXfHna4MnVk7n4MJF0DhpbDzFQiHwGU5+o0x8S/xavxeD/PlAJm126Hkc1C0BNoa4Pj2qLZ/dWsVz7xX7twcNsaYj6CErPHfk/Q6na5U3BfeCe1NzoLyd2DyRQNu2xkK8/VXtlPb3MGp1k4e+uuZo1BiY4yJrYSr8WfSzC1J77Ez9zpIzYKsAvAXQ/mfz7ntn/ZUU9vcwQX5mXz3tb38/J2Do1BiY4yJrYQL/LLt3/FKO+szbz49s2ixc4M3HB5w25c+qCTH5+HF//Rxrp07kW/+dicvlFaMcImNMSa2Eivwq0LpSvYkzWCHTj89v2gJtNUP2M7f0NLJ6ztPcMuCfFLdSfzfz17Mkpm5rHhxK2u2VY1C4Y0xJjYSK/BXvA8ndvJW+o1nZugsusJ5H6Bb52+3HqUjFOb2SwoBSElO4qd/cykXT/Xz8PMf8OaeEyNZcmOMiZnECvylKyElk105156ZqC2rEPxFA7bzv7S5ktkTM7ggP7N7nteTzMoHLmPWxAz+47Ob+MuBkyNYeGOMiY2oAr+ILBWRPSJSJiIr+lieIiK/jix/X0SKei2fKiJNIvLl2BR7CFpqYcdv4MLP4M3IOjs18wDt/Adrmtl8uJ5PXVKAiJyxLCvNzTOfX0ihP40Hny5la2X9SB6FMcYM2zkDv4gkAU8A1wNzgbtFZG6v1b4A1KnqDOD7wGO9lj8O/H74xR2GLb+EUDuUfK47Q2c43KMvftESaK2DEzvP2vTlzZW4BG69uKDPXeekp/CLBy8n2+vmvpUb2HOscaSOwhhjhi2aGv9CoExVD6hqB/A8sKzXOsuApyPTq4CrJFI1FpFbgYPAjtgUeQgiN3WZsggmXoDf5yGscKqtx9i70yLt/L2ae8Jh5aUPjnDFjFwmZqb2+xWTslL5xYOL8CS5uPep9ymvaR6JIzHGmGGLJvAXAD37LFZG5vW5jqoGgQYgR0TSga8C3xx+UYfh4Hqo3Q8lnwecDJ3Ame382VMgexqUv33GphvKa6msa+2+qTuQaTk+nntwEcFQmHv+9X2qGlpjdwzGGBMjI31z95+A76tq00ArichyESkVkdLq6urYl6J0JaT5Ya5zodKVofOsAVmKlsChd89o539pcyU+TxLXXTApqq+aNTGDZz6/iIbWTu751/epaWqPzTEYY0yMRBP4jwBTenwujMzrcx0RSQaygJPAIuA7IlIO/B3w30Xkod5foKpPqmqJqpbk5eUN+iAG1HQCdv8OFtwD7jTgdIbO2ubOM9ctugJaa6F6FwCtHSHWbDvGDfMnk+aJPoHb/MIsVj5wGUfrW7nvqQ00tHaeeyNjjBkl0QT+jcBMESkWEQ9wF7C61zqrgfsj03cAb6hjiaoWqWoR8APgf6nq/4tR2aPzwbMQDsKlD3TP6q7x9+7Z06ud/7Wdx2hqD/KpKJp5eltYHOAn917KvhONfO7nG2huDw6p+MYYE2vnDPyRNvuHgLXALuAFVd0hIo+KyC2R1Z7CadMvA74EnNXlc0yEQ7Dp35wmnNzTCdW6a/y9m3r80yB7anc7/0ubj1CQncai4sCQvv7K2RP44V0X80FFPT/90/4h7cMYY2ItquycqroGWNNr3iM9ptuAT59jH/80hPINz/43oP4wXH3mvWWvJwlPsuvsGj/AtMWwby0nGlp4e181//nKGbhccvZ6Ubp+/mQWz8jl5S1H+PtrZp31HIAxxoy2+H5yt3Ql+PLg/JvOmC0iBLyeM3v1dClaDC0n+dM7bxNW+NQlfffdH4xlCwqoqG1l82F7uMsYM/biN/A3HIG9f4CL/waSPWct9vs8Z/fqASfwAye2/pGLp2YzPS992EW57oKJpCS7eGVL73vixhgz+uI38G9+xnlw69L7+1wc8Ln7rvH7p9GRXkhR8wdDuqnbl4xUN1fPmcjvtlbRGRo49bMxxoy0+Az8oSBsfhpmXOUkX+uD3+uhrqXvbpa7UubzMdcubp4fXd/9aCxbkE9tcwd/3lcTs30aY8xQxGfg3/sHaKzqflK3LwFf3238wVCY39QVE5BGspsPxKxIV86eQFaam99Yc48xZozFZ+Df9HPIyIeZ1/W7it/roaG1k2Cvppf1+6pZ1zrL+RDFcIzR8iS7uGH+JF7bcZyWDuvTb4wZO/EX+GsPQtkfnbb9pP57q3b15a/v9VTti5uP0JJWgGYWxDTwg9O7p7UzxLqdx2O6X2OMGYz4C/ybnwZxwSX3Dbia33f207sNrZ2s23mcWxYUIEVLnPz8qv3sYfAWFgXIz0rlNx9Yc48xZuzEV+APdsDmZ2HWUsjMH3DVgLcrX8/pwL9mWxUdwbDTd79oMTRXQ83emBXP5RJuXpDP+n01nLTkbcaYMRJfgX/3b6GlZsCbul38kdTMPfvyv7S5khkT0plfkHV6HN5eaZqHa9lFBYTCagO0G2PGTHwF/tKfO7l2zvvrc67aO0PnoZPNbCyv4/ZLCp20Cv5iGIF2/jmTM5g1MZ3fbDka0/0aY0y04ifwn9zv1M4v/Ry4zn1YvXPyv7T5CCJw68WRJiIRp7mn/M8xbecXEZYtKGDToToqaltitl9jjIlW/AT+wHT4/GvnvKnbJdWdhNeTRG1zB6rKSx9UcsV5uUzOSju9Unc7/76YFnXZAufkYikcjDFjIX4CvwhMXQS+3Kg38Xs91DV3UHqojora1rMTsk0bmXb+Qr+Xy4r8/GbLUTSGVxPGGBON+An8QxDweaht6eClzZV4+xpeMTDdeRAsxu38ALcsKKDsRBM7q07FfN/GGDOQhA78fp+Hqvo2fre1iqXzJuFL6fXAV1c7f4z78wPcOH8yyS7hFbvJa4wZZQkd+ANeN3uON9LYFuSO/jJxFl0BTcfhZFlsv9vn4a9m5bF6y1FCYWvuMcaMnoQO/F1P7+ZnpXL59Jy+Vypa4rzHuJ0fYNnFBRw71caGg7Ux37cxxvQnoQN/19O7t15c0P/wioHpkDEZyt+J+fdfPWcCXk+S9e4xxoyqhA78Bf40XMLAA66IOL17YtyfH8DrSea6CyaxZlsV7cFQTPdtjDH9SejAf/NF+bzxD1cyY8I5hlcsWgxNx5yHxGJs2YJ8TrUFeXN3dcz3bYwxfUnowO9OclGU6zv3il3t/Idi361z8YxcctM9rP7QmnuMMaMjoQN/1HLOg/RJI9KfPznJxU0X5vP6rhOcaut7KEhjjImlqAK/iCwVkT0iUiYiK/pYniIiv44sf19EiiLzrxGRTSKyLfJ+7uxp45GI061zBNr5AW5ZkE9HMMwfth+L+b6NMaa3cwZ+EUkCngCuB+YCd4vI3F6rfQGoU9UZwPeBxyLza4CbVXU+cD/wbKwKPuqKFjvj+NbGbhzeLhdPyWZajtd69xhjRkU0Nf6FQJmqHlDVDuB5YFmvdZYBT0emVwFXiYio6geq2vVo6g4gTURSYlHwUdfdnz/2zT0iwrKL8nl3/0lOnGqL+f6NMaanaAJ/AVDR43NlZF6f66hqEGgAej8RdTuwWVU/mkNP5cwA34QRCfzg5O5RhdUfWgoHY8zIGpWbuyJyAU7zzxf7Wb5cREpFpLS6epx2axyh/PxdZkxIZ15BpuXuMcaMuGgC/xFgSo/PhZF5fa4jIslAFnAy8rkQeBm4T1X77Aivqk+qaomqluTl5Q3uCEZT0WJoPAp1B0dk97cuKGDbkQb2VzeNyP6NMQaiC/wbgZkiUiwiHuAuYHWvdVbj3LwFuAN4Q1VVRLKBV4EVqhr7nAejrWix8z5CzT03X5SPCFbrN8aMqHMG/kib/UPAWmAX8IKq7hCRR0XklshqTwE5IlIGfAno6vL5EDADeEREtkReE2J+FKMldxb48mDv2hFp7pmYmcrHz8vhlS1HbIAWY8yIiaqNX1XXqOosVT1PVb8VmfeIqq6OTLep6qdVdYaqLlTVA5H5/6yqPlVd0ON1YuQOZ4SJwIWfgd2/g198GhqPx/wrll1UwKGTLWypqI/5vo0xBuzJ3cG79p/hhu86aZp//DHY8/uY7n7p/El4kl3W3GOMGTEW+AdLBBb+B1j+J8jMh1/dBb/9O+hojsnuM1PdXHX+BH639SgtHcGY7NMYY3qywD9UE86HB9+AKx6GTf8GP/0EHNkck11/dtFUapo6uOn//pntRxpisk9jjOligX84kj1wzaNw/2+hsw2eugbW/x8IDy+3/pKZefziwUU0twe57Ufv8OT6/YRteMZ+VTW08va+ajpD4bEuijEfCTLeeo+UlJRoaWnpWBdj8Frr4NV/gO0vwpTL4VM/BX/RsHZZ19zBipe2snbHcRbPyOV7d17ExMzU2JR3vFKF9lOQmtXvKs3tQd4/eJK399Xw9r4ayk44zz3MmZzJd26/kPmF/W9rTLwSkU2qWhLVuhb4Y2zrC84JQBVu/K7TC0j6GdYxCqrK8xsrePS3O0lxu3js9gu57oJJMSzwKAkFofkENB5zBq9vOu70imo61uv9OIQ7YfaNcMdKcKcSDivbjzZEAn01mw7V0RlSUpJdLJqew5IZuQR8Hh77w25ONnfw4JJi/v7qWaS6k8b6qI0ZNRb4x1r9YXjpi3D4XbjgNrjp+5DmH9Yu91c38fDzH7D9yCnuXjiVr980B68nOUYFHiEdLbDuEdj5G2iuAfr4v5YWgIxJkD7x9HuoE/7yBFW5H+c72V/nrQNN1LU4YxXMnZzJklm5LJmRR0mR/4zg3tDaybd/v4tfbaigONfH//7UfC6f3jtllDHxyQL/eBAOwTv/Am9+ywlmCz4LE+ZA3hwn4VuyZ9C77AiG+d66PTy5/gDFuT5+eNfFzCsYp80ax3fCqs9D9S6Yd7tzzN3BfRJkTHSS3vX4d1BV3th9gsfX7WXu8dU8lvwzNrsu4MXZ32XR7KlcMSOXvIxzJ3d9t6yGFS9t43BtC/csmsqK688nI9U9kkdrzJizwD+eHP0AfvclqPoQNHLT15UMOTOdnkET5jonhAlznXsCrnM3T7xbVsPfv7CF2uYOvnLdbB5cPB2Xa+jNSTGlCqVPwdr/ASmZzr2O8849/s4Hh+v437/fzYaDtRTleLln0TRuYj2T3vx7pHAh3PPvkJoZdTFaO0I8vm4PT/35IBMzU/nWbfP46/MnDufIjBnXLPCPR8F2qNkHJ3bBiZ3Oe/UuqCs/vU5yqpMWYsJcmDQPzr8JAsV97q7njd8rZuTwvU8vYFLWGN/4bamF1f/FebL5vKvgtp9A+sAZOg7WNPN/1u5mzbZj5Pg8PHz1TO5eOBV3UqTD2Y6X4cUHYfJFcO+Lg24y21JRz1dXbWXP8UaWLcjnkZvmkpP+0RwSwpiBWOD/KGlvgpo9kRNCj1dpX6/4AAAP40lEQVRj5Mndgkth3h3OvYLMyWds2vvG71eum80nZuZR6E9DhnFDeUgOvQsv/gfn5uzV34DL/xZc/fcWrm5s54d/3MevNhzGk+ziwSXTWf6J6aSn9HHfYver8ML9zpXRfa+ANzCoonUEw/zorTKeeLOMjFQ337h5LrdclD/6/0bGjCAL/PGg/jBsfwm2r4Jj24DIeADzboe5y84Ifvurm/i757ewLfKw16TMVEqK/CwsDlAyLcDsSRkkjVRTUDjkPLvwp8ecpqrbn4KCS/pdvbk9yL++fZAn1++nLRjmrsum8PDVM5mQcY6rlX3r4Pl7nHsF970C6YNP3733eCP/bdVWtlTUc9X5E/jn2+YxOStt0PsxZjyywB9vqvc6zwdsXwUny5x7BOddBfPvgNnXQ0oG4bCy53gjpeW1bCivY+PBWo5FhnHMSE3m0ml+LisKcFlRgAsLs2LT1bGhEl5aDofecbqt3vg9SMnoc9XOUJjnN1bwL6/vo6apnevnTeLL183mvLz06L9v/5vwq7shewrct/qsK6BohMLKv71bznfX7iGkyi0X5fPAx4vG703ycaClI8iWinoqa1vJ8roJ+Dz4vR4CPg9Zae6Rq1SYQbHAH69U4dhW2LbKuRo4VQnJaTDrOuckMOMacKdGVlUq61opPVTLhoN1lJbXsi/yoJMnycWFhVmUFAWYMzmD4lwfRbk+MgfT82X3q/DK30KwA256HC66q58iK2t3HOM7f9jDgZpmFhYFWHHD+VwydYjdW8vfgV/e6dw7uP+3kFU4pN1U1Lbw0/X7eWnzEVo6Qlw6zc8DHy9i6bxJp+8vxBNVqNnrXDmVrYPDfwFvLuSc51xFdb/O46hMYFNFI5sO1bHpUB07q04R6ufJcRHITnPj93kIeD1nvvvcFOemUzLNj983+F5sZnAs8CeCcBgq3neuBHa8DC01zs3hKQudgeGLFjv3B5JP38isbe5g06E6NpbXsrG8lm2VDQR7/EHnpnuck0COcyKYHjkhFOX4SPNErhA6W+G1r8PGn8Hki2i6+Umqkgo42tDGsYZWjta3UdXQSlVDm/Oqb6W5I8TMCel8den5XDVnwvDb1is2wHO3Q1q2E/yH8YR0Q2sn/15awbN/OcShky1MzEzhnkXTuHvh1Ki6jo5r7U1w8E+RYP9HaDjszM+bA8VLoK0BrSkjXLOPpI5T3Zt1aBIVOoFDkk9zejGeibOYUDyXvOKLqJcs6lo6qG3uoK65g9qWzsh75HNzR/fyztDp/1szJ6RzWXGAhUUBSor8FPq9o/2vEfcs8CeaUND5Ay/7I5Svh2PbAe11IljitL33OBG0dYY4dLKFgzXNHKxppjzyfvBkM9WN7YCSRz3T5DgX+eq4IO0kH+t4j8ntB1nlWca32j9DXfuZRRGBvPQUJmenMTkzlcnZqVxYmMXNF+aTHMua9JHN8Oxt4EmH+1c7NddhCIeVt/ae4N/ePcT6vdV4klzceOFk7v94EQumZMeo0CNMFar3ODX6feucG+7hTuffaPqV6IyrOZZ3BR82ZrL9SAObDtWxpaKe1s4gfhq5LKOWJYEGLkyrpogqMlsOIbUHINh2+jsKSuD8G5wnq/Nm9/tUuqrS1B5kV1Vjd0VjU3kdje1Oxtn8rFQuKw50Nz/OnJA+frokf0RZ4E90LbVw+D1niMjyt3ucCNLOviIQcW4k1x2E2oNO99Lag4RqDyB15biCrd27DeGiUibz8/TlHM1bTH52GpOyUpmclepMZ6YyMTMVT/IoNZVUbYVnbwWX2wn+ebNjstv91U088245qzZV0twR4qIp2Tzw8WncMH8yKcnjLA3EGbX616GhAgDNm0ND4V+x07eIt9tnsK2qle1HG6iPPAGd5BLmTM6gZFqAS6b5KZnmJz+7jxvd4TCcOuLcWzpSCrvXwNFIFtrAdJh9A5x/I0xZdM5nUEJhZfexU2w8WMvGQ859qBONTs0h2+umZJqfkqIA8/KzmD0p46N/xTXKLPCbM7XUOrW/8j87r+PbnPlJKRAOnn6wDJyrBH8R+IudP+xAcWS6GLKmDOmJ4xF1Yhc8fYtzDJfcBzOvhcKFkDT8dBaNbZ28uKmSZ947xIGaZnLTU7hiRg752WkURF752WnkZ6eO3pPBPdvq973mnOBDHYTdPqpyLmez51LWtM7n7RMpNEVq154kF7MmpTMvP4sLCrKYl5/J+ZMyTzffDdapo7BnjXMSOLjeuarw5sCspc5JYPonwXPuphxV5XBtCxsjnRE2ltdyoOb0uBY5Pg+zJ2U4r4nO+6yJGfj66vJrLPCbc+g6ERx+zwn0PYN7+qQB+9+PSzX74NUvRZo2gpCSBed9EmZeAzOudtJEDEM4rLxdVsMv/nKIXcdOUVXfdsa9EYDM1OTuE0J+5FXgd66CUpJdJLkElwhJLiHJRff06Xmnp4PhMPUtnd3t6KdONeCrepdJx9+muP5dcjqPAXBAprJeL2Jtx4WUhmfTSTKpbhdzJ2cyryArEugzmTkhY+SuwtpOOVcae9bA3tegvcG5sjzvk87VwPQrnRvwUd7XqW3uYHfVKXYfa2TPsUZ2H29k77FGWjtPV06mBrzMnpTB+ZGTwvTcdDLTkklPScaXkjzmN+dVlY5QmLbOMI1tnZxqDXKqrZPGtiCnWjs5FZnX2HZ6umt5ca6PH9598ZC+1wK/SUxtDXDgrUhteJ2T8RNg0oXOlcDMa6GwJKq0GAMJhZXqxnaO1LdyNPI60v3extH6VhpaO4f1HUVSxSddW7jS9SGXu3aRIp00awqbky7kw7SFlGV+jHBmIQGfh2yvm2k5XublZzE9L33suleGOp2uvbvXOL2+TlU68705MHkB5C84/Z41JeqTQTjsXBl0nQz2HHdODOU1zfTV2Sgl2UVGqnMS6DoZpPeYzkhNJtklhMJKKKwEw0pYtftz96vXvI5QmI5gmPag8+5Mh5zpUJj2zjDtoTDuYDMTpY5cGvBJG+m04pM2fLSSTlv3vMykdjJdzis9srwx4zxmPbx6SP/8FviNUYXj253mkH3rnJ5AGoLUbCd30MxrYdrHIWPyiDRfNbZ1UtXQxrGGNjpDYULdwQVCqoTDSigUJqmjDm9jOb6mw6Q3HyK95TB5DTvIbHXa6tuzZ9A5/So851+HZ/riM27Oj2uqzoOHFe/D0S1QtcVplutqVkwLnHkimLwAsqcOKoV5W2eIshNNHDrZQnN7kMb2IM3tQZq6Xm3BM+Z3LWtsCxIKa/eVVpIISUmRd1cfr8j8zKR2JlDPRFcdeVpLjtYSCNfhD9eSHawhK3iSjM4aPOHWAcutyWmQko540iElHTwZ4PE503nnw5UrhvRPboHfmN5a65wHwMped04EzSdOL0sLRDKHTjydOTR9kvOsQM9sov08nDYgVee7T+6H2v1Qe+DM6bYeQ2uKy6kJT5gLM65ymqqGOZjPuNLZCsd3OIkLq7bA0Q+dfFXhyNjSaX4nJ1P6JGc6Lds5UZ/1HlnW30lQ1bn6CLY6I+P19d7RAu2N0NHkDPzT3uR87m9eeyN09jGudnKa838kY3LkfdLpz748J1FhSrrTsyolHdy+mNx/6osFfmMGEg47D8JVbTk9+EvT8TMHiQl1nL2d29f9gFy3s/5+en0OdTqBpEtXcA9Md7qgBqZD4DxnOnvqR6dGHyudbXBix+mrgmPbnLEbWuuho3HgbZPTnBOA2+skQewZ2HWQw3C6kp0Te0qGE6w96ZHpyLsnw0kT0h3gI+8pmcMaaCmWBhP4ozr1iMhS4F+AJOBfVfXbvZanAM8AlwIngc+oanlk2deALwAh4L+q6tooj8OYkeFyOc0L+Qv6Xt5VS+99Mmg8DqH2Pjbo9YffMxBIkpNiIhAJ8v5piRfcB+JOdboVF1x69rJQ0Lkiaqt3TgRtdc57a12PefVO7d2d5nRU6H5PdU4M/b17vGcG+eSUcRPAR8M5A7+IJAFPANcAlcBGEVmtqjt7rPYFoE5VZ4jIXcBjwGdEZC5wF3ABkA+8LiKzVHV4o5EbM5JEnCR43oCTEdSMjaRk8OU4LxNT0fR7WgiUqeoBVe0AngeW9VpnGfB0ZHoVcJU4z+UvA55X1XZVPQiURfZnjDFmjEQT+AuAih6fKyPz+lxHVYNAA5AT5baIyHIRKRWR0urq6uhLb4wxZtDGxZM6qvqkqpaoakle3uDzrBtjjIleNIH/CDClx+fCyLw+1xGRZCAL5yZvNNsaY4wZRdEE/o3ATBEpFhEPzs3a3o+WrQbuj0zfAbyhTj/R1cBdIpIiIsXATGBDbIpujDFmKM7Zq0dVgyLyELAWpzvnSlXdISKPAqWquhp4CnhWRMqAWpyTA5H1XgB2AkHgb61HjzHGjC17gMsYY+LAYB7gGhc3d40xxoyecVfjF5Fq4NAwdpEL1MSoOB81duyJK5GPP5GPHU4f/zRVjapb5LgL/MMlIqXRXu7EGzv2xDx2SOzjT+Rjh6EdvzX1GGNMgrHAb4wxCSYeA/+TY12AMWTHnrgS+fgT+dhhCMcfd238xhhjBhaPNX5jjDEDiJvALyJLRWSPiJSJyNAGrfwIE5FyEdkmIltEJK6fgBORlSJyQkS295gXEJF1IrIv8u4fyzKOpH6O/59E5Ejk998iIjeMZRlHiohMEZE3RWSniOwQkYcj8+P+9x/g2Af928dFU09ksJi99BgsBri712AxcU1EyoESVY37/swi8gmgCXhGVedF5n0HqFXVb0dO/H5V/epYlnOk9HP8/wQ0qep3x7JsI01EJgOTVXWziGQAm4BbgQeI899/gGO/k0H+9vFS449msBgTJ1R1PU5OqJ56Dgb0NM4fRFzq5/gTgqpWqermyHQjsAtnjI+4//0HOPZBi5fAH9WAL3FOgddEZJOILB/rwoyBiapaFZk+Bkwcy8KMkYdEZGukKSjumjp6E5Ei4GLgfRLs9+917DDI3z5eAr+Bxap6CXA98LeR5oCEFEkJ/tFvwxycHwPnAQuAKuB7Y1uckSUi6cCLwN+p6qmey+L99+/j2Af928dL4E/4AV9U9Ujk/QTwMok3tvHxSBtoV1voiTEuz6hS1eOqGlLVMPAz4vj3FxE3TuD7haq+FJmdEL9/X8c+lN8+XgJ/NIPFxC0R8UVu9iAiPuBaYPvAW8WdnoMB3Q+8MoZlGXVdQS/iNuL09xcRwRn/Y5eqPt5jUdz//v0d+1B++7jo1QMQ6cL0A04PFvOtMS7SqBGR6Ti1fHAG1/llPB+/iPwKuBInK+Fx4BvAb4AXgKk42V3vVNW4vAHaz/FfiXOpr0A58MUebd5xQ0QWA28D24BwZPZ/x2nrjuvff4Bjv5tB/vZxE/iNMcZEJ16aeowxxkTJAr8xxiQYC/zGGJNgLPAbY0yCscBvjDEJxgK/McYkGAv8xhiTYCzwG2NMgvn/P9KWbMFlLYsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot history\n",
    "pyplot.plot(history.history['loss'], label='train')\n",
    "pyplot.plot(history.history['val_loss'], label='test')\n",
    "pyplot.legend()\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Test the prediction of test data\n",
    "y = model.predict(test_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = test_y[:,0]\n",
    "b = y[:,0]\n",
    "c = np.append(b, y[-1], axis=0)\n",
    "\n",
    "# Show how the model fits the test data\n",
    "pyplot.plot(a[:100], label='original')\n",
    "pyplot.plot(b[:100], label='model')\n",
    "pyplot.legend()\n",
    "pyplot.show()\n",
    "\n",
    "# Show how the model predicts data\n",
    "pos = int(a.shape[0]-n_out*4)\n",
    "pyplot.plot(a[pos:], label='original')\n",
    "pyplot.plot(c[pos:], label='model')\n",
    "pyplot.legend()\n",
    "pyplot.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Prediction on public data!\n",
    "period = CONFIG['period']\n",
    "import time\n",
    "import urllib2\n",
    "import json\n",
    "\n",
    "# Download a live bitcoin price data set\n",
    "def dl_X(now = None, points = n_lag, period = period, pair=CONFIG['pair']):\n",
    "    if now == None:\n",
    "        now = time.time() \n",
    "    end = now - now % period\n",
    "    #print end, time.strftime(\"%a, %d %b %Y %H:%M:%S +0000\", time.gmtime(end))\n",
    "    start = end - points*period\n",
    "    #print start, time.strftime(\"%a, %d %b %Y %H:%M:%S +0000\", time.gmtime(start))\n",
    "    url = \"https://poloniex.com/public?command=returnChartData&currencyPair=%s&start=%d&end=%d&period=%d\" % (pair, start, end, period)\n",
    "    openUrl = urllib2.urlopen(url)\n",
    "    r = openUrl.read()\n",
    "    openUrl.close()\n",
    "    d = json.loads(r.decode())[-n_lag:]\n",
    "    df = pd.DataFrame(d)\n",
    "    original_columns=[u'close', u'date', u'high', u'low', u'open',u'volume']\n",
    "    new_columns = ['Close','Timestamp','High','Low','Open','Volume']\n",
    "    df = df.loc[:,original_columns]\n",
    "    df.columns = new_columns\n",
    "    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict(when=None):\n",
    "    rt_df = dl_X(when)\n",
    "    rt_values = rt_df.loc[:,columns].values\n",
    "    rt_scaled = scaler.transform(rt_values)\n",
    "    rt_x = rt_scaled.reshape((1, n_lag, n_features))\n",
    "    print rt_x.shape\n",
    "    return rt_scaled, model.predict(rt_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 30, 4)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 30, 4)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 30, 4)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 30, 4)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 30, 4)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 30, 4)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAD8CAYAAAB3u9PLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzt3Xl8VPW5+PHPk8m+r2wJIWEnrAKyFBG1RdFat6rVutS2bu31V9teu3jbW29te2+9t9faBWut1dpeV2zrUqGKFdmlArIGAgEChOwJ2ffM9/fHOROGkGWSzDDJzPN+veaVzJlzznznQOY53+35ijEGpZRSKsTfBVBKKTU0aEBQSikFaEBQSill04CglFIK0ICglFLKpgFBKaUUoAFBKaWUTQOCUkopQAOCUkopW6gnO4nICuAXgAN4xhjz0272uRn4D8AAu40xn7e3fwH4vr3bj40xz9vbPwBGA032a5cbY8p6K0dqaqrJysrypMhKKaVsO3bsqDDGpPW1X58BQUQcwEpgOVAIfCQibxpjct32mQQ8DCwxxpwWkRH29mTgEWA+VqDYYR972j70NmPMdk8/VFZWFtu3e7y7UkopQESOe7KfJ01GC4B8Y8xRY0wr8DJwbZd97gFWur7o3e70rwDWGmOq7NfWAis8KZhSSqnzy5OAkA6cdHteaG9zNxmYLCKbReRDu4nJk2OfE5FdIvLvIiL9LLtSSikv8lancigwCbgEuBX4nYgk9nHMbcaYmcBS+3FHdzuJyL0isl1EtpeXl3upuEoppbrypFP5FDDW7XmGvc1dIbDNGNMGHBORQ1gB4hRWkHA/9gMAY8wp+2ediLyI1TT1x65vbox5GngaYP78+ZqrW6kA0NbWRmFhIc3Nzf4uSkCJjIwkIyODsLCwAR3vSUD4CJgkItlYX/C3AJ/vss/rWDWD50QkFasJ6ShwBPhPEUmy97sceFhEQoFEY0yFiIQBVwPvDegTKKWGncLCQuLi4sjKykJbi73DGENlZSWFhYVkZ2cP6Bx9BgRjTLuIPAC8gzXs9FljzH4ReRTYbox5037tchHJBTqAbxljKgFE5EdYQQXgUWNMlYjEAO/YwcCBFQx+N6BPoJQadpqbmzUYeJmIkJKSwmCa1j2ah2CMWQ2s7rLtB26/G+Cb9qPrsc8Cz3bZ1gDMG0B5lVIBQoOB9w32mupMZeVzTqfh5X+eoLmto9/HNrS0s2r7SXSpV6V8TwOC8rkPj1by3b/sZc2+4n4f+5edhXzrtT3sLqzxQcmU8p8PPviALVu2+LsYZ9GAoHxuf1EtALn2z4Ecu7ew2qtlUmqg2tvbz3pujMHpdPb7PBoQVFA6UGx9qe8fQEDItY/dozUE5QN//OMfmTVrFrNnz+aOO+7grrvu4rXXXut8PTY2FrC+vJcuXco111xDTk4OBQUFTJkyhTvvvJMZM2Zw8uRJ3n33XRYvXszcuXO56aabqK+vB6yUO4888ghz585l5syZHDx4kIKCAp566il+/vOfM2fOHDZu3OiXz9+VR53KSg2G60s9t7gWY4zHHV9tHU4OltQBsPeUBoRA9cO39g+o9tibnDHxPPKZ6b3us3//fn784x+zZcsWUlNTqaqq4pvfPGdcTKedO3eyb98+srOzKSgo4PDhwzz//PMsWrSIiooKfvzjH/Pee+8RExPDY489xuOPP84PfmCNvUlNTWXnzp08+eST/OxnP+OZZ57h/vvvJzY2loceesirn30wNCAon2ptd3KkvJ7kmHCqGlopqmkmPTHKo2OPljfQ2u5kXEo0h0rraGrtICrc4eMSq2Dx/vvvc9NNN5GamgpAcnJyr/svWLDgrPH948aNY9GiRQB8+OGH5ObmsmTJEgBaW1tZvHhx57433HADAPPmzeMvf/mLVz+HN2lAUD51uKyOtg7DNbPH8IctBeQW1XocEPYXWbWCz104lv/+ex65xTXMG9f7H60afvq6kz+fQkNDO/sDnE4nra2tna/FxMScta/7c2MMy5cv56WXXur2vBEREQA4HI5z+iCGEu1DUD51oNhq8rlhbjoi/etYzi2qJSI0hOvmWPkQd5/UZiPlPZdddhmrVq2isrISgKqqKrKystixYwcAb775Jm1tbR6da9GiRWzevJn8/HwAGhoaOHToUK/HxMXFUVdXN4hP4H0aEJRPHSiuJTIshOljEshOjem86/fE/qJapo6OZ0xiFCPjI7QfQXnV9OnT+d73vseyZcuYPXs23/zmN7nnnntYv349s2fPZuvWrefUCnqSlpbGH/7wB2699VZmzZrF4sWLOXjwYK/HfOYzn+Gvf/3rkOpUluE04Wf+/PlGF8gZXj7/uw9paGnnjQcu4oEXd7LrZDWbvnNZn8cZY5jz6Fqumjma/7phJnc/v51jFfX8418v8X2hlc8dOHCAadOm+bsYAam7aysiO4wx8/s6VmsIymeMMeQW1zJtdDwA08ckUHi6iZrGvqvhRTXN1DS1MX2MdeysjASOVjRQ1+xZFV4p1X8aEJTPlNQ2U93YRo79pe766RqG2pv9dvOQ65iZGQkYM7C5DEopz2hAUD7jmpDmqiHkjPY8IOQW1yICU0fFATAzPQGAvTpBTSmf0YCgfMY1wsj1pZ4WF8GIuAiPOpb3F9UyPjWG6HBrZHRqbATpiVHs0Y5lpXxGA4LymdyiWsYmRxEXeWb1ppwx8R4NPc0tqiVnTMJZ22ZlJLBHcxop5TMaENRZ2jv6n6SrJweKazubiVymj4knv6yelvaeU2FXN7Zyqrqps0PZZWZGAscrGz3qlFZK9Z8GBAVYI4J++NZ+Fv3XPyipGfw6t42t7RyrbOjsP3DJGZ1Au9NwuLS+x2NdfQxdg8ms9ERA8xqpocmVCK+oqIgbb7yx132feOIJGhsbO59fddVVVFf7v/arAUEB8NsNR3lucwEV9a386O3cQZ8vr6QOYzg3INh3/b31I7ialHK61hDsjuU9p/z/h6OCQ0dH/xd1GjNmzFkZU7vTNSCsXr2axMTEfr+Xt2lAULyx6xQ/XXOQz8wew9c/NYm39xSz6XDFoM7Z013+uORoYsIdvfYj5BbVMjI+gtTYiLO2J0SHMS4lWkcaKa8oKChg6tSp3HbbbUybNo0bb7yRxsZGsrKy+M53vsPcuXNZtWoVR44cYcWKFcybN4+lS5d2zkA+duwYixcvZubMmXz/+98/67wzZswArIDy0EMPMWPGDGbNmsWvfvUrfvnLX1JUVMSll17KpZdeClgpsisqrL+5xx9/nBkzZjBjxgyeeOKJznNOmzaNe+65h+nTp3P55ZfT1NTk9Wuiye2C3JYjFTy0ajcLs5P52U2zMAZe//gUP3hjH2u+vpSI0IFlFz1QXEtcRCgZSWcnsgsJEaaNju916GluN30PLjPTE/j4hNYQAsqa70LJXu+ec9RMuPKnfe6Wl5fH73//e5YsWcKXvvQlnnzySQBSUlLYuXMnAJ/85Cd56qmnmDRpEtu2beOrX/0q77//Pg8++CBf+cpXuPPOO1m5cmW353/66acpKChg165dhIaGUlVVRXJyMo8//jjr1q3rzLTqsmPHDp577jm2bduGMYaFCxeybNkykpKSOHz4MC+99BK/+93vuPnmm/nzn//M7bffPsgLdTatIQSxgyW13PfHHWSnxvD0nfOJCHUQGebgh9fO4GhFA7/bcHTA5z5QXMe00fHdrn0w3R5p5HSemzalua2Dw2X1TO8ywshldkYip6qbqKxvGXDZlHIZO3ZsZ8rq22+/nU2bNgHwuc99DoD6+nq2bNnCTTfdxJw5c7jvvvsoLraWgt28eTO33norAHfccUe353/vvfe47777CA217r37SrG9adMmrr/+emJiYoiNjeWGG27ozHOUnZ3NnDlzACuNdkFBwSA+efe0hhCkimua+OJzHxEd4eC5Ly4gIerM0NBlk9O4csYofr0un2vnpDM2Obpf53Y6DQeLa7lxXka3r+eMiadhawcnqhrJSj07edjh0no6nOac/gOXmRmufoQaLp0yol/lUkOUB3fyvtL1hsX13JXUzul0kpiYyK5duzw63pdcKbTBSqPtiyYjrSEEodrmNr743EfUNbfz3F0Lul2f4N+vziFEhB++1f8O5hNVjTS0dpzToeySM9r6Uu8uDYWrs7mnJqPpY+IR0RnLyjtOnDjB1q1bAXjxxRe56KKLzno9Pj6e7OxsVq1aBVij8Xbv3g3AkiVLePnllwF44YUXuj3/8uXL+e1vf9u5BkJVVRXQc+rrpUuX8vrrr9PY2EhDQwN//etfWbp0qRc+qWc0IASZ1nYnX/m/HeSX1fPU7fN6vBMfkxjFg5+cxHsHSnkvt7Rf79E1ZUVXk0bGEhoi5Baf+6WeW1xLbEQomT3USuIiwxifGqNrLCuvmDJlCitXrmTatGmcPn2ar3zlK+fs88ILL/D73/+e2bNnM336dN544w0AfvGLX7By5UpmzpzJqVOnuj3/3XffTWZmZue6zS+++CIA9957LytWrOjsVHaZO3cud911FwsWLGDhwoXcfffdXHDBBV7+1D3T9NdBpLmtg4dW7eZve4p5/ObZ3DC3+yYdl7YOJ1f9YiNNbR2s/cYyj5evfPzdPH69Lp/cR1cQGdb9MSue2MCohEj+8MUFZ22/8TdbEIFV93+ix/N/45VdbDlSwbZ/+5RH5VFDz1BIf11QUMDVV1/Nvn37/FoOb9P016pPxyoauP7JLfxtTzEPXzm1z2AAEOYI4UfXzaDwdBNPfpDv8XvlFtcxPi22x2AA3aewcDoNB4pre+xQdpmZnkBpbQultYOfQKeUOkMDQhB4e08xn/nVJoprmnjurgu5b9kEj49dND6F6y9I57frj3K0vOfZxe4OuK2B0JOc0fGU1bVQXndmtNBxu++hp/4Dl9lj7Y5lbTZSg5CVlRVwtYPB0oAQwFraO3jkjX38y4s7mTQylre/tpRLp/Z/ZM7DV00lIjSER97cT19NjDWNbZyqbmLa6Lhe9+tubYTODuUe+jU6jx2dQIjAXk10N6wNp+bq4WKw11QDQoA6WdXITU9t5fmtx7n7omxeuXdxt6OJPDEiLpKHrpjCxsMVrN5b0uu+B0q6n6Hc1XR7pJF7s1FuUS2hIcKkkbG9HhsV7mDyyDhNhT2MRUZGUllZqUHBi4wxVFZWEhkZOeBz6DwEPzHG+GwM87v7S3ho1W4M8Ns75nHF9FGDPufti8bxykcn+cnbuVw6Na1znYKuDvSQsqKrhOgw0hOjzspptL+olokjYj2aHT0zPYH3D5b59Doq38nIyKCwsJDy8nJ/FyWgREZGkpHRd/9gTzQgnGfVja28tqOQF7adICI0hDUPLvXqF9rWI5Xc+6cdzExPYOXn55KZ0r9JZT1xhAg/vHY6Nz21ld98cIR/vXxKt/vlFtWSEhNOWlxEt6+7mz7m7BQWucW1XDwpzaPyzMpIYNWOQopqmgdc81H+ExYWRnZ2tr+LobrQgHAeGGPYdbKa//vwBH/bU0RLu5PRCZEcq2ggv6yeSSN7b2/vjzX7iokKc/DqfYs9HibqqQuzkrluzhh+u+EoN80b222wOVBS22PKiq5yxsSz9kApja3t1Le0U17X0mf/gcvMDDsVdmG1BgSlvET7EHyosbWdl/55gqt/tYnrn9zCmn3FfHZeBqu/tpRV9y8GYP0h71aZNxwqZ9H4ZK8HA5eHr5pGWIjw6N/OncHc3uHkUGl9nx3KLjmj4zHGynvUmfK6j6Yml2mj4whzCLt1pJFSXqM1BB9pae/giic2cLKqiSkj4/jRtdO57oL0s5aTnJAWw4bDFdy9dLxX3vN4ZQMFlY3c9Yksr5yvOyPjI/l/n5zET9cc5IO8Mi5xyyd0tKKB1nanx3f50+31DXKLa6lrtlZB8/TYiFAHU0bFaQoLpbxIawg+8s9jVZysauKxz87k719fyh2Ls84KBgDLJo9g29FKmtv6vwhHdzbYtY2LJ3vWDj9QX1qSzfjUGH74Vu5ZS2G67vL7moPgMiYhkoSoMHKLathfVEtGUtRZSfb6MjM9kT2F1TpSRSkv0YDgI+8fLCM8NIRrZqf32J5+8eRUWtqdbDtW5ZX3XH+ogrHJUWR3ySDqbeGhITxyzXSOVTTw7KaCzu0HimsJd4QwIa33YaMuItKZCvtAUe05ayj3ZVZGArXN7Zyoaux7Z6VUnzQg+MgHeeUsHp/Sa1v+wuwUwkNDWJ83+H6E1nYnW49UcPGktPMyDHPZ5DSW54zkV+8f7lyDObfYGjYa5vD8v1XO6HgOlNRxrLKhMwuqpzqX1NRmI6W8wqO/XBFZISJ5IpIvIt/tYZ+bRSRXRPaLyItu278gIoftxxfcts8Tkb32OX8pATSY/FhFA8cqGrisj1nBUeEOFmYns+Hw4APCjuOnaWjt8Hlzkbt//3QO7U7Df605AFidw572AbjkjImntd2JMZ73H7hMGRVHeGgIu07qjGWlvKHPgCAiDmAlcCWQA9wqIjld9pkEPAwsMcZMB75ub08GHgEWAguAR0QkyT7sN8A9wCT7scIbH2goWHewDMCjBVyWTU4jv6yeU9WDW+xiw+FyQkOET0xIGdR5+iMzJZr7Lx7PG7uKWL23mIr6Fo/7D1zcE9n1t8kozBHC3MxEth6p7NdxSqnueVJDWADkG2OOGmNagZeBa7vscw+w0hhzGsAYU2ZvvwJYa4ypsl9bC6wQkdFAvDHmQ2P1CP4RuM4Ln2dIWJdXxvi0GI8mhS2z7+g3DHL46fq8cuaOSzqn49rXvnLJRNITo/jWKmvREE+HnLqMT4shPDSExOgwRif0f8r9RRNTyS2upaqhtd/HKqXO5klASAdOuj0vtLe5mwxMFpHNIvKhiKzo49h0+/fezjksNba2s+1oFZd5uLzjxBGxjE6IHFRAKK9rIbe4tjO4nE9R4Q6+9+lpNLRao408nUfgEuYIYU5GIvMykwbU97FkorVI+ZYjFf0+Vil1Nm/NQwjFava5BMgANojITG+cWETuBe4FyMzM9MYpfWpzfiWtHU6Ps4qKCBdPSmP1vmLaO5yE9qND1mWj3Qfhj4AAcOWMUVw0MZVT1U0kRof3+/jf3jGPkJCBdSHNTE8gLiKUzfmVXD1rzIDOoZSyePLtcwoY6/Y8w97mrhB40xjTZow5BhzCChA9HXvK/r23cwJgjHnaGDPfGDM/Lc0/X3j9sS6vjJhwBxdmJXt8zLIpadQ1tw+4c3T9oXJSYsL7fXfuLSLC03fO49X7Fg/o+KSY8H7NP3AX6ghh0YQUNudrDUGpwfIkIHwETBKRbBEJB24B3uyyz+tYtQNEJBWrCeko8A5wuYgk2Z3JlwPvGGOKgVoRWWSPLroTeMMbH8ifjDF8cLCMiyalEh7q+Z3+kgmphMjA+hGcTsPGwxUsnZQ64Ltsb4gOD/UooZ0vLJmQwomqRk7qfASlBqXPby1jTDvwANaX+wHgVWPMfhF5VESusXd7B6gUkVxgHfAtY0ylMaYK+BFWUPkIeNTeBvBV4BkgHzgCrPHi5/KLvNI6imqaPRpd5C4hOow5YxNZf7j/d7n7i6wO1WVThn7tyVcummT1I2gtQanB8agPwRizGljdZdsP3H43wDftR9djnwWe7Wb7dmBGP8s7pK07aN3hD2RVsmWTR/DEPw5R1dBKcozn7fDrD1kDupZ6mDY6EE1Ii2VkfASb8iu4ZcHQ72dSaqjSmcpetO5gGTmj4xkZ3//hkxdPTsUY2NTPu9wNhyqYkR5Paqx/mmuGAhFhyYRUth6pxOnUvEZKDZQGBC+paWxjx4nTXDp1YHfqszISSYwO61cai9pm6z09XVQmkC2ZmEplQysHS+r8XRSlhq2gDwgnqxr58h8+Ir9scF8kG/PL6XCaPtNV9MQRIlw0MZWNh8s9zt65Jb+SDqc5r+kqhiqdj6DU4AV9QHj5oxP842AZ9/xxBzVNbQM+z/sHy0iMDmPO2KS+d+7BxZPTKKtr8fgud8PhcmIjQpmbOfD3DBSjEiKZkBbT7yY3pdQZQR0QjDGs3ltCVko0J6sa+frLH9MxgDZop9OwPq+ciyel4RjE0E/XxDJPVlEzxnrPxRNS+jXENZBdNDGVbUeraG13+rsoSg1LQf1Nkldax7GKBu65eDyPXDOddXnlPL42r9/n2XuqhsqG1gE3F7mMjI9k6qg4j+YjHK1o4FR1kzYXufnExFSa2jo0+6lSAxTUAWH13hJCBC7PGcXtCzO5dcFYVq47wtt7ivt1nvcPliHinZXKlk1OY3vBaRpa2nvdzxU0lmmHcqdF41MIkf6P1FJKWYI8IBSzIDuZtLgIRIT/uGY6czMTeWjVbg4U13p8ng/yypgzNrFf8wd6cvHkNFo7nHx4tPeUzusPlZOd6llG1WCREBXGzIxEtmhAUGpAgjYgHC6tI7+snqtmju7cFhHq4Knb5xEfFcq9f9rOaQ9SKpfXtbC7sMbj7KZ9mZ+VRFSYo9dmo+a2Dj48WsnF9gxddcZFE1P4+GQ1dc0DHyCgVLAK2oCwem8JIrBi+qizto+Ij+Sp2+dRWtPCAy/tpL2j9w5KVwfwQGYndyci1MHiCSls6CWNxfaC0zS3OYM6XUVPlkxMpcNp+KeX1qlWKph4K/31sLNmXzEXjktmRDezii/ITOLH18/g26/t4adrDvL9q3O6OYNlXV4ZI+Ii+r3aV28unpTK+wfLWLkun+hu1mTecKiccEcIi8afv9XRhou5mUlEhIawOb+ST04b6e/iKDWsBGVAOFJez8GSOh75TM9f9DfPH0tuUS3PbDpGQ2sHszISmDgilglpsZ19Be0dTjYcKufKGaO8urD9J6eN5D/XHOR/3ul5xNPlOSOJDg/Kf75eRYY5WJCdrInulBqAoPxG+fu+EgBWzBjV637f+/Q0Kupb+MvOQl7654nO7ckx4UxIiyElJoK65vZ+Zzfty9jkaHb/4HJa2jt63Od8L5U5nHxiQiqP/f0gZXXNjIjrf14ppYJVUAaEt/cUMzczkdEJUb3uF+YI4defn4vTaThV3UR+eT1Hyuo5Ul7PkbIG/llQRWpseGf6ZW+KCncQ1U1zkerbRRNTeQzYeqSSa+cExMqsSp0XQRcQCioayC2u5fufnubxMSEhwtjkaMYmR59TG3A6jV8XplHnyhkTT2J0GJvzKzQgKNUPQTfKaI3dXHSl23DTwdBgMPQ4QoTF41PYdLiix0SBxhgKKhoGlKpEqUAVhAGhmNljE0lP7L25SA1vSyamUlTTTEHl2ctqtnU4eWPXKa759WYu+dkH/HlnoZ9KqNTQE1RNRierGtlTWMPDV071d1GUj7nSYW/OryA7NYaaxjZe+ugEf9hcQEltM+PTYggPDeFgsa6foJRLUAUE1+iiq7zUXKSGrqyUaNITo3h7TzGHS+tYtaOQxtYOlkxM4T9vmMElk0dw5S82cqKqwd9FVWrICKqA8PbeYmamJzA2WfP/BDoRYcnEFF7dXsj241VcOyedLy3JJsdtAmFmSjQFFRoQlHIJmoBwqrqJXSer+faKKf4uijpP/uXSiUweGcc1s8d0OyN9XHI0Gw6V60gxpWxB06nsai66coY2FwWLcSkx3L10fLfBwHo9mpZ2J+X1Lee5ZEoNTUETENbsLWba6HiyU2P8XRQ1RGSmWP8XjncZiaRUsAqKgFBS08z246e5qo9UFSq4jLP7ko5Xaj+CUhAkAeGd/fboolnaXKTOSE+KwhEinKjSGoJSECQBYfXeYqaMjGNCWqy/i6KGkDBHCGMSI7XJSClbUIwy+sUtF1Bc0+TvYqghKDM5muNaQ1AKCJIawqiESC7ITPJ3MdQQlJkcwwntQ1AKCJKAoFRPxqVEc7qxjVpdg1kpDQgquLlGGp3QfgSlNCCo4JaZYgcE7UdQSgOCCm7jdHKaUp00IKigFhsRSkpMuGY9VQoNCEqRmRKtNQSl0ICglDUXQQOCUhoQlBqXHE1xTROt7U5/F2VI+n8vfcyTH+T7uxjqPNCAoIJeZkoMTgOFp7WW0FVbh5N39pWw9Uilv4uizgMNCCrojdOhpz06XtlAa4eT8jpdMyIYeBQQRGSFiOSJSL6IfLeb1+8SkXIR2WU/7nZ77TER2Wc/Pue2/Q8icsztmDne+UhK9U/n5DQNCOfIK6kHoEIXEQoKfSa3ExEHsBJYDhQCH4nIm8aY3C67vmKMeaDLsZ8G5gJzgAjgAxFZY4yptXf5ljHmtcF+CKUGIy0ugqgwh3YsdyOvxPpTrWxopb3DSahDGxUCmSf/uguAfGPMUWNMK/AycK2H588BNhhj2o0xDcAeYMXAiqqUb4iIjjTqQV5pHQDGQFVDq59Lo3zNk4CQDpx0e15ob+vqsyKyR0ReE5Gx9rbdwAoRiRaRVOBSYKzbMT+xj/m5iEQM5AMo5Q2ZKdE6Oa0beSV1RIZZXxNl2o8Q8LxV/3sLyDLGzALWAs8DGGPeBVYDW4CXgK1Ah33Mw8BU4EIgGfhOdycWkXtFZLuIbC8vL/dScZU627jkaE5UNWKM8XdRhoym1g6OVzWyIDsFgHLtRwh4ngSEU5x9V59hb+tkjKk0xrj+tzwDzHN77SfGmDnGmOWAAIfs7cXG0gI8h9U0dQ5jzNPGmPnGmPlpaWmefi6l+iUzJZrmNqfeBbvJL6vHGFg6MRVARxoFAU8CwkfAJBHJFpFw4BbgTfcdRMR9seJrgAP2doeIpNi/zwJmAe+6HyMiAlwH7BvcR1Fq4DLtkUbaj3DGQbtD+RMT7RqCBoSA1+coI2NMu4g8ALwDOIBnjTH7ReRRYLsx5k3gayJyDdAOVAF32YeHARut73xqgduNMe32ay+ISBpWrWEXcL/3PpZS/ePKenqiqpEF2cl+Ls3QcKi0jojQEKaOiicuIlQDQhDwaE1lY8xqrL4A920/cPv9Yaw+ga7HNWONNOrunJf1q6RK+VB6YhQhgi6n6SavtJ5JI2NxhAhpcRHahxAEdFCxUkB4aAhjEqM4rpPTOuWV1DJ5ZBwAqXERWkMIAhoQlLKN0zTYnaobWymtbWHqKCsgpMVFUKEBIeBpQFDKlpkco+krbIdKrZQVrhpCWqzWEIKBBgSlbONSoqmXYUkKAAAZE0lEQVRqaKWuuc3fRfE7V8qKKW41hLqWdppaO3o7TA1zGhCUsunQ0zPySuuIjwxlVHwkYAUE0CR3gU4DglK2TM162imvpI4po+Kwh4x3BgSduBfYNCAoZdN1ESzGmM6A4JIWawUE7UcIbBoQlLLFRYaRHBMe9E1GpbUt1Da3M2XkmYAwwq4h6FyEwKYBQSk3mcma9dSVsmKyW0BIjglHRGsIgU4DglJudC6ClbICOKvJKNQRQkpMuAaEAKcBQSk345KjKapuorXd6e+i+E1eST0j4yNIjA4/a3uqzkUIeBoQlHKTmRKD08Cp6iZ/F8Vv8kprz2ouctF8RoFPA4JSblwjjY4HaZK7DqfhcGl9Z8oKd5q+IvBpQFDKTbDPRThe2UBLu7PnGkJdi64qF8A0ICjlZkRcBJFhIZwI0o5lV4fy1FHx57yWFhtBa4eT2qb2c15TgUEDglJuRITM5OigTYOdV1KPCEwcEXvOa2mdcxGaz3ex1HmiAUGpLjKTY4K2hpBXWsu45Giiwh3nvKbpKwKfBgSluhiXEs2JqsagbCvvmrLCXedsZQ0IAUsDglJdjEuJpqmtY8BffM9sPMrj7+bR2Dq82tqb2zooqGw8K2WFu7Q4K/OpBoTApQFBqS4602APoB+hsbWd/3knj1++n8/yxzfwjwOl3i6ezxwpr6fDaZjcQw0hPjKU8NAQnYsQwDQgKNXFuJQYYGDrImw8XEFLu5NvXTGF6HAHX35+O/f9aTvFNUN/otuZEUbdBwQR0ZXTApwGBKW6SE+MIkTgxAAmp63NLSU+MpR7Lx7P219byrdXTGH9oXI+9b/reWbjUdo7/JMS4+drD3HvH7f3+v4HS+oId4R0BsTuuOYiqMCkAUGpLsJDQxidENXvyWkdTsP7B8u4dOoIwhwhhIeG8NVLJrL2G8tYkJ3Mj98+wGd+vZmPT5z2Ucl7tv5QOe/mlvLz9w71uM+hkjomjIglzNHz14IGhMCmAUGpboxLieZYP5uMdhw/TVVDK8tzRp61fWxyNM/edSG/uW0uVQ0t3PzbrRRUnN/UGCU1zYQ5hJXrjrAur6zbfQ6V1jNl5LnzD9ylxUXoMpoBTAOCUt2Ym5nE3sJqSmo8n4S1NreEMIewbHLaOa+JCFfOHM2bD1yEMfDCtuPeLG6v2juclNU1c9cnspg6Ko5vvLKLoi7J+2qb2zhV3dRjh7JLWmwElQ2tfmv6Ur6lAUGpbtw0PwOngT/vLPRof2MMa3NLWTwhlbjIsB73GxkfyRUzRvHq9kKaWju8Vdxelde34DSQnRrLk7fNpb3D8C8v7jwrxffhPjqUXdLiIjAGqhpafVpm5R8aEJTqxriUGBaNT+aVj07idPY9QS2/rJ6CysZzmou6c+eicdQ0tfHWniJvFLVPxXYtZ3RCJOPTYvnpZ2fy8Ylq/vvvBzv3ySupB+g2qZ07na0c2DQgKNWDWy7M5ERVIx8eq+xz37X2fINPTRvR574LspOZPDKWP209fl5mQxdXWwFhVII1sezqWWP4wuJxPLPpGH/fVwJAXkktsRGhpCdG9XquNF1bOaBpQFCqBytmjCIuMpRXPjrZ575rc0uZmZ7A6ITev1DB6k+4Y3EWe0/VsLuwxuPy7Dh+mt+uP+Lx/i6uORCj7YAA8G+fnsasjAS+9dpuTlQ2kldax+SRsYhIr+dKi9X0FYFMA4JSPYgMc3D9Bems2VdCTWNbj/uV1TWz62S1R81FLtdfkE5MuIM/bi3waP+m1g6+9tLH/PTvB/u9vGdJTTNRYQ4Sos70bUSEOlj5+bkI8NUXd3CwlxxG7tI0n1FA04CgVC9unj+W1nYnr+861eM+/zhQhjH0KyDERoRyw9wM/ran2KMO2pXr8jlV3YQx9HvWc3FtM6MTIs+5+x+bHM3/3jyHfadqqW5s6zGHkbvIMAdxkaEaEAKUBgSlejEjPYEZ6fG9NhutzS0lIymqzxE6Xd2xeByt7U5e3d57k9TR8nqe3nCU8WnWDOLC0/0MCNVNnf0HXS3PGcm9F48HIGdMgkfn08lpgUsDglJ9+Nz8seQW17Lv1Lnt/Q0t7WzKr2B5zsg+29+7mjwyjoXZybyw7TgdPYxkMsbwyJv7iQgN4X9unA3AqX4GhJKa5h4DAsC3r5jCi/cs5MKsJI/Op/mMApcGBKX6cM2cdCJCQ3j5oxPnvLbxcDmt7c5+NRe5u3NxFiermlh/qPvZw+/sL2Hj4Qq+eflkZmUkECJQeNrzGdQdTkNpXQtjeunsDnWE8IkJqR4HtLS4CB1lFKA0ICjVh4SoMK6aOZo3dhWdM5nsXTuZ3YVZyQM69+XTRzIiLoI/bj135nJjazuPvpXL1FFx3LFoHGGOEEbFR/aryaiivoUOp+m1htBf2mQUuDQgKOWBm+ePpa65nTX7iju3tXc4WXewjMvsZHYDEeYI4dYFmaw/VM7xLtlVf/1+PkU1zfzouhmE2ufPSIqmsNrzgOA+Kc1b0uIiqG9pH3YLAKm+aUBQygOLxieTlRJ9VufyjuOnOd3YxvKcUYM6960LMgkR4YVtZ5qkjpTX87uNR/ns3Iyzah8ZSVH96kMotoOHV2sI9lyEijpNXxFoPAoIIrJCRPJEJF9EvtvN63eJSLmI7LIfd7u99piI7LMfn3Pbni0i2+xzviIi4d75SEp5n4hw0/yxbDtWxTE7U+na3FLCHSEsm3JuMrv+GJUQyRXTR/Lq9pM0t3VgjOE/3txPZJiD71459ax905OiKK5pos3D5HJnagh9T5jz1JnZyp4n/lPDQ58BQUQcwErgSiAHuFVEcrrZ9RVjzBz78Yx97KeBucAcYCHwkIjE2/s/BvzcGDMROA18edCfRikfunFeBo4Q4dXtJ61kdgdKWTwhhdiI0EGf+/ZF46hubOOt3UWs2Wd1JD90+ZTOL1+XjKQonAaPs7CW1DYTERpCUnTPCff6SyenBS5PaggLgHxjzFFjTCvwMnCth+fPATYYY9qNMQ3AHmCFWMMZLgNes/d7Hriuf0VX6vwaGR/JpVPSeG1HIQeK6zjuYTI7Tywen8LEEbE8u7mAH/0tl5zR8dy2MPOc/TKSrPWePe1YLq7pflLaYGhACFyeBIR0wH3mTKG9ravPisgeEXlNRMba23ZjBYBoEUkFLgXGAilAtTHG1SvV0zmVGlJunj+W8roW/v2NfQB8app3AoKIcMeicRworqW4ppkfXTe9syPZnSv5nKdDT3ublDZQKTERhIgGhEDkrU7lt4AsY8wsYC3WHT/GmHeB1cAW4CVgK9CvJPAicq+IbBeR7eXl5V4qrlIDc+nUEaTFRbDj+GlmZSR49cv2hrnpJESF8bn5Y5k3rvthrKMTIxGBUx6ONLJqCN7rPwBwhAjJMToXIRB5EhBOYd3Vu2TY2zoZYyqNMa7/Hc8A89xe+4ndr7AcEOAQUAkkikhoT+d0O/5pY8x8Y8z8tLTBdd4pNVhhjhA+OzcDgOVeqh24xEWGse6hS/jJ9TN63Cci1MHIOM/mIjidhlI7j5G36VyEwORJQPgImGSPCgoHbgHedN9BREa7Pb0GOGBvd4hIiv37LGAW8K6xksCvA260j/kC8MZgPohS58sdi8exMDuZ6y7wfitnckx4t01F7tKTojxqMqpoaKHdaTQgKI/1OTzCGNMuIg8A7wAO4FljzH4ReRTYbox5E/iaiFwDtANVwF324WHARrtDqxa43a3f4DvAyyLyY+Bj4Pfe+1hK+U56YhSv3LfYb++fkRTFzhOn+9zvzMI43m0yAmsuQr697KYKHB6NlzPGrMbqC3Df9gO33x8GHu7muGaskUbdnfMo1ggmpVQ/ZCRF8faeYto7nL3WJnwxS9nFlc/IGOPVEUzKv3SmslLDTHpiNO120rrelHSzUpq3pMVF0NZhqGnqeeEgNfxoQFBqmMlIspqA+kphUVzbTLgjhOQY7ycB0LkIgUkDglLDjCsg9NWx7FoHwRdNOrq2cmDSgKDUMDOmc3JaHzWE6t4XxhmMM/mMNCAEEg0ISg0zkWEO0uIiPGgyamKMrwOC1hACigYEpYah9MQoCqt7bjJyOg2lNS0+GXIKEB8ZSnhoiAaEAKMBQalhKCMpqtcmo6rGVlo7nD4ZYQRW7iVdWznwaEBQahjKSIqmqLoJp9N0+/qZSWm+CQigaysHIg0ISg1D6UlRtHUYynq4Qy+25yCM8VGTEWj6ikCkAUGpYaivoaclteephqABIaBoQFBqGBrrmpzWQxrs4ppmwhxCig8mpbmkxUZQ1djq8XKeaujTgKDUMNTXXISSmmZGxkcSEuK7PEMj4iMwBqoaWn32Hur80oCg1DAUHR5KSkx4j01GRdVNPu0/AJ2tHIg0ICg1TPU29LSk1nezlF10clrg0YCg1DCVnhTV7WxlY4y9dKYGBNU/GhCUGqYykqIp7GYuwunGNlrbnT6vIaTGaj6jQKMBQalhKiMpitZ2JxUNZ38hF1X7bh0Ed5FhDuIjQ7WGEEA0ICg1TKX3MNKopHOlNN92KoPORQg0GhCUGqYykqKBcwNCca3vls7sSgNCYNGAoNQwld7DymklNU2Ehggpdhu/L6XFRWofQgDRgKDUMBUbEUpidNg5cxGK7UlpDh9OSnPRjKeBRQOCUsNYd3MRiqt9P+TUJS0ugvqWdhpb28/L+ynf0oCg1DCWkRh9Tj6j8zEpzWVWRgI3z8+grb37NNxqeAn1dwGUUgOXnhTFB4fKMMYgIvaktCY+NW3EeXn/JRNTWTIx9by8l/I9rSEoNYxlJEXR3Oak0k4wV9PURnOb02dLZ6rApgFBqWHMNfTUNdKoyF4pbcx5ajJSgUUDglLDWNfJaSW11s/z1YegAosGBKWGsfQuK6cVn8dZyirwaEBQahhLiAojPjK0c6RRSU0zjhDpzESqVH9oQFBqmEtPiu5sMiqqbmZkXMR5mZQ2aIXbYd1/Qk2hv0uibBoQlBrmrMlpVpNRSW3T8Ok/2PA/sP4x+MVs+Ov9UHbA3yUKehoQlBrmMuyFcs4sjDMM+g+cHXB8C0y9Gi68B3LfgCcXwYu3wIkP/V26oKUBQalhLj0xiobWDk43tlFSc/5mKQ9KyV5oqYXp18OVP4Vv7IdLHoaT2+DZK+D3V0DeGjA6A/p80oCg1DDnmotwoLiWxtaO85bHaFAKNlk/xy2xfkYnwyXfhW/sgyv/G2qL4KVb4JXboanaf+UMMhoQlBrmMuyhpx8VVAHDZMhpwSZIngDxo8/eHh4DC++Dr+2E5T+CQ3+Hp5dB8W7/lDPIaEBQaphzBYTtBaeBYTApzdV/kHVRz/s4wmDJ1+Cut6G9FZ5ZDjv+oE1IPqYBQalhLiEqjNiIUHaesALCkG8yKt0HLTWQtbTvfTMXwf0bIWsJvPWgNRqptcH3ZQxSGhCUGuZEhIykKBpbOwgRGDHUJ6W5+g+ylni2f0wq3Paa1em85xX43Seh/JDvyhfEPAoIIrJCRPJEJF9EvtvN63eJSLmI7LIfd7u99t8isl9EDojIL0VE7O0f2Od0HXN+8vUqFYBcOY1GxEUS6hji93kFmyB5PMSP8fyYEIfV6XzHX6ChDJ6+BHa/ok1IXtbn/xwRcQArgSuBHOBWEcnpZtdXjDFz7Mcz9rGfAJYAs4AZwIXAMrdjbnM7pmyQn0WpoOXqRxge/Qebe+8/6M2Ey+D+TTB6Fvz1XmsUUr1+dXiLJ7cSC4B8Y8xRY0wr8DJwrYfnN0AkEA5EAGFA6UAKqpTqmWvo6bDoP2j2sP+gJ/FjrM7mT/0QDq+FlQth31+8V8Yg5klASAdOuj0vtLd19VkR2SMir4nIWABjzFZgHVBsP94xxrjPT3/Obi76d1dTklKq/1xZT4f8kNOCzdbPcR72H/QkxAEXfR3u2wBJWfDaF+HVO6GhYtBFDGbeamx8C8gyxswC1gLPA4jIRGAakIEVRC4TEdetwW3GmJnAUvtxR3cnFpF7RWS7iGwvLy/3UnGVCiwZnQFhiNcQXP0HCd3dUw7AiKnw5bXwyUesmc0rF8L+171z7iDkSUA4BYx1e55hb+tkjKk0xrTYT58B5tm/Xw98aIypN8bUA2uAxfYxp+yfdcCLWE1T5zDGPG2MmW+MmZ+WlubZp1IqyEwaEcfSSalcNGkIr2/sdFr9B4OtHXTlCIWl34R710NCBqz6Aqy6C04XePd9goAnAeEjYJKIZItIOHAL8Kb7DiLiPt3wGsDVLHQCWCYioSIShtWhfMB+nmofGwZcDewb3EdRKnhFhTv405cXMm10vL+L0rPSfdBcPbj+g96MzIG734PLvg8HV8Ov5sEbD0DVMd+8XwDqMyAYY9qBB4B3sL7oXzXG7BeRR0XkGnu3r9lDS3cDXwPusre/BhwB9gK7gd3GmLewOpjfEZE9wC6sGsfvvPexlFJDznG7/8DT+QcD4QiDi78FD+6C+V+GPa9ageH1r0LlEd+9b4AQM4zG8c6fP99s377d38VQSg3Ey7dB6X7ry/p8qS2Gzb+AHc9BRxvMutkKGCkTzl8ZhgAR2WGMmd/XfkN8BotSKiA4nVaHsi9rB92JH22l135wNyy83+pw/vV8eOlW2PuapsHoItTfBVBKBYGy/b7tP+hL3ChY8Z+w5EH4cKXVlJS3GsKiYcpVMOOzMPGTEDrE0374mAYEpZTvdV3/wF/iRsLyR61hqse3wL4/W6u17XsNIhNg2mcg53orqV5ErH/L6gcaEJRSvlewyZpAlji2z13PixAHZC+1Hlf9DxxZZwWH/a/Dx/8H4oBRMyFzMWQutH7GjfJ3qX1OA4JSyrdc8w+mftrfJemeIwwmX2492pqssp7YBie2WmswbPuNtV9SlhUYRs6A1EnWI3GcFVwChAYEpZRvleVC02kYN8CEdudTWBRM/JT1AGtkUvEeKzic2Ar5/4DdL53Z3xFurfyWOhFSJlmzsONGW01TcaMhKhlChs/YHQ0ISinf6u/6B0OJIwwy5lmPTzxgbWusgorDUHEIKg9DRT6UHbRSZzjbzz4+JBRiR1nNTXGjICoRIl2PBPt5gvU8Itbq5A6LhnD753mufWhAUEr51vFNVtNKYqa/S+Id0cl2v8LCs7d3tEHtKagrhbpiqLd/up5XHYWmamu0VVujZ+/liLBqLeEx8IW3fD5/QgOCUsp3nE4rw+mUq/xdEt9zhFn9DElZfe/b3mqlAW+usQJEUzW01ll9GK0N1s+2RuvR2mg9j4jz9SfQgKCU8qHyA9BUNfAFcQJVaDjEplmPIWT49HYopYaf4dx/EIQ0ICilfKdgo9V3ECj9BwFOA4JSyjecTms2sL/SVah+04CglPKN1nqYdAVMudLfJVEe0k5lpZRvRMbD9b/xdylUP2gNQSmlFKABQSmllE0DglJKKUADglJKKZsGBKWUUoAGBKWUUjYNCEoppQANCEoppWxijPF3GTwmIuXA8QEengpUeLE4gUqvk+f0WnlGr5NnfHmdxhlj+kytOqwCwmCIyHZjzHx/l2Oo0+vkOb1WntHr5JmhcJ20yUgppRSgAUEppZQtmALC0/4uwDCh18lzeq08o9fJM36/TkHTh6CUUqp3wVRDUEop1YugCAgiskJE8kQkX0S+6+/yDBUi8qyIlInIPrdtySKyVkQO2z+T/FnGoUBExorIOhHJFZH9IvKgvV2vlRsRiRSRf4rIbvs6/dDeni0i2+y/v1dEJNzfZR0KRMQhIh+LyN/s536/TgEfEETEAawErgRygFtFJMe/pRoy/gCs6LLtu8A/jDGTgH/Yz4NdO/CvxpgcYBHwL/b/Ib1WZ2sBLjPGzAbmACtEZBHwGPBzY8xE4DTwZT+WcSh5EDjg9tzv1yngAwKwAMg3xhw1xrQCLwPX+rlMQ4IxZgNQ1WXztcDz9u/PA9ed10INQcaYYmPMTvv3Oqw/4nT0Wp3FWOrtp2H2wwCXAa/Z24P+OgGISAbwaeAZ+7kwBK5TMASEdOCk2/NCe5vq3khjTLH9ewkw0p+FGWpEJAu4ANiGXqtz2M0gu4AyYC1wBKg2xrTbu+jfn+UJ4NuA036ewhC4TsEQENQAGWsImg5Ds4lILPBn4OvGmFr31/RaWYwxHcaYOUAGVu18qp+LNOSIyNVAmTFmh7/L0lWovwtwHpwCxro9z7C3qe6VishoY0yxiIzGutMLeiIShhUMXjDG/MXerNeqB8aYahFZBywGEkUk1L771b8/WAJcIyJXAZFAPPALhsB1CoYawkfAJLsHPxy4BXjTz2Uayt4EvmD//gXgDT+WZUiw23d/Dxwwxjzu9pJeKzcikiYiifbvUcByrP6WdcCN9m5Bf52MMQ8bYzKMMVlY30fvG2NuYwhcp6CYmGZH4icAB/CsMeYnfi7SkCAiLwGXYGVZLAUeAV4HXgUysTLL3myM6drxHFRE5CJgI7CXM22+/4bVj6DXyiYis7A6Qx1YN5uvGmMeFZHxWIM5koGPgduNMS3+K+nQISKXAA8ZY64eCtcpKAKCUkqpvgVDk5FSSikPaEBQSikFaEBQSill04CglFIK0ICglFLKpgFBKaUUoAFBKaWUTQOCUkopAP4/fJ+FhH/lMu8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 30, 4)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# do some now & past predictions\n",
    "for t in [0, 100, 200, 300, 500, 1000, 2000]:\n",
    "    rt_x, prediction = predict(time.time()-t*period)\n",
    "\n",
    "    current = rt_x[:,0]\n",
    "    prediction = prediction[0]\n",
    "\n",
    "    pyplot.plot(current, label='current')\n",
    "\n",
    "    # shift train predictions for plotting\n",
    "    predictPlot = np.empty_like(current)\n",
    "    predictPlot[:] = np.nan\n",
    "    predictPlot = np.append(predictPlot, prediction)\n",
    "\n",
    "    pyplot.plot(predictPlot, label='prediction')\n",
    "    pyplot.legend()\n",
    "    pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sat, 30 Jun 2018 01:40:00 +0000\n",
      "(1, 30, 4)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD8CAYAAACb4nSYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzt3Xl8lOW98P/PdyYr2VcICVnY9y2RRQRFa4u2xVax1rrx+FhbW6s9Hp+q55ynnsfHc361i9X20KdatNVTd60ttiiigIoIEhDQQNiSAFkgO9nXuX5/zCQMIcskuZOZTL7v12teM/d25Zob5jvXfO/rvi4xxqCUUmp0sHm7AkoppYaPBn2llBpFNOgrpdQookFfKaVGEQ36Sik1imjQV0qpUUSDvlJKjSIa9JVSahTRoK+UUqNIgLcr0FV8fLxJT0/3djWUUmpE2bNnT7kxJqGv/Xwu6Kenp5Odne3taiil1IgiIic82U/TO0opNYp4FPRFZJWIHBaRYyLyYDfb14pImYjscz3ucK1f6bZun4g0icg3rH4TSimlPNNnekdE7MA64EqgENgtIhuMMQe77PqKMeZu9xXGmK3AfFc5scAx4F0rKq6UUqr/PMnpLwKOGWPyAETkZeAaoGvQ78sa4G1jTEM/j1NKjVCtra0UFhbS1NTk7ar4jZCQEFJSUggMDBzQ8Z4E/WTglNtyIbC4m/2uE5EVwBHgn4wxp7ps/zbw+IBqqZQakQoLC4mIiCA9PR0R8XZ1RjxjDBUVFRQWFpKRkTGgMqy6kPsWkG6MmQtsBp5z3ygiScAcYFN3B4vInSKSLSLZZWVlFlVJKeVtTU1NxMXFacC3iIgQFxc3qF9OngT9ImCC23KKa10nY0yFMabZtbgeyOxSxreAN40xrd39AWPM08aYLGNMVkJCn91MlVIjiAZ8aw32fHoS9HcDU0QkQ0SCcKZpNnSpRJLb4mrgUJcybgReGkxFlVKDk3u6ho+O6i/p0a7PoG+MaQPuxpmaOQS8aozJEZFHRGS1a7d7RCRHRPYD9wBrO44XkXScvxQ+sLbqSqn+eOztXO58fg81Td3+4FZDYNu2bezYscPb1TiPR3fkGmM2Ahu7rPup2+uHgId6OLYA58VgpZQH3v68hEmJ4UwdG2FpuTnFNTS2tvPXz4q4dWm6pWX7m7a2NgICzoVHYwzGGGy2/l0G3bZtG+Hh4Vx88cVWV3HA9I5cpXxITVMrP3rpM/5zY9cM6eCU1TZTWuu87PbCzpMYYywt35c9//zzzJ07l3nz5nHLLbewdu1aXn/99c7t4eHhgDNAL1++nNWrVzNz5kwKCgqYNm0at956K7Nnz+bUqVO8++67LF26lIULF3L99ddTV1cHOIePefjhh1m4cCFz5swhNzeXgoICfv/73/PrX/+a+fPn89FHH3nl/Xflc2PvKDWafXSknDaHYcexCmqaWokMGVhf7K4OltQA8I354/nrvmL2nqwiMy3WkrI99X/eyuFgcY2lZc4cH8nDX5/V4/acnBweffRRduzYQXx8PJWVldx333097r93716++OILMjIyKCgo4OjRozz33HMsWbKE8vJyHn30Ud577z3CwsJ47LHHePzxx/npT51Jj/j4ePbu3cvvfvc7fvnLX7J+/Xq+//3vEx4ezv3332/p+x4Mbekr5UPezz2D3Sa0tDvYmltqWbkdwfaBq6YTERzACztPWla2L9uyZQvXX3898fHxAMTG9v5Ft2jRovP6v6elpbFkyRIAdu7cycGDB1m2bBnz58/nueee48SJc2OcXXvttQBkZmZSUFBg8Tuxjrb0lfIR7Q7DB4fLuHpOEjvzKng35wzXzLfmctjBkhqSo0NJigrlGwuSeSX7FP/7azOJCQuypHxP9NYiH04BAQE4HA4AHA4HLS0tndvCwsLO29d92RjDlVdeyUsvdd8RMTg4GAC73U5bW5vV1baMtvSV8hH7C6upqG/hSzMSuXLmWLYeLqWptd2SsnOKzzJrfCQA31mcSkubgzf2FlpSti+7/PLLee2116ioqACgsrKS9PR09uzZA8CGDRtobfWsN9OSJUv4+OOPOXbsGAD19fUcOXKk12MiIiKora0dxDuwngZ9pXzE1txSbAKXTk1g1axxNLS0s/1o+aDLbWhpI7+8npmuoD8jKZKFqdG8uMv/L+jOmjWLf/3Xf+XSSy9l3rx53HfffXz3u9/lgw8+YN68eXzyyScXtO57kpCQwJ/+9CduvPFG5s6dy9KlS8nNze31mK9//eu8+eabPnUhV3ztHz0rK8voJCpqNLr6yY8IDw7g1e8vpaXNQeajm1k1axy/uH7eoMrde7KKa3+3g6dvyeTLs8YB8MaeQv75tf28+N3FXDwp3orqd+vQoUPMmDFjyMofrbo7ryKyxxiT1dex2tJXygecPtvEwZIaVk5PBCAowMYV0xN579AZ2todgyo7x3URd1ZyVOe6r85NIio0kBd3jY4LuuocDfpK+YAtrp46V8xI7Fy3avY4qhpa+bSgclBlHyyuISo0kPFRIZ3rQgLtXLcwhU05pymrbe7laOVvNOgr5QO25JaSHB3KlMTwznUrpiYQHGDj3Zwzgyr7YEkNM5MiLxio6zuLU2ltN7y2p+so6MqfadBXysuaWtv5+Fg5V8xIPC8wjwkK4NKpCWzKOT3gC65t7Q5yS2o6L+K6m5wYzuKMWF769CQOh29d21NDR4O+Ul62M6+Cxtb2zny+u6/MGkfJ2SYOFJ4dUNn55fU0tzk6u2t2ddOSNE5VNvLRscH3ElIjgwZ9pbxsS24poYF2lk6Mu2DbFTMSsduEd3JOD6jsjuEXumvpA3xl1ljiwoJ4cdeJbrcr/6NBXykvMsawJbeUZZPjCAm0X7A9ekwQSyfGsWmAQT+nuIagABuTEsK73R4cYGdNVgrvHSrl9Fmdx9YTHQO0FRcXs2bNml73feKJJ2hoODct+NVXX011dfWQ1q8vGvSV8qKjpXUUVjVy+fSxPe7zlVljySur51hp/+/sPFhcw7SxEQTae/6of2dRKu0Owyu7R+8F3fb2/t/5PH78+PNG6+xO16C/ceNGoqOj+/23rKRBXykv6uiquXJ6z9OEdtxQ9c4X/WvtG2M6e+70Ji0ujOVT4nl598lB3xPgiwoKCpg+fTo33XQTM2bMYM2aNTQ0NJCens4DDzzAwoULee211zh+/DirVq0iMzOT5cuXd95tm5+fz9KlS5kzZw7/9m//dl65s2fPBpxfGvfffz+zZ89m7ty5/Pa3v+U3v/kNxcXFrFy5kpUrVwLOIZjLy53XTx5//HFmz57N7NmzeeKJJzrLnDFjBt/97neZNWsWX/7yl2lsbLT0fOiAa0p50ZZDpcxIiiQpKrTHfcZGhrAgNZpNOWe4+/IpHpd9uqaJyvqWHvP57r6zKJW7XtjLx8cruHTqEM1T/faDcPpza8scNweu+lmfux0+fJhnnnmGZcuWcfvtt/O73/0OgLi4OPbu3QvAFVdcwe9//3umTJnCrl27+MEPfsCWLVu49957ueuuu7j11ltZt25dt+U//fTTFBQUsG/fPgICAqisrCQ2NpbHH3+crVu3do7y2WHPnj388Y9/ZNeuXRhjWLx4MZdeeikxMTEcPXqUl156iT/84Q9861vf4o033uDmm28e5Ik6R1v6SnlJdUMLe05WcUU3vXa6+sqscXxedJbCqoY+9+3QMZxyTz133K2cnkhIoM3S4Zx9yYQJE1i2bBkAN998M9u3bwfghhtuAKCuro4dO3Zw/fXXM3/+fL73ve9RUlICwMcff8yNN94IwC233NJt+e+99x7f+973Omfb6msI5+3bt/PNb36TsLAwwsPDufbaazvH5snIyGD+/PnA0AzTrC19pbzkgyNltDtMt101u/rKrHH87O1c3s05w+2XZPS5P5wL+tP7SO+A8w7diyfFsyW3lIe/PvOCG7ks4UGLfKh0fT8dyx2DrTkcDqKjo9m3b59Hxw+ljiGawTlMs9XpHW3pK+UlW3NLiQ0LYv6Evi/sZcSHMW1sRL968eQU15AeN4bwYM/adiunJ3KysoG88nqP/8ZIcfLkST755BMAXnzxRS655JLztkdGRpKRkcFrr70GOK+H7N+/H4Bly5bx8ssvA/DCCy90W/6VV17JU0891TmOfmWlc+iMnoZWXr58OX/9619paGigvr6eN998k+XLl1vwTvumQV8pL2h3GLYdKeOyqQnYbZ61Ir8yexy7CyqpqPNsrJyDJTXMGh/V944uK6c5c/n+mOKZNm0a69atY8aMGVRVVXHXXXddsM8LL7zAM888w7x585g1axZ/+9vfAHjyySdZt24dc+bMoaioqNvy77jjDlJTUzvn4n3xxRcBuPPOO1m1alXnhdwOCxcuZO3atSxatIjFixdzxx13sGDBAovfdQ86Znn3lUdmZqZRyt/tzq8waQ/83by1v8jjY74oqjZpD/zdvPzpiT73PdvYYtIe+Lv5ry1H+1WvKx/fZr7zh0/6dUxvDh48aFlZA5Wfn29mzZrl7WpYqrvzCmQbD2KstvSV8oL3c0ux24TlUzzvKTMzKZIJsaFs8mAAtkOufH5f3TW7WjktkU/zK6lr9t3p/tTgaNBXygu25pZyUXoMUaGBHh8jIqyaNY6PjpZRcrb3i3sdwy940nPH3crpibS2G0tm7PIV6enpfPHFF96uhs/QoK/UMCuqbiT3dC2Xe9Brp6tbl6YD8Jv3j/a638HiGuLDg0iICO51v64y02KICA5g22Hr8vrGx2bnG+kGez416Cs1zDa7euD0NvRCTybEjuGmxWm8ml1IXlldj/vlFNcwo5sx9PsSaLexfGo8Ww+XWhKsQ0JCqKio0MBvEWMMFRUVhISE9L1zD7SfvlLDbOMXp5mSGM7kxO4HQevL3ZdP5tXsU/xq8xHWfWfhBdtb2hwcLa1lxdSJAyp/5bRENn5+ut+9f7qTkpJCYWEhZWVlgypHnRMSEkJKSsqAj9egr9QwKq1pYndBJff0YziFruLDg7njkgx+s+UY319xljkp5wfmY6V1tLYbj4Zf6M6lbl03Bxv0AwMDycjw7GYyNTw0vaPUMHon5zTGOCcmH4w7VkwkZkwgP9+Ue8G2nGLnhCv97bnTITEihLkpUWw9rK1zf6RBX6lh9I8DJUxODGfq2IhBlRMZEsgPLpvMR0fL2XH8/J42B0tqCA20kxEfNuDyL5uWyGcnq6iqbxlUPZXv8Sjoi8gqETksIsdE5MFutq8VkTIR2ed63OG2LVVE3hWRQyJyUETSrau+UiNHaW0TnxZUcvWcwbXyO9yyNI2kqBB+/s7h8y6UHiyuYXpShMd3+nbn8umJOAx8eFRb+/6mz6AvInZgHXAVMBO4UURmdrPrK8aY+a7Herf1zwO/MMbMABYB/nePtxoy7Q7DgcJqNn5eMuJ7gGz6wpXasSjohwTaufeKKew7Vc3mg84btoyHY+j3ZW5yFHFhQX45JMNo58mF3EXAMWNMHoCIvAxcAxzs60DXl0OAMWYzgDGm5z5mSgEOhzNo7cyrYGdeBbvyK6ltct4d+tJ3l7B00oXzyI4U//i8hEkJYUwdO7BeO91Zk5nC0x/m8YtNh7lixliKqxupbWob9AVYm024dFoCW3NLaXeYQf1qUL7Fk/ROMuA+j1qha11X14nIARF5XUQmuNZNBapF5C8i8pmI/ML1y0Gp8+w4Vs6dz2ez4P9u5mu/3c6j/zjEsdI6vjY3iZ+vmQvA7oJKL9dy4Mpqm/k0v5KvzkmydJjeALuN+78yjaOldbz5WRE5xb1PhN4fK6clUtXQyr5T3p3TVVnLqi6bbwEvGWOaReR7wHPA5a7ylwMLgJPAK8Ba4Bn3g0XkTuBOgNTUVIuqpEaSxzcf4ciZWlbNHsfSSXEsmRh33mxSz3yUT/aJKi/WcHDeyTmNw8DVg+y1052rZo9jTnIUv958hK/NS8ImMG2QF4oBVkxxjgC6NbeUzLQYC2qqfIEnLf0iYILbcoprXSdjTIUxpmO81/VAput1IbDPGJNnjGkD/gpccDeJMeZpY0yWMSYrIWGIpmpTPq2wqpEvzxrHz9fM45sLUi6YPjArPYa9J6pod4zMvP7GAyVMTAizJBh3JSL8ZNU0iqob+dPHBUxKCCc0aPA/qKPGBJKZGsNWC4dkUN7nSdDfDUwRkQwRCQK+DWxw30FE3Jsvq4FDbsdGi0hHJL8cD64FqNGlpc3BmdomkqN7nif2ovRY6prbyD1dM4w1s0Z5XTO78issT+24u2RyPEsnxtHc5rAktdPhsukJ5BTXcKamybIylXf1GfRdLfS7gU04g/mrxpgcEXlERFa7drtHRHJEZD9wD84UDsaYduB+4H0R+RwQ4A/Wvw01kp0+24QxkBzTc9DPSnemF7ILRl6K550vXKkdi3rtdKejtQ8wN6Xvmbg81TEonJUDsCnv8iinb4zZCGzssu6nbq8fAh7q4djNwNxB1FH5ucJq52TfKb209JOjQ0mKCmF3QSW3XZw+TDWzxsbPS5gYH8b0cdandtwtSI3h7z+6ZMBj+nRn2tgIkqJC2JJbyg0X6fU2f6B35CqvK652pg7G9xL0RYSs9Fh2F1SOqP765XXN7Myr4OohTO24m50cRUigdR3kRISV0xPZfrScljaHZeUq79Ggr7yuqMo5IUhSdO/DxV6UHsOZmmYKq3qfQMSXbMoZ+tTOUFs5LZH6lvYR3WVWnaNBX3ldUXUDiRHBBAf03kLNSosFIPvEyAk+Gz8vISM+jBlJQ5vaGUoXT4ojyG7Tu3P9hAZ95XVF1Y29XsTtMG1cBBHBAeweIRdzK+qa+eR4BVfPGTcsqZ2hEhYcwPIp8Tpvrp/Q8fSV1xVVNTI7ue9hA+w2YWFaDNkjJM2wKefMiE/tdFh/W9aI/uJS52hLX3mVw2EoPtt7H313F6XHcORMHdUNvj/k78bPS0iPGzPowc98gQZ8/6FBX3lVeX0zLW0Oj9I7AFnpzrz+Hh8fkqGyvoVPhrHXjlKe0qCvvKqj546nLf15KdEE2sXnx+HZlHOadofxi9SO8i8a9JVXFVW7gr6HLf3QIDuzk6N8Pq//zhenSYsbwywLh0RQygoa9JVX9belD85xePafOktTa/tQVWtQ2tod7C6o5NKpCZraUT5Hg77yqqLqRiJCAogICfT4mMy0GFraHXxRdHYIazZwh0pqaWhp77z+oJQv0aCvvKq4urFfrXyALNfY7r7aX7/jztWL0nUMeuV7NOgrryqsaiTFw3x+h7jwYCYmhPlsXj/7RKVrgLj+vS+lhoMGfeVVRQNo6QNclBZL9okqHD42qYoxht0FVdrKVz5Lg77ympqmVmqb2jzuueMuKz2Gs42tHCurG4KaDdzJygbKaps1n698lgZ95TUdPXd6G1K5Jxe5gqqvjfzYcZ3hIg36ykdp0FdeU1zd/+6aHdLixhAfHuxzM2llF1QSGRLAFAsnMlHKShr0ldf098YsdyLCRekxPtjSryQrPRabTfvnK9+kQV95TVFVI0EBNuLDggd0fFZ6LIVVjZw+6xuTdlfUNXO8rL5zPl+lfJEGfeU1ha6eOwNtFXf0kPGVSVU6BoFbpPl85cM06CuvKapqZHwfUyT2ZmZSJGOC7D6T188+UUVQgI05KX3PDaCUt2jQV14zkLtx3QXYbSxIjfaZvP7ugkrmpUT1Oe2jUt6kQV95RXNbO6W1zSRHjxlUOVlpsRwqqaG2qdWimg1MY0s7XxSd1f75yudp0FdeUVLtvPg6kJ477i5Kj8Vh4LOT1VZUa8D2F1bT2m70Tlzl8zToK68oGkQffXfzU6Ox24SPj5dbUa0B6xgHKDNVW/rKt2nQV14xkHH0uxMeHMDl0xP54/YC9p3yXmt/d0EV08ZGEDXG8yGilfIGDfrKK4qqGxGBcVED773T4efXzSUxMpi7/ryH8rpmC2rXP+0Ow94TVdo/X40IGvSVVxRVNzI2IoSggMH/F4wJC+L3N2dSWd/CD1/YS1u7w4Iaeu7w6Vpqm9t0vB01ImjQV15RVNU46Iu47mYnR/Gz6+awK7+Sn72da1m5nui4OUxb+mok0KCvvGKg4+j35psLUlh7cTrrt+ezYX+xpWX3ZndBFUlRIZa/H6WGgkdBX0RWichhETkmIg92s32tiJSJyD7X4w63be1u6zdYWXk1MjkchpKzjQMaUrkv/3L1DC5Kj+GB1w9wqKTG8vK7MsawO985yJpOgq5Ggj6DvojYgXXAVcBM4EYRmdnNrq8YY+a7Huvd1je6rV9tTbXVSFZa20xru7E0vdMhKMDGupsWEhESwPf/vIezDUN701ZhVSOna5q0f74aMTxp6S8Cjhlj8owxLcDLwDVDWy3lzzr66KcMUTokMSKE/3dzJsXVjfz4lc+GdErFznx+ml7EVSODJ0E/GTjltlzoWtfVdSJyQEReF5EJbutDRCRbRHaKyDcGU1nlHwYzjr6nMtNi+OnXZ7H1cBmPvZNLXXPbkPyd3QVVRAQHMG1cxJCUr5TVAiwq5y3gJWNMs4h8D3gOuNy1Lc0YUyQiE4EtIvK5Mea4+8EicidwJ0BqaqpFVVK+ajDTJPbHzYtTOXCqmqc+zGP99nxmJ0exZGIsSybGkZUWQ0TI4G+kyi6oZGFaDHadNEWNEJ4E/SLAveWe4lrXyRhT4ba4Hvi527Yi13OeiGwDFgDHuxz/NPA0QFZW1tD9Flc+oai6gajQQMKDrWpzdE9EeOy6uXxjQTI78yrYmVfBs9vzeeqDPGwCc5KjWDIxjuuzJjB5ANMbVje0cORMHavnjR+C2is1NDz51O0GpohIBs5g/23gO+47iEiSMabEtbgaOORaHwM0uH4BxAPLcPtCUKNTUZX13TV7YrMJyybHs2xyPOAcDXPvySp25VWwM6+SP35cwPrt+dy4aAI//tJU4sM9n8WrY9IUHVlTjSR9Bn1jTJuI3A1sAuzAs8aYHBF5BMg2xmwA7hGR1UAbUAmsdR0+A3hKRBw4rx/8zBhzcAjehxpBiqubSI0b3JDKAxUaZD/vS6C8rpnfvH+UF3ad5K+fFXPXZZO4fVkGoUF9j4m/u6CKQLswLyV6qKutlGXEGN/KpmRlZZns7GxvV0MNEWMMc/79XdZkpvDvq2d5uzqdjpfV8bO3c9l88AzjIkO4/yvTuHZBcq9TOa75fztoN4Y3f7BsGGuqVPdEZI8xJquv/YY2qapUFzWNbdQ1t/nc3auTEsL5w61Z7Mqr4D83HuL+1/bz7PZ8brs4rduZsAyGA4VnWbssffgrq9QgaNBXw6qwugEY2u6ag7F4Yhxv/mAZbx0o5ufvHOaBNz7vdf9LXGkipUYKDfpqWFk1jv5QstmEa+Ync9XspM57CroTHGAb8m6nSllNg74aVsXDcGOWVYICbGTEh3m7GkpZSkfZVMOqqLqR4AAbcWFB3q6KUqOSBn01rDqGVNYRKZXyDg36alhZPXmKUqp/NOirYTUUk6copTynF3LVsGlqbae8rkWD/mjSUg9v3Qt1pRCTBjHpEJ0GMRnO5TFxoKm+YaVBfwi1Owz/viGHVbPHdd72P5qNpJ47ygLtbfD67XD0XRi/AA6/DfVl5+8TFA5xkyBlEUxYDBMucn4p6BfBkNGgP4Se+vA4/73zBMXVjRr0OTeOvvZtHwWMgY33w5F34Ku/gotcM6g210H1Sag+AVUFUHUCSg/C/pdg9x+c+4SPhQmuL4GURZA0FwL1/4xVNOgPkYPFNfx68xGC7DY+yaugpc1BUMDovoQyEm7MUhbZ/jjs+SMs+/G5gA8QHA5jZzof7hztzuB/ahec2u18PvSWc5vYIX4qJM079xg3B0Iih+/9+BEN+kOgua2d+17dR1RoED9ZNY2fvH6A7BOVXDxpdLf2i6obsQmMiwrxdlXUUNr/Crz/CMy5Hq542LNjbHZnIB8359yXRF0pFO6Gkv3OR/4HcODlc8fEToTEmRA32Zkiip3kfA4fq+mhXmjQHwJPvneU3NO1PHNbFosnxvEvf/mcD4+Ua9CvamRcZAiB9tH9i8ev5X0Af/shpC+Ha9aBbRD/1uGJMP2rzkeHulIoOQAl+5xfBGW5zmsG7S3n9gkKh9gM55dCZDJEJLke45zPkUkQNHrvtNagb7E9Jyr5/QfHuSFrAlfMGAs452v98EgZD1413cu1866iau2j79fO5MArNztb3jf8GQI8n5DGY+GJMOVLzkcHRzucPQUVx6Eyz/lcccxZn6PvQWv9heUER0JYPITGQEi089n9ERLlTEUFhUFgmPM5KMz5hRIU5rzGYOt7zgVfpEHfQg0tbfzzq/tJigrl3742o3P9iqkJ/GLTYcpqm0mIGIIPwghRVN1IZlqMt6uhhsLZIvjzGmdAvPl1CB3GiWVsdmdX0Jh04IoLtzfVQO1pqC1xey6BhgporHI+qvJdr6sBD+cYsQWAPdj55RYQAgFBzmd7INgC3Z4D3JYDXA+781nszl9DHa9j0uDiH1l3brqhQd9CP3s7l4KKBl767pLzJt1eMcUZ9D86Wsa1C1O8WEPvaXcYTp9t0ou4/qjpLLywBppr4fZ3IMrH/o+HRDofCVP73tfhgOYaaKp23mPQUg8tddDS4Pa6HtqaoK35wuf2ZtdzKzhand1WW+pdy23OZ9Pu/HXS8dz5us15kVqD/sjw0dEynv/kBLcvy2DppLjzts0aH0lcWBAfHhm9Qf90TRNtDqPpHX904FUoPwI3vQ7jZnu7NoNjszl/pQznL5VhpkHfAmcbWvlfrx1gUkIYP1k17YLtNptwyZR4th8rx+EwvU7B56/2n6oGYEaSdrPzOxfdAWnLLuyGqXySdqOwwL+/lUNZXTO/vmE+IYHdX9xZMSWB8roWDpbUDHPtfMOuvArGBNmZkxzl7aooq4lowB9BNOgP0o5j5bz5WRE/XDmZuSk9/yRcPtXZXfPDo2U97uPPduZVkpUeq901lfIy/QQO0meutMX3Vkzsdb/EiBBmJEXy4ZHRF/Qr6po5fKaWJRNjvV0VpUY9DfqDVFjVQFxYEGHBfV8eWTElnj0nqqhvbhuGmvmOT/MrAVicEdfHnkqpoaZBf5AK+zGvJU+OAAAV+klEQVQpyIqpCbS2Gz45XjHEtfItO/MqCA20MzdF8/lKeZsG/UHqz6QgWekxhAbaR11ef1d+JVnpMZrPV8oH6KdwEIwxFFU1kuJhSz84wM6SibGjKq9fWd9C7ulalkzU1I5SvkCD/iCU17XQ3Obo112mK6YmUFDRwMmKhiGsme/4NN+ZytKLuEr5Bg36g1DUORPUGI+PWTE1ARg9XTd35lUSGmhnTrL/3uGo1EiiQX8QCqucrXVP0zsAE+PDSI4OHTUpnp15FWSlx4z6CWSU8hX6SRyEzpmg+hH0RYQVU+PZcbyC1nbHUFXNJ1RpPl8pn+NR0BeRVSJyWESOiciD3WxfKyJlIrLP9bijy/ZIESkUkf+yquK+oKi6kYiQACLdRtT0xIopCdQ1t/HZyeohqplv2NXZP1/z+Ur5ij6DvojYgXXAVcBM4EYR6W6gjVeMMfNdj/Vdtv1f4MNB19bHFFY1ktKPfH6HiyfHY7eJ36d4duZVEBJo63V4CqXU8PKkpb8IOGaMyTPGtAAvA9d4+gdEJBMYC7w7sCr6rqIqz/vou4sKDWT+hOgRdzG33WF4Y08hJWcbPdp/Z14FWWmxms9Xyod48mlMBk65LRe61nV1nYgcEJHXRWQCgIjYgF8B9w+6pj7GGENRted99LtaMSWBz4vOUlnf0vfOPqC0tolbntnFP7+2n5+8fqDP/asbWnS8HaV8kFVNsLeAdGPMXGAz8Jxr/Q+AjcaYwt4OFpE7RSRbRLLLykZG6/dsYyt1zW0DD/pT4zEGth8rt7hm1vvkeAVf/c129pyo4kszEvnoaHnneDo92ZVfiTHoRVylfIwnQb8ImOC2nOJa18kYU2GMaXYtrgcyXa+XAneLSAHwS+BWEflZ1z9gjHnaGJNljMlKSEjo51vwjsKOnjsDnP5vbko0UaGBPp3XdzgM/7XlKDet30lESAB/u3sZv71xIQkRwfzy3cMY0/NcoprPV8o3eRL0dwNTRCRDRIKAbwMb3HcQkSS3xdXAIQBjzE3GmFRjTDrOFM/zxpgLev+MROduzBpY0LfbhEsmx/PR0bJeg6e3VNa3sPZPu/nlu0f4+rzxbLj7EqaPiyQ0yM4PL5vEp/mVfHys54HjduZVkpmm/fOV8jV9fiKNMW3A3cAmnMH8VWNMjog8IiKrXbvdIyI5IrIfuAdYO1QV9hUdLf2B9N7psGRSHGdqmjvL8hXZBZVc/eRH7Myr4D++OZsnbphPuNvQ0TcuTmV8VAi/2tx9a7+6oYXc0zUs0aGUlfI5Hs2Ra4zZCGzssu6nbq8fAh7qo4w/AX/qdw19VFFVI6GBdmLG9K+PvruFqc7Ux54TVUyIHfiXh5XeP3SGO/97DykxofzlrouZ3c30hsEBdn50xRQe+svnbD1cyuXTx563/dOOfP4kDfpK+Rr97T1ARdUNJMeEIjLwSc6njY0gLMjO3pNVFtZs4FraHDzy94NMSgjjrR9d0m3A77AmM4XU2DH86t0jF7T2d+ZVEhxg0/HzlfJBGvQHaDDdNTsE2G3MmxDNnhO+EfRf+vQkJyoaeOjqGX3eZRxot3HvFVPIKa5hU87p87btzKsgMy2G4IDuJ4lXSnmPBv0BKhzgjVldZabFkHu61utTKNY2tfLk+0dZOjGOy6Z61oPqGwuSmZgQxuObj9DucLb2qxtaOHS6RrtqKuWjNOgPQH1zG9UNrQPuueNuYWoM7Q7D/kLvjsPz1Ad5VNa38NDV0z1OWdltwj99aSpHztTx9wPFgFs+X4O+Uj5Jg/4AdHTXHEzPnQ4LXBdzvTn42pmaJtZvz2P1vPH97lf/1TlJTB8XwRPvHaWt3cGufGc+f94Ezecr5Ys06A9Axzj6VqR3oscEMSkhzKt5/V+70jP/6yvT+n2szSb805VTyS+v583PitiZV8HCVM3nK+WrNOgPQFFnH/3BB31w5vX3nqzyyk1aR8/U8mr2KW5Zkj7gbqNfnjmWOclRPL75CAdLNJ+vlC/ToD8AhdWNBNltJIQHW1LewtQYqhtaySuvt6S8/njsnVzCggK4+/LJAy5DRLjvy1MpOdvkyufrIGtK+SoN+gNQWNXI+OgQbLaB99F3l5kWA8DeYU7x7Mqr4L1Dpdy1chKxYUGDKuuyqQmubprObqhKKd/k0R256nxFVY2W9NzpMCkhnMiQAPaerOL6rAl9H2ABYwz/+XYu4yJDuH1ZxqDLExGe/PZ8TlU2EhKo+XylfJW29AegqLqRlGjrhk2w2YQFqTHsPTF8PXg2fn6a/aeque/LUy0L0ikxY1iqQy8o5dM06PdTU2s7ZbXNlrb0wZnXP1JaS01Tq6XldqelzcHPN+UybWwE1y1MGfK/p5TyHRr0+6m4enDj6PckMy0GY2DfMPTX7xhu4cGrpmO36LqEUmpk0KDfT+duzLI26M+bEIVNGPL++mdqmnh88xGWTIzlsmkjY8IapZR1NOj3U+eMWRYH/YiQQKaOjRjSETeNMTzwxgGa29r5j2/OGdQIoUqpkUmDfj8VVTVitwnjIkMsLzszLYZ9J6txOIbmJq2XPj3FtsNlPLhqOpMSwofkbyilfJsG/X4qqm5kXGQIAXbrT93C1Bhqm9s4WlpnedknKxp49B8HuXhSHLcuTbe8fKXUyKBBv58KqxosT+106LxJy+IUT7vDcP9r+7GL8Ivr51l2U5lSauTRoN9PRVWNpFjcc6dDWtwYYsOCLL+Y+8z2PD4tqOTh1bMs73WklBpZNOj3Q2u7g9M1TZb33OkgIixMjbG0pX/kTC2/3HSEK2eO5bqFyZaVq5QamTTo98Pps004jPU9d9wtTIsmr6yeqvqWQZfV0ubgn17ZR0RIAP/ftdpbRymlQb9fOrtrWjgEQ1eZqc68/menBt/a/68tR8kpruE/vjmHeItGBFVKjWwa9PthqG7Mcjc3JZoAmww6r7//VDXrth3n2gXJrJo9zqLaKaVGOg36/dAxeUpStPV99DuEBtmZOT5yUIOvNbW2c9+r+0iMCObh1bMsrJ1SaqTToN8PhVUNJEYED/lUgAtTY9h3qpq2dseAjn/+kwKOl9Xz2HVziQoNtLZySqkRTYN+PxRVNw5paqfDwrQYGlvbyT1d2+9jG1raeOqDPJZPiWfFVB1bRyl1Pg36/VBU3UhyzNBdxO2wMNU589RAum7+9ycnqKhv4cdfmmJ1tZRSfkCDvoccDkNxdeOw3NyUHB3K2Mjgfk+f2NDSxlMfOlv5mWk6T61S6kIa9D1UWttMa7sZlvROx01ae/rZ0n/+kxNU1rfw4y9NHaKaKaVGOg36HiqqbgCG9sYsd5lpMZyqbKS0tsmj/eub23j6wzxWuCYoV0qp7ngU9EVklYgcFpFjIvJgN9vXikiZiOxzPe5wrU8Tkb2udTki8n2r38Bw6bgxa6jG3elqgesmrfcPlXq0/7lWvubylVI9C+hrBxGxA+uAK4FCYLeIbDDGHOyy6yvGmLu7rCsBlhpjmkUkHPjCdWyxFZUfTkM1eUpP5qVEsTA1moc35DApIZxFGT3n6J2t/ONcOjWBhanayldK9cyTlv4i4JgxJs8Y0wK8DFzjSeHGmBZjTLNrMdjDv+eTiqobiQ0LYkxQn9+Tlgiw23jmtotIiQnljud2c7iX7pvPf3KCqoZWbeUrpfrkSRBOBk65LRe61nV1nYgcEJHXRWRCx0oRmSAiB1xlPDYSW/ngbOkP97DEMWFBPH/7IkKD7Nz27Kedw0C4q3O18i+bltCZElJKqZ5Y1fJ+C0g3xswFNgPPdWwwxpxyrZ8M3CYiY7seLCJ3iki2iGSXlZVZVCVrFVU1DEvPna5SYsbw3O2LqG9p47ZnP6W64fzRN5//pICqhlbuvUJb+UqpvnkS9IuACW7LKa51nYwxFW5pnPVAZtdCXC38L4Dl3Wx72hiTZYzJSkjwvbtIjTHOG7O8NAHJ9HGR/OHWLE5WNnD7n3bT2NIOOFv5f/gwT1v5SimPeRL0dwNTRCRDRIKAbwMb3HcQkSS3xdXAIdf6FBEJdb2OAS4BDltR8eFUUd9CU6tj2C7idmfJxDievGE+n52q5kcv7aWt3cFzOwpcuXztl6+U8kyfVyWNMW0icjewCbADzxpjckTkESDbGLMBuEdEVgNtQCWw1nX4DOBXImIAAX5pjPl8CN7HkOoYXTNlGIZg6M1Vc5J4ZPUs/vffcvjJGwfYklvKymkJzJ8Q7dV6KaVGDo+6ohhjNgIbu6z7qdvrh4CHujluMzB3kHX0uo4LqL4wv+wtS9MprW3mt1uOAXCvtvKVUv0wPP0PR7jCquG9G7cv9105ldZ2Q0ubQ1v5Sql+0aDvgaKqRiJCAnxmbHoR4cGrpnu7GkqpEWjE3iw1nLzZc0cppaykQd8Dx8vqSY317kVcpZSyggb9PpyqbCC/vJ4lE+O8XRWllBo0Dfp92HbYOcrlZdN876YxpZTqLw36ffjgSBmpsWPIiA/zdlWUUmrQNOj3ormtnR3HK7hsWgIi4u3qKKXUoGnQ78Xu/CoaWtq5dKqmdpRS/mHUBP3H3snl1mc/7dcx2w6XEmS3sXSSXsRVSvmHURP039pfzIdHyjhUUuPxMduOlLF4YuywTZyilFJDbVQE/aLqxs7pDt/YU+jRMYVVDRwrrdPUjlLKr4yKoL8rrwKASQlh/HVfEa3tjj6P2XbYOZnLZdMSh7RuSik1nEZJ0K8kKjSQn6yaTnldCx8e6Xt2rm2Hy0iJCWVSgnbVVEr5j9ER9PMruCg9lsunJxIXFsTrfaR4Wtoc7DhezqVTtaumUsq/+H3QP1PTREFFA0smxhJot3HN/GTeO3SGqvqWHo/JLqikoaVdUztKKb/j90F/pyufvzjD2e1yTWYKre2GDfuLezxm25Eyguw2LtaumkopP+P3QX9XfiURwQHMHB8JwMzxkcxMiuSNvT2neLYdLuWijBjCgrWrplLKv/h/0M+rICs9BrvtXG5+TWYKBwrPcvh07QX7F1c3cuRMHZdN1dSOUsr/+HXQL6tt5nhZPYu7DIt8zfzxBNik29b+ua6a2j9fKeV//Drof5pfCcDijNjz1seFB7NyeiJ/2VtEW5c++9sOl5IcHcrkxPBhq6dSSg0Xvw76u/IrGBNkZ3Zy1AXb1mSmUF7XzEdHyzvXtbQ5+PhYOSu0q6ZSyk/5d9DPqyQzLYZA+4Vvc+W0RGK79NnPPlFJfUu7pnaUUn7Lb4N+ZX0Lh8/U9jjNYVCAjWvmj2fzwTNUNzj77H9wpIxAu7BscvxwVlUppYaN3wb9nvL57tZkptDS7uAtV5/9Dw6XkZUWS7h21VRK+Sm/Dfq78isIDrAxNyW6x31mjY9i+rgIXt9bRMnZRnJP12pqRynl1/w36OdVsjA1hqCA3t/imswU9p+q5pmP8gEdVVMp5d/8MuifbWjl0OkaFk/sObXT4RsLkgmwCc98nM+4yBCmjtWumkop/+WXQX93QSXGnBtvpzfx4cFcNi0RY9AJ0JVSfs8vg/6u/AqC7DYWpPacz3d3fVYKAJdP19SOUsq/eRT0RWSViBwWkWMi8mA329eKSJmI7HM97nCtny8in4hIjogcEJEbrH4D3dmVX8n8CdGEBNo92v/LM8fyxl1LuXLm2CGumVJKeVeffRNFxA6sA64ECoHdIrLBGHOwy66vGGPu7rKuAbjVGHNURMYDe0RkkzGm2orKd6e2qZUvis7yw5WTPT5GRMhM6zv/r5RSI50nLf1FwDFjTJ4xpgV4GbjGk8KNMUeMMUddr4uBUmBI+0TuOVGFw8N8vlJKjTaeBP1k4JTbcqFrXVfXuVI4r4vIhK4bRWQREAQcH1BNPbQrv5IAm7AwzbN8vlJKjSZWXch9C0g3xswFNgPPuW8UkSTgv4H/YYxxdD1YRO4UkWwRyS4r63vS8t7syqtgbkoUY4L0rlqllOrKk6BfBLi33FNc6zoZYyqMMc2uxfVAZsc2EYkE/gH8qzFmZ3d/wBjztDEmyxiTlZAw8OxPQ0sbBwrPXjB+vlJKKSdPgv5uYIqIZIhIEPBtYIP7Dq6WfIfVwCHX+iDgTeB5Y8zr1lS5Z3tPVNPmML2Ot6OUUqNZnzkQY0ybiNwNbALswLPGmBwReQTINsZsAO4RkdVAG1AJrHUd/i1gBRAnIh3r1hpj9ln7Npx25VdgtwlZ6Rr0lVKqOx4lvo0xG4GNXdb91O31Q8BD3Rz3Z+DPg6yjx3blVTJ7fKSOkqmUUj3wmztym1rb2XeqWvP5SinVC78J+jVNrayaPY7LpurQyEop1RO/yYMkRoTwmxsXeLsaSinl0/ympa+UUqpvGvSVUmoU0aCvlFKjiAZ9pZQaRTToK6XUKKJBXymlRhEN+kopNYpo0FdKqVFEjDHersN5RKQMODGIIuKBcouq48/0PHlGz5Nn9Dx5bqjOVZoxps8hCXwu6A+WiGQbY7K8XQ9fp+fJM3qePKPnyXPePlea3lFKqVFEg75SSo0i/hj0n/Z2BUYIPU+e0fPkGT1PnvPqufK7nL5SSqme+WNLXymlVA/8JuiLyCoROSwix0TkQW/Xx5eIyLMiUioiX7itixWRzSJy1PUc4806+gIRmSAiW0XkoIjkiMi9rvV6rtyISIiIfCoi+13n6f+41meIyC7XZ/AVEQnydl19gYjYReQzEfm7a9mr58kvgr6I2IF1wFXATOBGEZnp3Vr5lD8Bq7qsexB43xgzBXjftTzatQH/bIyZCSwBfuj6f6Tn6nzNwOXGmHnAfGCViCwBHgN+bYyZDFQB/9OLdfQl9wKH3Ja9ep78IugDi4Bjxpg8Y0wL8DJwjZfr5DOMMR8ClV1WXwM853r9HPCNYa2UDzLGlBhj9rpe1+L8oCaj5+o8xqnOtRjoehjgcuB11/pRf54ARCQF+Cqw3rUsePk8+UvQTwZOuS0Xutapno01xpS4Xp8GxnqzMr5GRNKBBcAu9FxdwJWy2AeUApuB40C1MabNtYt+Bp2eAH4COFzLcXj5PPlL0FeDYJxduLQbl4uIhANvAD82xtS4b9Nz5WSMaTfGzAdScP7Snu7lKvkcEfkaUGqM2ePturjzl4nRi4AJbssprnWqZ2dEJMkYUyIiSThbbKOeiATiDPgvGGP+4lqt56oHxphqEdkKLAWiRSTA1YrVzyAsA1aLyNVACBAJPImXz5O/tPR3A1NcV8WDgG8DG7xcJ1+3AbjN9fo24G9erItPcOVbnwEOGWMed9uk58qNiCSISLTrdShwJc7rH1uBNa7dRv15MsY8ZIxJMcak44xJW4wxN+Hl8+Q3N2e5vk2fAOzAs8aY//BylXyGiLwEXIZzdL8zwMPAX4FXgVSco5p+yxjT9WLvqCIilwAfAZ9zLgf7Lzjz+nquXERkLs4LkHacDcdXjTGPiMhEnJ0oYoHPgJuNMc3eq6nvEJHLgPuNMV/z9nnym6CvlFKqb/6S3lFKKeUBDfpKKTWKaNBXSqlRRIO+UkqNIhr0lVJqFNGgr5RSo4gGfaWUGkU06Cul1Cjy/wMVAzZVQDLhVQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sat, 30 Jun 2018 01:45:00 +0000\n",
      "(1, 30, 4)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Prediction on live data!\n",
    "starttime=time.time()\n",
    "while True:\n",
    "    now = time.time() \n",
    "    end = now - now % period\n",
    "    print time.strftime(\"%a, %d %b %Y %H:%M:%S +0000\", time.gmtime(end))\n",
    "    rt_x, prediction = predict()\n",
    "\n",
    "    current = rt_x[:,0]\n",
    "    prediction = prediction[0]\n",
    "\n",
    "    pyplot.plot(current, label='current')\n",
    "\n",
    "    # shift train predictions for plotting\n",
    "    predictPlot = np.empty_like(current)\n",
    "    predictPlot[:] = np.nan\n",
    "    predictPlot = np.append(predictPlot, prediction)\n",
    "\n",
    "    pyplot.plot(predictPlot, label='prediction')\n",
    "    pyplot.legend()\n",
    "    pyplot.show()\n",
    "\n",
    "    time.sleep(period - ((time.time() - starttime) % period))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}