predicting daily increasing determined cases in China with LSTM model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "file=r'chinancoron2020.csv'\n",
    "coro=pd.read_csv(file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>determined</th>\n",
       "      <th>Inc_deter</th>\n",
       "      <th>undetermined</th>\n",
       "      <th>death</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1/16/2020</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1/17/2020</td>\n",
       "      <td>62</td>\n",
       "      <td>58</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1/18/2020</td>\n",
       "      <td>121</td>\n",
       "      <td>59</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1/19/2020</td>\n",
       "      <td>198</td>\n",
       "      <td>77</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1/20/2020</td>\n",
       "      <td>291</td>\n",
       "      <td>93</td>\n",
       "      <td>54</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  determined  Inc_deter  undetermined  death\n",
       "0  1/16/2020           4          4             0      2\n",
       "1  1/17/2020          62         58             0      2\n",
       "2  1/18/2020         121         59             0      3\n",
       "3  1/19/2020         198         77             0      4\n",
       "4  1/20/2020         291         93            54      6"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coro.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>date</th>\n",
       "      <th>determined</th>\n",
       "      <th>Inc_deter</th>\n",
       "      <th>undetermined</th>\n",
       "      <th>death</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>2/4/2020</td>\n",
       "      <td>24363</td>\n",
       "      <td>3892</td>\n",
       "      <td>23260</td>\n",
       "      <td>491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>2/5/2020</td>\n",
       "      <td>28060</td>\n",
       "      <td>3697</td>\n",
       "      <td>24702</td>\n",
       "      <td>564</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>2/6/2020</td>\n",
       "      <td>31211</td>\n",
       "      <td>3151</td>\n",
       "      <td>26359</td>\n",
       "      <td>637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>2/7/2020</td>\n",
       "      <td>34598</td>\n",
       "      <td>3401</td>\n",
       "      <td>27657</td>\n",
       "      <td>723</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>2/8/2020</td>\n",
       "      <td>37251</td>\n",
       "      <td>2657</td>\n",
       "      <td>28942</td>\n",
       "      <td>812</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  determined  Inc_deter  undetermined  death\n",
       "19  2/4/2020       24363       3892         23260    491\n",
       "20  2/5/2020       28060       3697         24702    564\n",
       "21  2/6/2020       31211       3151         26359    637\n",
       "22  2/7/2020       34598       3401         27657    723\n",
       "23  2/8/2020       37251       2657         28942    812"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coro.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "coro['date'] =  pd.to_datetime(coro['date']) #, format='%d%b%Y:%H:%M:%S.%f')"
   ]
  },
  {
   "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>date</th>\n",
       "      <th>determined</th>\n",
       "      <th>Inc_deter</th>\n",
       "      <th>undetermined</th>\n",
       "      <th>death</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>2020-02-04</td>\n",
       "      <td>24363</td>\n",
       "      <td>3892</td>\n",
       "      <td>23260</td>\n",
       "      <td>491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>2020-02-05</td>\n",
       "      <td>28060</td>\n",
       "      <td>3697</td>\n",
       "      <td>24702</td>\n",
       "      <td>564</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>2020-02-06</td>\n",
       "      <td>31211</td>\n",
       "      <td>3151</td>\n",
       "      <td>26359</td>\n",
       "      <td>637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>2020-02-07</td>\n",
       "      <td>34598</td>\n",
       "      <td>3401</td>\n",
       "      <td>27657</td>\n",
       "      <td>723</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>2020-02-08</td>\n",
       "      <td>37251</td>\n",
       "      <td>2657</td>\n",
       "      <td>28942</td>\n",
       "      <td>812</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         date  determined  Inc_deter  undetermined  death\n",
       "19 2020-02-04       24363       3892         23260    491\n",
       "20 2020-02-05       28060       3697         24702    564\n",
       "21 2020-02-06       31211       3151         26359    637\n",
       "22 2020-02-07       34598       3401         27657    723\n",
       "23 2020-02-08       37251       2657         28942    812"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coro.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "coro.set_index('date', inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DatetimeIndex(['2020-01-16', '2020-01-17', '2020-01-18', '2020-01-19',\n",
       "               '2020-01-20', '2020-01-21', '2020-01-22', '2020-01-23',\n",
       "               '2020-01-24', '2020-01-25', '2020-01-26', '2020-01-27',\n",
       "               '2020-01-28', '2020-01-29', '2020-01-30', '2020-01-31',\n",
       "               '2020-02-01', '2020-02-02', '2020-02-03', '2020-02-04',\n",
       "               '2020-02-05', '2020-02-06', '2020-02-07', '2020-02-08'],\n",
       "              dtype='datetime64[ns]', name='date', freq=None)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coro.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x22aeec68da0>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "coro.plot() #color='b',linestyle='--',marker='o')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Using the data of daily increasing determined cases in China from Jan 21 through Jan 30 to build the Deep learning model "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from math import sqrt\n",
    "from numpy import concatenate\n",
    "from matplotlib import pyplot as plt\n",
    "from pandas import read_csv\n",
    "from pandas import DataFrame\n",
    "from pandas import concat\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import LSTM\n",
    "from keras.layers import Dropout\n",
    "import keras as K"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# convert an array of values into a dataset matrix\n",
    "def create_dataset(dataset, look_back=1):\n",
    "    dataX, dataY = [], []\n",
    "    for i in range(len(dataset)-look_back): #-1):\n",
    "        a = dataset[i:(i+look_back), 0]\n",
    "        dataX.append(a)\n",
    "        dataY.append(dataset[i + look_back, 0])\n",
    "    return np.array(dataX), np.array(dataY)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "det=coro['determined'].to_list()\n",
    "  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "inc_det=[]\n",
    "inc_det.append(4)\n",
    "for i in range(1,len(det)):\n",
    "    inc_det.append(det[i]-det[i-1])\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "inc_det"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "coro['Inc_deter']=inc_det"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>determined</th>\n",
       "      <th>Inc_deter</th>\n",
       "      <th>undetermined</th>\n",
       "      <th>death</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-01-16</th>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-01-17</th>\n",
       "      <td>62</td>\n",
       "      <td>58</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-01-18</th>\n",
       "      <td>121</td>\n",
       "      <td>59</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-01-19</th>\n",
       "      <td>198</td>\n",
       "      <td>77</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-01-20</th>\n",
       "      <td>291</td>\n",
       "      <td>93</td>\n",
       "      <td>54</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            determined  Inc_deter  undetermined  death\n",
       "date                                                  \n",
       "2020-01-16           4          4             0      2\n",
       "2020-01-17          62         58             0      2\n",
       "2020-01-18         121         59             0      3\n",
       "2020-01-19         198         77             0      4\n",
       "2020-01-20         291         93            54      6"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coro.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df=coro['Inc_deter'] #['2020-01-21':'2020-01-31']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.columns=['cases']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x22aeefd4dd8>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.plot(kind='bar')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "look_back=3\n",
    "splitratio=1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "date\n",
       "2020-02-04    3892\n",
       "2020-02-05    3697\n",
       "2020-02-06    3151\n",
       "2020-02-07    3401\n",
       "2020-02-08    2657\n",
       "Name: Inc_deter, dtype: int64"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset1=np.log(df.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "#if look_back==1:\n",
    "dataset1=dataset1.reshape(-1,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.38629436],\n",
       "       [4.06044301],\n",
       "       [4.07753744],\n",
       "       [4.34380542],\n",
       "       [4.53259949],\n",
       "       [5.00394631],\n",
       "       [4.87519732],\n",
       "       [5.55682806],\n",
       "       [6.12468339],\n",
       "       [6.5482191 ],\n",
       "       [6.65415252],\n",
       "       [7.48099216],\n",
       "       [6.75925527],\n",
       "       [7.75705114],\n",
       "       [7.59337419],\n",
       "       [7.6501687 ],\n",
       "       [7.85941315],\n",
       "       [7.94697136],\n",
       "       [8.08116578],\n",
       "       [8.26667844],\n",
       "       [8.21527696],\n",
       "       [8.05547514],\n",
       "       [8.13182479],\n",
       "       [7.88495295]])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "sc= MinMaxScaler(feature_range=(0,1))\n",
    "dataset = sc.fit_transform(dataset1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.        ],\n",
       "       [0.3886627 ],\n",
       "       [0.39114722],\n",
       "       [0.4298468 ],\n",
       "       [0.45728626],\n",
       "       [0.52579215],\n",
       "       [0.50707968],\n",
       "       [0.60614839],\n",
       "       [0.68868089],\n",
       "       [0.75023788],\n",
       "       [0.76563432],\n",
       "       [0.88580779],\n",
       "       [0.78091003],\n",
       "       [0.9259304 ],\n",
       "       [0.90214147],\n",
       "       [0.91039603],\n",
       "       [0.94080777],\n",
       "       [0.95353354],\n",
       "       [0.97303745],\n",
       "       [1.        ],\n",
       "       [0.99252927],\n",
       "       [0.96930356],\n",
       "       [0.98040027],\n",
       "       [0.94451974]])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(24, 1) (24, 1)\n"
     ]
    }
   ],
   "source": [
    "# split into train and test sets\n",
    "train_size = int(len(dataset) * splitratio)\n",
    "test_size = train_size \n",
    "train, test = dataset[0:train_size,:], dataset[0:train_size,:]\n",
    "print(train.shape, test.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "24"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# reshape into X=t and Y=t+1\n",
    " #   look_back = 24\n",
    "trainX, trainY = create_dataset(train, look_back)\n",
    "testX, testY = create_dataset(test, look_back)\n",
    "all_X,all_Y=create_dataset(dataset,look_back)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.        , 0.3886627 , 0.39114722],\n",
       "       [0.3886627 , 0.39114722, 0.4298468 ],\n",
       "       [0.39114722, 0.4298468 , 0.45728626],\n",
       "       [0.4298468 , 0.45728626, 0.52579215],\n",
       "       [0.45728626, 0.52579215, 0.50707968],\n",
       "       [0.52579215, 0.50707968, 0.60614839],\n",
       "       [0.50707968, 0.60614839, 0.68868089],\n",
       "       [0.60614839, 0.68868089, 0.75023788],\n",
       "       [0.68868089, 0.75023788, 0.76563432],\n",
       "       [0.75023788, 0.76563432, 0.88580779],\n",
       "       [0.76563432, 0.88580779, 0.78091003],\n",
       "       [0.88580779, 0.78091003, 0.9259304 ],\n",
       "       [0.78091003, 0.9259304 , 0.90214147],\n",
       "       [0.9259304 , 0.90214147, 0.91039603],\n",
       "       [0.90214147, 0.91039603, 0.94080777],\n",
       "       [0.91039603, 0.94080777, 0.95353354],\n",
       "       [0.94080777, 0.95353354, 0.97303745],\n",
       "       [0.95353354, 0.97303745, 1.        ],\n",
       "       [0.97303745, 1.        , 0.99252927],\n",
       "       [1.        , 0.99252927, 0.96930356],\n",
       "       [0.99252927, 0.96930356, 0.98040027]])"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainX"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.4298468 , 0.45728626, 0.52579215, 0.50707968, 0.60614839,\n",
       "       0.68868089, 0.75023788, 0.76563432, 0.88580779, 0.78091003,\n",
       "       0.9259304 , 0.90214147, 0.91039603, 0.94080777, 0.95353354,\n",
       "       0.97303745, 1.        , 0.99252927, 0.96930356, 0.98040027,\n",
       "       0.94451974])"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainY"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# reshape input to be [samples, time steps, features]\n",
    "trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))\n",
    "testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))\n",
    "all_X = np.reshape(all_X, (all_X.shape[0], 1, all_X.shape[1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[0.        , 0.3886627 , 0.39114722]],\n",
       "\n",
       "       [[0.3886627 , 0.39114722, 0.4298468 ]],\n",
       "\n",
       "       [[0.39114722, 0.4298468 , 0.45728626]],\n",
       "\n",
       "       [[0.4298468 , 0.45728626, 0.52579215]],\n",
       "\n",
       "       [[0.45728626, 0.52579215, 0.50707968]],\n",
       "\n",
       "       [[0.52579215, 0.50707968, 0.60614839]],\n",
       "\n",
       "       [[0.50707968, 0.60614839, 0.68868089]],\n",
       "\n",
       "       [[0.60614839, 0.68868089, 0.75023788]],\n",
       "\n",
       "       [[0.68868089, 0.75023788, 0.76563432]],\n",
       "\n",
       "       [[0.75023788, 0.76563432, 0.88580779]],\n",
       "\n",
       "       [[0.76563432, 0.88580779, 0.78091003]],\n",
       "\n",
       "       [[0.88580779, 0.78091003, 0.9259304 ]],\n",
       "\n",
       "       [[0.78091003, 0.9259304 , 0.90214147]],\n",
       "\n",
       "       [[0.9259304 , 0.90214147, 0.91039603]],\n",
       "\n",
       "       [[0.90214147, 0.91039603, 0.94080777]],\n",
       "\n",
       "       [[0.91039603, 0.94080777, 0.95353354]],\n",
       "\n",
       "       [[0.94080777, 0.95353354, 0.97303745]],\n",
       "\n",
       "       [[0.95353354, 0.97303745, 1.        ]],\n",
       "\n",
       "       [[0.97303745, 1.        , 0.99252927]],\n",
       "\n",
       "       [[1.        , 0.99252927, 0.96930356]],\n",
       "\n",
       "       [[0.99252927, 0.96930356, 0.98040027]]])"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainX"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(21, 1, 3) (21, 1, 3)\n"
     ]
    }
   ],
   "source": [
    "print(trainX.shape, testX.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING: Logging before flag parsing goes to stderr.\n",
      "W0209 11:58:33.585304 15784 deprecation_wrapper.py:119] From C:\\ProgramData\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:74: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.\n",
      "\n",
      "W0209 11:58:34.556283 15784 deprecation_wrapper.py:119] From C:\\ProgramData\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:517: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.\n",
      "\n",
      "W0209 11:58:34.937251 15784 deprecation_wrapper.py:119] From C:\\ProgramData\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:4138: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.\n",
      "\n",
      "W0209 11:58:37.439876 15784 deprecation_wrapper.py:119] From C:\\ProgramData\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:133: The name tf.placeholder_with_default is deprecated. Please use tf.compat.v1.placeholder_with_default instead.\n",
      "\n",
      "W0209 11:58:37.461864 15784 deprecation.py:506] From C:\\ProgramData\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:3445: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.\n",
      "W0209 11:58:38.380442 15784 deprecation_wrapper.py:119] From C:\\ProgramData\\Anaconda3\\lib\\site-packages\\keras\\optimizers.py:790: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_1 (LSTM)                (None, 1, 50)             10800     \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 1, 50)             0         \n",
      "_________________________________________________________________\n",
      "lstm_2 (LSTM)                (None, 1, 130)            94120     \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 1, 130)            0         \n",
      "_________________________________________________________________\n",
      "lstm_3 (LSTM)                (None, 30)                19320     \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 1)                 31        \n",
      "=================================================================\n",
      "Total params: 124,271\n",
      "Trainable params: 124,271\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# create and fit the LSTM network\n",
    "model = Sequential()\n",
    "model.add(LSTM(units=50, return_sequences=True,input_shape=(trainX.shape[1], trainX.shape[2])))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(LSTM(units=130, return_sequences=True))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(LSTM(units=30))\n",
    "#model.add(Dropout(0.2))\n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mean_squared_error', optimizer='adam')\n",
    "print(model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "W0209 11:58:39.484322 15784 deprecation.py:323] From C:\\Users\\Changbing Yang\\AppData\\Roaming\\Python\\Python37\\site-packages\\tensorflow\\python\\ops\\math_grad.py:1250: add_dispatch_support.<locals>.wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n",
      "W0209 11:58:42.577618 15784 deprecation_wrapper.py:119] From C:\\ProgramData\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:986: The name tf.assign_add is deprecated. Please use tf.compat.v1.assign_add instead.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 18 samples, validate on 3 samples\n",
      "Epoch 1/200\n",
      " - 8s - loss: 0.6092 - val_loss: 0.8206\n",
      "Epoch 2/200\n",
      " - 0s - loss: 0.5114 - val_loss: 0.6576\n",
      "Epoch 3/200\n",
      " - 0s - loss: 0.3759 - val_loss: 0.4010\n",
      "Epoch 4/200\n",
      " - 0s - loss: 0.1755 - val_loss: 0.1009\n",
      "Epoch 5/200\n",
      " - 0s - loss: 0.0288 - val_loss: 0.0171\n",
      "Epoch 6/200\n",
      " - 0s - loss: 0.0253 - val_loss: 0.0200\n",
      "Epoch 7/200\n",
      " - 0s - loss: 0.0089 - val_loss: 0.0016\n",
      "Epoch 8/200\n",
      " - 0s - loss: 0.0084 - val_loss: 0.0041\n",
      "Epoch 9/200\n",
      " - 0s - loss: 0.0076 - val_loss: 1.8984e-04\n",
      "Epoch 10/200\n",
      " - 0s - loss: 0.0087 - val_loss: 2.0135e-04\n",
      "Epoch 11/200\n",
      " - 0s - loss: 0.0068 - val_loss: 1.9779e-04\n",
      "Epoch 12/200\n",
      " - 0s - loss: 0.0066 - val_loss: 2.1156e-04\n",
      "Epoch 13/200\n",
      " - 0s - loss: 0.0047 - val_loss: 1.9547e-04\n",
      "Epoch 14/200\n",
      " - 0s - loss: 0.0048 - val_loss: 3.1680e-04\n",
      "Epoch 15/200\n",
      " - 0s - loss: 0.0062 - val_loss: 9.0429e-04\n",
      "Epoch 16/200\n",
      " - 0s - loss: 0.0059 - val_loss: 5.7916e-04\n",
      "Epoch 17/200\n",
      " - 0s - loss: 0.0099 - val_loss: 0.0016\n",
      "Epoch 18/200\n",
      " - 0s - loss: 0.0081 - val_loss: 5.2305e-04\n",
      "Epoch 19/200\n",
      " - 0s - loss: 0.0073 - val_loss: 3.3870e-04\n",
      "Epoch 20/200\n",
      " - 0s - loss: 0.0096 - val_loss: 0.0040\n",
      "Epoch 21/200\n",
      " - 0s - loss: 0.0097 - val_loss: 2.6320e-04\n",
      "Epoch 22/200\n",
      " - 0s - loss: 0.0072 - val_loss: 2.6333e-04\n",
      "Epoch 23/200\n",
      " - 0s - loss: 0.0043 - val_loss: 7.3644e-04\n",
      "Epoch 24/200\n",
      " - 0s - loss: 0.0040 - val_loss: 2.0032e-04\n",
      "Epoch 25/200\n",
      " - 0s - loss: 0.0038 - val_loss: 4.6420e-04\n",
      "Epoch 26/200\n",
      " - 0s - loss: 0.0050 - val_loss: 3.2271e-04\n",
      "Epoch 27/200\n",
      " - 0s - loss: 0.0062 - val_loss: 1.8031e-04\n",
      "Epoch 28/200\n",
      " - 0s - loss: 0.0094 - val_loss: 0.0021\n",
      "Epoch 29/200\n",
      " - 0s - loss: 0.0056 - val_loss: 4.1783e-04\n",
      "Epoch 30/200\n",
      " - 0s - loss: 0.0095 - val_loss: 0.0026\n",
      "Epoch 31/200\n",
      " - 0s - loss: 0.0045 - val_loss: 1.7563e-04\n",
      "Epoch 32/200\n",
      " - 0s - loss: 0.0055 - val_loss: 3.4638e-04\n",
      "Epoch 33/200\n",
      " - 0s - loss: 0.0078 - val_loss: 0.0031\n",
      "Epoch 34/200\n",
      " - 0s - loss: 0.0058 - val_loss: 5.1608e-04\n",
      "Epoch 35/200\n",
      " - 0s - loss: 0.0043 - val_loss: 2.2445e-04\n",
      "Epoch 36/200\n",
      " - 0s - loss: 0.0059 - val_loss: 0.0021\n",
      "Epoch 37/200\n",
      " - 0s - loss: 0.0070 - val_loss: 0.0016\n",
      "Epoch 38/200\n",
      " - 0s - loss: 0.0048 - val_loss: 2.4943e-04\n",
      "Epoch 39/200\n",
      " - 0s - loss: 0.0037 - val_loss: 0.0066\n",
      "Epoch 40/200\n",
      " - 0s - loss: 0.0057 - val_loss: 6.1578e-04\n",
      "Epoch 41/200\n",
      " - 0s - loss: 0.0036 - val_loss: 1.9966e-04\n",
      "Epoch 42/200\n",
      " - 0s - loss: 0.0072 - val_loss: 0.0045\n",
      "Epoch 43/200\n",
      " - 0s - loss: 0.0041 - val_loss: 7.2563e-04\n",
      "Epoch 44/200\n",
      " - 0s - loss: 0.0074 - val_loss: 0.0022\n",
      "Epoch 45/200\n",
      " - 0s - loss: 0.0061 - val_loss: 0.0120\n",
      "Epoch 46/200\n",
      " - 0s - loss: 0.0063 - val_loss: 1.7224e-04\n",
      "Epoch 47/200\n",
      " - 0s - loss: 0.0058 - val_loss: 2.4149e-04\n",
      "Epoch 48/200\n",
      " - 0s - loss: 0.0122 - val_loss: 0.0022\n",
      "Epoch 49/200\n",
      " - 0s - loss: 0.0059 - val_loss: 1.9504e-04\n",
      "Epoch 50/200\n",
      " - 0s - loss: 0.0050 - val_loss: 0.0077\n",
      "Epoch 51/200\n",
      " - 0s - loss: 0.0059 - val_loss: 0.0029\n",
      "Epoch 52/200\n",
      " - 0s - loss: 0.0051 - val_loss: 7.7694e-04\n",
      "Epoch 53/200\n",
      " - 0s - loss: 0.0058 - val_loss: 0.0059\n",
      "Epoch 54/200\n",
      " - 0s - loss: 0.0030 - val_loss: 4.8668e-04\n",
      "Epoch 55/200\n",
      " - 0s - loss: 0.0041 - val_loss: 0.0018\n",
      "Epoch 56/200\n",
      " - 0s - loss: 0.0032 - val_loss: 0.0035\n",
      "Epoch 57/200\n",
      " - 0s - loss: 0.0079 - val_loss: 5.9729e-04\n",
      "Epoch 58/200\n",
      " - 0s - loss: 0.0088 - val_loss: 0.0097\n",
      "Epoch 59/200\n",
      " - 0s - loss: 0.0075 - val_loss: 0.0011\n",
      "Epoch 60/200\n",
      " - 0s - loss: 0.0076 - val_loss: 4.8975e-04\n",
      "Epoch 61/200\n",
      " - 0s - loss: 0.0048 - val_loss: 0.0059\n",
      "Epoch 62/200\n",
      " - 0s - loss: 0.0046 - val_loss: 2.2456e-04\n",
      "Epoch 63/200\n",
      " - 0s - loss: 0.0027 - val_loss: 0.0017\n",
      "Epoch 64/200\n",
      " - 0s - loss: 0.0059 - val_loss: 0.0012\n",
      "Epoch 65/200\n",
      " - 0s - loss: 0.0059 - val_loss: 0.0036\n",
      "Epoch 66/200\n",
      " - 0s - loss: 0.0096 - val_loss: 9.4489e-04\n",
      "Epoch 67/200\n",
      " - 0s - loss: 0.0026 - val_loss: 3.3945e-04\n",
      "Epoch 68/200\n",
      " - 0s - loss: 0.0040 - val_loss: 4.7784e-04\n",
      "Epoch 69/200\n",
      " - 0s - loss: 0.0045 - val_loss: 0.0130\n",
      "Epoch 70/200\n",
      " - 0s - loss: 0.0071 - val_loss: 0.0012\n",
      "Epoch 71/200\n",
      " - 0s - loss: 0.0068 - val_loss: 7.4204e-04\n",
      "Epoch 72/200\n",
      " - 0s - loss: 0.0040 - val_loss: 0.0209\n",
      "Epoch 73/200\n",
      " - 0s - loss: 0.0099 - val_loss: 0.0013\n",
      "Epoch 74/200\n",
      " - 0s - loss: 0.0060 - val_loss: 1.8014e-04\n",
      "Epoch 75/200\n",
      " - 0s - loss: 0.0055 - val_loss: 0.0072\n",
      "Epoch 76/200\n",
      " - 0s - loss: 0.0034 - val_loss: 0.0031\n",
      "Epoch 77/200\n",
      " - 0s - loss: 0.0067 - val_loss: 1.7770e-04\n",
      "Epoch 78/200\n",
      " - 0s - loss: 0.0071 - val_loss: 0.0037\n",
      "Epoch 79/200\n",
      " - 0s - loss: 0.0063 - val_loss: 0.0016\n",
      "Epoch 80/200\n",
      " - 0s - loss: 0.0056 - val_loss: 0.0025\n",
      "Epoch 81/200\n",
      " - 0s - loss: 0.0037 - val_loss: 0.0025\n",
      "Epoch 82/200\n",
      " - 0s - loss: 0.0032 - val_loss: 4.2077e-04\n",
      "Epoch 83/200\n",
      " - 0s - loss: 0.0031 - val_loss: 0.0052\n",
      "Epoch 84/200\n",
      " - 0s - loss: 0.0051 - val_loss: 1.7112e-04\n",
      "Epoch 85/200\n",
      " - 0s - loss: 0.0043 - val_loss: 0.0026\n",
      "Epoch 86/200\n",
      " - 0s - loss: 0.0070 - val_loss: 0.0031\n",
      "Epoch 87/200\n",
      " - 0s - loss: 0.0031 - val_loss: 0.0041\n",
      "Epoch 88/200\n",
      " - 0s - loss: 0.0035 - val_loss: 0.0029\n",
      "Epoch 89/200\n",
      " - 0s - loss: 0.0028 - val_loss: 0.0025\n",
      "Epoch 90/200\n",
      " - 0s - loss: 0.0035 - val_loss: 0.0031\n",
      "Epoch 91/200\n",
      " - 0s - loss: 0.0048 - val_loss: 0.0116\n",
      "Epoch 92/200\n",
      " - 0s - loss: 0.0030 - val_loss: 3.8835e-04\n",
      "Epoch 93/200\n",
      " - 0s - loss: 0.0082 - val_loss: 0.0035\n",
      "Epoch 94/200\n",
      " - 0s - loss: 0.0052 - val_loss: 0.0108\n",
      "Epoch 95/200\n",
      " - 0s - loss: 0.0059 - val_loss: 0.0018\n",
      "Epoch 96/200\n",
      " - 0s - loss: 0.0045 - val_loss: 0.0015\n",
      "Epoch 97/200\n",
      " - 0s - loss: 0.0070 - val_loss: 0.0025\n",
      "Epoch 98/200\n",
      " - 0s - loss: 0.0068 - val_loss: 1.7737e-04\n",
      "Epoch 99/200\n",
      " - 0s - loss: 0.0042 - val_loss: 0.0085\n",
      "Epoch 100/200\n",
      " - 0s - loss: 0.0080 - val_loss: 5.3503e-04\n",
      "Epoch 101/200\n",
      " - 0s - loss: 0.0029 - val_loss: 7.7575e-04\n",
      "Epoch 102/200\n",
      " - 0s - loss: 0.0024 - val_loss: 0.0020\n",
      "Epoch 103/200\n",
      " - 0s - loss: 0.0046 - val_loss: 0.0019\n",
      "Epoch 104/200\n",
      " - 0s - loss: 0.0050 - val_loss: 0.0047\n",
      "Epoch 105/200\n",
      " - 0s - loss: 0.0070 - val_loss: 1.8304e-04\n",
      "Epoch 106/200\n",
      " - 0s - loss: 0.0054 - val_loss: 0.0045\n",
      "Epoch 107/200\n",
      " - 0s - loss: 0.0041 - val_loss: 0.0045\n",
      "Epoch 108/200\n",
      " - 0s - loss: 0.0039 - val_loss: 0.0040\n",
      "Epoch 109/200\n",
      " - 0s - loss: 0.0046 - val_loss: 0.0015\n",
      "Epoch 110/200\n",
      " - 0s - loss: 0.0025 - val_loss: 6.7162e-04\n",
      "Epoch 111/200\n",
      " - 0s - loss: 0.0028 - val_loss: 1.6329e-04\n",
      "Epoch 112/200\n",
      " - 0s - loss: 0.0078 - val_loss: 2.0383e-04\n",
      "Epoch 113/200\n",
      " - 0s - loss: 0.0037 - val_loss: 0.0126\n",
      "Epoch 114/200\n",
      " - 0s - loss: 0.0054 - val_loss: 0.0019\n",
      "Epoch 115/200\n",
      " - 0s - loss: 0.0042 - val_loss: 1.8058e-04\n",
      "Epoch 116/200\n",
      " - 0s - loss: 0.0043 - val_loss: 0.0120\n",
      "Epoch 117/200\n",
      " - 0s - loss: 0.0050 - val_loss: 2.4390e-04\n",
      "Epoch 118/200\n",
      " - 0s - loss: 0.0030 - val_loss: 1.9209e-04\n",
      "Epoch 119/200\n",
      " - 0s - loss: 0.0044 - val_loss: 0.0025\n",
      "Epoch 120/200\n",
      " - 0s - loss: 0.0050 - val_loss: 3.7273e-04\n",
      "Epoch 121/200\n",
      " - 0s - loss: 0.0054 - val_loss: 0.0055\n",
      "Epoch 122/200\n",
      " - 0s - loss: 0.0034 - val_loss: 1.6285e-04\n",
      "Epoch 123/200\n",
      " - 0s - loss: 0.0040 - val_loss: 3.2575e-04\n",
      "Epoch 124/200\n",
      " - 0s - loss: 0.0074 - val_loss: 0.0029\n",
      "Epoch 125/200\n",
      " - 0s - loss: 0.0065 - val_loss: 9.2160e-04\n",
      "Epoch 126/200\n",
      " - 0s - loss: 0.0041 - val_loss: 4.5676e-04\n",
      "Epoch 127/200\n",
      " - 0s - loss: 0.0046 - val_loss: 0.0049\n",
      "Epoch 128/200\n",
      " - 0s - loss: 0.0035 - val_loss: 0.0020\n",
      "Epoch 129/200\n",
      " - 0s - loss: 0.0042 - val_loss: 0.0020\n",
      "Epoch 130/200\n",
      " - 0s - loss: 0.0040 - val_loss: 3.9163e-04\n",
      "Epoch 131/200\n",
      " - 0s - loss: 0.0032 - val_loss: 0.0049\n",
      "Epoch 132/200\n",
      " - 0s - loss: 0.0032 - val_loss: 0.0087\n",
      "Epoch 133/200\n",
      " - 0s - loss: 0.0038 - val_loss: 0.0031\n",
      "Epoch 134/200\n",
      " - 0s - loss: 0.0041 - val_loss: 0.0032\n",
      "Epoch 135/200\n",
      " - 0s - loss: 0.0053 - val_loss: 9.7920e-04\n",
      "Epoch 136/200\n",
      " - 0s - loss: 0.0063 - val_loss: 0.0021\n",
      "Epoch 137/200\n",
      " - 0s - loss: 0.0028 - val_loss: 2.0635e-04\n",
      "Epoch 138/200\n",
      " - 0s - loss: 0.0088 - val_loss: 0.0027\n",
      "Epoch 139/200\n",
      " - 0s - loss: 0.0041 - val_loss: 0.0020\n",
      "Epoch 140/200\n",
      " - 0s - loss: 0.0042 - val_loss: 1.8855e-04\n",
      "Epoch 141/200\n",
      " - 0s - loss: 0.0049 - val_loss: 0.0104\n",
      "Epoch 142/200\n",
      " - 0s - loss: 0.0068 - val_loss: 2.0008e-04\n",
      "Epoch 143/200\n",
      " - 0s - loss: 0.0055 - val_loss: 5.8759e-04\n",
      "Epoch 144/200\n",
      " - 0s - loss: 0.0056 - val_loss: 0.0046\n",
      "Epoch 145/200\n",
      " - 0s - loss: 0.0035 - val_loss: 0.0018\n",
      "Epoch 146/200\n",
      " - 0s - loss: 0.0034 - val_loss: 5.8684e-04\n",
      "Epoch 147/200\n",
      " - 0s - loss: 0.0034 - val_loss: 0.0021\n",
      "Epoch 148/200\n",
      " - 0s - loss: 0.0048 - val_loss: 9.3751e-04\n",
      "Epoch 149/200\n",
      " - 0s - loss: 0.0046 - val_loss: 0.0031\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 150/200\n",
      " - 0s - loss: 0.0044 - val_loss: 0.0037\n",
      "Epoch 151/200\n",
      " - 0s - loss: 0.0031 - val_loss: 1.7671e-04\n",
      "Epoch 152/200\n",
      " - 0s - loss: 0.0038 - val_loss: 7.5050e-04\n",
      "Epoch 153/200\n",
      " - 0s - loss: 0.0039 - val_loss: 0.0017\n",
      "Epoch 154/200\n",
      " - 0s - loss: 0.0024 - val_loss: 0.0015\n",
      "Epoch 155/200\n",
      " - 0s - loss: 0.0040 - val_loss: 0.0014\n",
      "Epoch 156/200\n",
      " - 0s - loss: 0.0029 - val_loss: 0.0016\n",
      "Epoch 157/200\n",
      " - 0s - loss: 0.0041 - val_loss: 0.0032\n",
      "Epoch 158/200\n",
      " - 0s - loss: 0.0045 - val_loss: 0.0013\n",
      "Epoch 159/200\n",
      " - 0s - loss: 0.0035 - val_loss: 0.0014\n",
      "Epoch 160/200\n",
      " - 0s - loss: 0.0034 - val_loss: 0.0033\n",
      "Epoch 161/200\n",
      " - 0s - loss: 0.0038 - val_loss: 0.0026\n",
      "Epoch 162/200\n",
      " - 0s - loss: 0.0036 - val_loss: 5.0386e-04\n",
      "Epoch 163/200\n",
      " - 0s - loss: 0.0052 - val_loss: 0.0021\n",
      "Epoch 164/200\n",
      " - 0s - loss: 0.0040 - val_loss: 5.4462e-04\n",
      "Epoch 165/200\n",
      " - 0s - loss: 0.0053 - val_loss: 0.0020\n",
      "Epoch 166/200\n",
      " - 0s - loss: 0.0030 - val_loss: 0.0015\n",
      "Epoch 167/200\n",
      " - 0s - loss: 0.0027 - val_loss: 0.0021\n",
      "Epoch 168/200\n",
      " - 0s - loss: 0.0033 - val_loss: 0.0026\n",
      "Epoch 169/200\n",
      " - 0s - loss: 0.0029 - val_loss: 4.0251e-04\n",
      "Epoch 170/200\n",
      " - 0s - loss: 0.0032 - val_loss: 0.0044\n",
      "Epoch 171/200\n",
      " - 0s - loss: 0.0023 - val_loss: 0.0021\n",
      "Epoch 172/200\n",
      " - 0s - loss: 0.0053 - val_loss: 0.0036\n",
      "Epoch 173/200\n",
      " - 0s - loss: 0.0048 - val_loss: 3.1675e-04\n",
      "Epoch 174/200\n",
      " - 0s - loss: 0.0038 - val_loss: 0.0044\n",
      "Epoch 175/200\n",
      " - 0s - loss: 0.0049 - val_loss: 4.1073e-04\n",
      "Epoch 176/200\n",
      " - 0s - loss: 0.0053 - val_loss: 0.0015\n",
      "Epoch 177/200\n",
      " - 0s - loss: 0.0051 - val_loss: 0.0030\n",
      "Epoch 178/200\n",
      " - 0s - loss: 0.0065 - val_loss: 1.6123e-04\n",
      "Epoch 179/200\n",
      " - 0s - loss: 0.0037 - val_loss: 0.0041\n",
      "Epoch 180/200\n",
      " - 0s - loss: 0.0037 - val_loss: 9.9079e-04\n",
      "Epoch 181/200\n",
      " - 0s - loss: 0.0042 - val_loss: 7.7863e-04\n",
      "Epoch 182/200\n",
      " - 0s - loss: 0.0041 - val_loss: 0.0052\n",
      "Epoch 183/200\n",
      " - 0s - loss: 0.0029 - val_loss: 1.6546e-04\n",
      "Epoch 184/200\n",
      " - 0s - loss: 0.0031 - val_loss: 2.2264e-04\n",
      "Epoch 185/200\n",
      " - 0s - loss: 0.0037 - val_loss: 0.0053\n",
      "Epoch 186/200\n",
      " - 0s - loss: 0.0055 - val_loss: 2.8053e-04\n",
      "Epoch 187/200\n",
      " - 0s - loss: 0.0026 - val_loss: 5.2735e-04\n",
      "Epoch 188/200\n",
      " - 0s - loss: 0.0038 - val_loss: 0.0018\n",
      "Epoch 189/200\n",
      " - 0s - loss: 0.0039 - val_loss: 7.0026e-04\n",
      "Epoch 190/200\n",
      " - 0s - loss: 0.0025 - val_loss: 0.0013\n",
      "Epoch 191/200\n",
      " - 0s - loss: 0.0048 - val_loss: 0.0070\n",
      "Epoch 192/200\n",
      " - 0s - loss: 0.0041 - val_loss: 1.6458e-04\n",
      "Epoch 193/200\n",
      " - 0s - loss: 0.0038 - val_loss: 0.0013\n",
      "Epoch 194/200\n",
      " - 0s - loss: 0.0028 - val_loss: 0.0029\n",
      "Epoch 195/200\n",
      " - 0s - loss: 0.0029 - val_loss: 5.5729e-04\n",
      "Epoch 196/200\n",
      " - 0s - loss: 0.0021 - val_loss: 0.0017\n",
      "Epoch 197/200\n",
      " - 0s - loss: 0.0020 - val_loss: 0.0013\n",
      "Epoch 198/200\n",
      " - 0s - loss: 0.0036 - val_loss: 5.3774e-04\n",
      "Epoch 199/200\n",
      " - 0s - loss: 0.0023 - val_loss: 8.6770e-04\n",
      "Epoch 200/200\n",
      " - 0s - loss: 0.0019 - val_loss: 9.9385e-04\n"
     ]
    }
   ],
   "source": [
    "history=model.fit(trainX, trainY, epochs=200, batch_size=2, validation_split=0.1,verbose=2)  #validation_split=0.05,"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot history\n",
    "plt.plot(history.history['loss'], label='train')\n",
    "plt.plot(history.history['val_loss'], label='cros_validation')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# make predictions\n",
    "trainPredict = model.predict(trainX)\n",
    "testPredict = model.predict(testX)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(21, 1)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainPredict.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(21,)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainY.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "y1=trainPredict.reshape(trainPredict.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x22aff8685f8>]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(y1,trainY,'ro')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# invert predictions\n",
    "#trainPred = sc.inverse_transform(trainPredict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Score: 0.27 RMSE\n",
      "Test Score: 0.27 RMSE\n"
     ]
    }
   ],
   "source": [
    "# invert predictions\n",
    "trainPredict = sc.inverse_transform(trainPredict)\n",
    "trainY = sc.inverse_transform([trainY])\n",
    "testPredict = sc.inverse_transform(testPredict)\n",
    "testY = sc.inverse_transform([testY])\n",
    "# calculate root mean squared error\n",
    "trainScore = sqrt(mean_squared_error(trainY[0], trainPredict[:,0]))\n",
    "print('Train Score: %.2f RMSE' % (trainScore))\n",
    "testScore = sqrt(mean_squared_error(testY[0], testPredict[:,0]))\n",
    "print('Test Score: %.2f RMSE' % (testScore))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot baseline and predictions\n",
    "fig = plt.figure(figsize=(10, 8))\n",
    "plt.plot(trainY[0,:],trainPredict[:,0],'bo',label='Train')\n",
    "#plt.plot(testY[0,:],testPredict[:,0],'y^',label='Test')\n",
    "upperb=np.amax([np.amax(trainPredict[:,0]),np.amax(testPredict[:,0])])\n",
    "lowera=np.amin([np.amin(trainPredict[:,0]),np.amin(testPredict[:,0])])\n",
    "plt.plot([0,10],[0,10],'r-',label='1:1 line')\n",
    "#plt.plot(trainPredictPlot)\n",
    "#plt.plot(testPredictPlot)\n",
    "plt.ylim(5,9)\n",
    "plt.xlim(5,9)\n",
    "\n",
    "plt.ylabel('log(predicted increase number)')\n",
    "plt.xlabel('log(actual increased number)')    \n",
    "    \n",
    "    #plt.xlim(5000,1000000)\n",
    "#plt.ylim(5000,1000000)\n",
    "\n",
    "\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "def recover_to_time_series(Yhat,initime):\n",
    "    \n",
    "    nperiod=len(Yhat)\n",
    "    date=pd.period_range(initime,periods=nperiod,freq='D')\n",
    "    print(len(date),len(Yhat[:,0]))\n",
    "    y=pd.DataFrame({'date':date,'predicted':Yhat[:,0]})\n",
    "    y=y.reset_index(drop=True).set_index('date')\n",
    "    \n",
    "    return y\n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "21 21\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlwAAAH6CAYAAADSsZLeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nO3de5zVVb3/8ddCUERlvCBFIjMW3gVRRtSTedfUxGOlpo2px36OmpestGNRXkq62eVkmUon02rC0jwmqalhZl7SQFFBUigBUfMCOorcYf3+WHtghJlhhvl+57v3zOv5eMxjz/7uvdf+zGXvec9a67tWiDEiSZKk/PQqugBJkqTuzsAlSZKUMwOXJElSzgxckiRJOTNwSZIk5czAJUmSlLPeRRcAMGDAgFhTU1N0GZIkSes0efLk12OMW3fkMWURuGpqapg0aVLRZUiSJK1TCGF2Rx/jkKIkSVLODFySJEk5M3BJkiTlrNA5XCGE0cDooUOHrnXbsmXLmDt3LosXL+76wqQO6Nu3L4MHD6ZPnz5FlyJJKlOhHDavrq2tjWtOmn/++efZbLPN2GqrrQghFFSZ1LYYI/PmzePtt99mu+22K7ocSVIXCCFMjjHWduQxZTukuHjxYsOWyl4Iga222sqeWElSm8o2cAGGLVUEf08lSetS1oGraCEEPvWpT626vnz5crbeemuOPvroDrVTU1PD66+/3un7NHfDDTdw7rnndqiOcrXBBhswYsQIdtttN0aPHs2bb76Z+3PecMMNvPTSSx16zKxZs9htt91yqkiS1J0ZuNqwySabMHXqVBYtWgTAvffeyzbbbFNwVcVqaICaGujVK102NHS+zY033pgpU6YwdepUttxyS66++urON9qGFStWrFfgkiRpfRm41uHII4/kjjvuAGD8+PGcdNJJq26bP38+xx57LMOHD2efffbhqaeeAmDevHkcfvjh7LHHHpx55pk0PzHhV7/6FaNGjWLEiBGceeaZrFixYq3nbO0+P//5z9lhhx044IADeOihh/L8slvU0AD19TB7NsSYLuvrswldTfbdd19efPHFVdevvPJK9tprL4YPH86ll14KpJ6mnXbaiVNPPZXhw4dz3HHHsXDhQgAmTpzIHnvswbBhwzj99NNZsmQJkHoQv/a1r7Hffvsxfvx4Jk2aRF1dHSNGjGDRokVMnjyZAw44gJEjR/LhD3+Yl19+GYDJkyez++67s+++++YeBCVJ3VflBK4DD1z74yc/SbctXNjy7TfckG5//fW1b2unE088kZtuuonFixfz1FNPsffee6+67dJLL2WPPfbgqaee4hvf+AannHIKAJdffjn77bcfTzzxBMcccwxz5swBYPr06fzmN7/hoYceYsqUKWywwQY0rJFWWrvPyy+/zKWXXspDDz3EvffeyzPPPNOR714mxoxJ3+rmFi5Mx7OwYsUKJk6cyDHHHAPAPffcw4wZM3jssceYMmUKkydP5oEHHgDg2Wefpb6+nqeeeor+/fvzk5/8hMWLF3Paaafxm9/8hqeffprly5dzzTXXrGq/b9++PPjgg5x88snU1tbS0NDAlClT6N27N+eddx633HILkydP5vTTT2dM6Yv6r//6L6666ioeeeSRbL5ISVKPVBZ7KZaz4cOHM2vWLMaPH89RRx31rtsefPBBfve73wFw8MEHM2/ePBobG3nggQe49dZbAfjIRz7CFltsAaTel8mTJ7PXXnsBsGjRIgYOHPiuNlu7z6OPPsqBBx7I1lunvTI/8YlP8Nxzz+X3hbeglBvbfby9Fi1axIgRI5g1axYjR47ksMMOA1Lguueee9hjjz0AWLBgATNmzGDIkCFsu+22fPCDHwTg5JNP5qqrruKwww5ju+22Y4cddgDg1FNP5eqrr+aCCy4A0vesJc8++yxTp05d9bwrVqxg0KBBNDY28uabb3LAAQcA8KlPfYq77rqrc1+sJKlHqpzAdf/9rd/Wr1/btw8Y0Pbt63DMMcdw4YUXcv/99zNv3rxVx1taw6zpjLWWzlyLMXLqqafyzW9+s9Xnau0+t912W+Fnww0ZkoYRWzreGU1zuBobGzn66KO5+uqrOf/884kx8qUvfYkzzzzzXfefNWvWWt+LEEKLP4/mNtlkkxaPxxjZdddd1+rFevPNNwv/nkuSuofKGVIs0Omnn84ll1zCsGHD3nV8//33XzUkeP/99zNgwAD69+//ruN33XUXb7zxBgCHHHIIt9xyC6+++iqQ5oDNXiPBtHafvffee1XgW7ZsGTfffHOuX3NLxo5N2ba5fv3S8SxUVVVx1VVX8d3vfpdly5bx4Q9/mOuvv54FCxYA8OKLL676vsyZM2dVQBo/fjz77bcfO+20E7NmzWLmzJkA/PKXv1zVO7WmzTbbjLfffhuAHXfckddee21Ve8uWLWPatGlsvvnmVFVV8eCDDwKsNfwrSVJ7GbjaYfDgwXz2s59d6/hll13GpEmTGD58OBdffDE33ngjkOZ2PfDAA+y5557cc889DCl1Ae2yyy5cccUVHH744QwfPpzDDjts1eTsJq3dZ9CgQVx22WXsu+++HHrooey55575f+FrqKuDceOguhpCSJfjxqXjWdljjz3Yfffduemmmzj88MP55Cc/yb777suwYcM47rjjVoWknXfemRtvvJHhw4czf/58zj77bPr27cvPf/5zjj/+eIYNG0avXr0466yzWnye0047jbPOOosRI0awYsUKbrnlFv77v/+b3XffnREjRvDwww8D6USFc845h3333ZeNN944uy9Ukjooj7PE1XXKdmuf6dOns/POOxdUkcrZrFmzOProo5k6dWrRpazi76ukPDWdJd78xKV+/bL/p1ftU3Fb+4QQRocQxjU2NhZZhiRJZS3vs8SVv0IDV4xxQoyxvqqqqsgyVGFqamrKqndLkvKW11nizTlkma/KOUtRkqQeKq+zxJusOWTZtLA1rOeQZYzwyiswYwbsuy/07g3/+79w1VXwwgvQpw/07Zs+pkxJ46NXXw2///3q400f112XJg7ffjs8/XQ6tvHG6XLTTeGEE9JzTpsGjY3p+JZbwtZbQytnpxehrANXjNHT8lX2ymEepKTubezYludwZXWWeFtDlm0GrjfeSIVstBH89a9pQfLnnktBq3SSE889B9tvn4JQTQ3svz+sXAmLF8OSJbDhhul+y5fDggVpsfLFi9PHihUpbAHceiuUTk5bZcstVweuSy5J92lu6NBUC8DXvgazZsHAgas/qqvhQx9Kt69YARts0IHvWseU7aT5559/ns0224ytttrK0KWyFWNk3rx5vP3222y33XZFlyOpG2toSAFozpzUszV2bHYT5nv1Sp1SawoBVq6I6ZPZs+GXv0wBpilUzZsH990HBx0Et90Gn/887LBDClhNl/vtl3qiOitGWLZsdVBbvDhdf//70+1PPw0vvwyLFsH8+fDqq6ln7QtfSLefckqq9dVX0+MARo2CRx9Nn++5Jzz/fOoZawpk//EfcOGF6fa77kqhceBAwm67dXjSfNkGrmXLljF37lwWL15cUFVS+/Tt25fBgwfTp0+fokuRpPVSU9PykGV1rxeY9fM/p7Dy2GOw994weHAKUk2h6uMfTw1UihjhrbdWB69ddknHr7oKZs5Mx5s+Ro2C669Ptw8cCK+9Bn37EhYv7nDgKtshxT59+thjIElSHhYtgieeSCFqhx0YO/Yo6s9YycJFq8+l69drEWM/eCdU75QO7LknvPPO2itgl5F29QKGAFVV6aO5889vu/H7708h7O23obTnb0eUbeCSJEkZihHOPRcefjgNv61YkY6feSZ11x4F9GLMFxYz59WNGDIkMHbsxtTVNdtarXfv9FGmMp/4v6ZddlndG7YeynZIUZIkdVCMaWL4Y4+lj7//PfXkTJiQbj/44BSaRo1KH3vtBYMGFVpyVlodFq1O35Isrc/Cp+UbVSVJqiB5Tmpv1WuvpeUQDjwwXT/hBLjllvT5RhulYcDhw1ff/777ci6oOF2xVllnGLgkSeqk3IezmkyfDnfeuboHa9asNCepsRE22yw92SGHpN6rYcPSelc9RN5rlXWWm1dLktRJmW+9s3w5PPVUWiy0vj4tFgrwpz+lZQoefTQNB155Jfz5z2m5AoBjj4Wzzko9W2UWtvJeyX7s2LXn82e5Vlln2cMlSVIndWo4K8Y0gb13b3jySTjvPJg8eXWC22IL+OQnYdtt4eST07Dhe96TWe1doSt6AJva6fJh3XZy0rwkSZ3UoQnb8+enyexNw4KPPQZf/Wo6g3D2bDjppNWT2keNgg98YPVq6xWqKye0dwUnzUuSVIBWt965bCn87fF0YJ990jpWW2+dtrYJAXbeGY46Kl1CSiAPP9z1X0DOyn1Ce1cwcEmS1EnvHs6KDNlkPmMHfJ+6M76T5mMdfjjcfXfaTPmaa9Iq7SNHQv/+xRbeRcp9QntXKHRIMYQwGhg9dOjQM2Y0bS4pSVK5izF1z/z97+njjTdg3Lh028EHw6RJaVJ786HBbbYptuYCrTmHC1IP4Lhx5TPHqiPWZ0jROVySpG6v02tkzZsHW22VPv/Od+B730vbvEA6G7C2Fh58MJ2C9/rrsOWW6XOtUsg6ZTlxDpckSWvo8Bly77yTeqiaeq+a1ruaOzf1Ug0aBEceuXql9uHD0yKjTQYMyPtLqkh1dZUbsLJgD5ckqVtr8wy555amfQUfewyOOAK22w5+9Sv41KdW36kpWP3XfxmmBNjDJUnSWlo9Q272yjRpfcmSdOBnP0uB67DD4I470jDhwIFdV6jKXtOwaDrjoWMMXJKkyvPii2leVdPH/PlpvapDDoGlS+H441cdHxLuZnbcdq0mhmz0alpktKkHq7o63fCe96SlGqRmWpr43xEGLklS4RoaYMyXVjJnbq80oXrU76nb4s7VYWrePNh/f/jRj9IDdtkF3nrr3Y2cdloKXH36pEDWvz/suitj33s39X89hYXLN1x11379ImPHvRfqruy6L1IVraXtmzrCwCVJKs4rr9Dwhcep//WBLIwbA6VJ7XMOh03vom7wM+nswJqa9NHkmmvSRPUtt0y3N31AWlC02bzgOoC1zpALPXoCtzqus4u0OmlektT1Jk6EH/wA/vhHalbMZDY1a92lUrd9Uff07pMvaolxUof2W3KREElS/lauhAceWD0MOH162qj5oouYE6pbfEhP2vZF5W/s2LRY6/oycEmS8jNjBlxySZrQfsABcPPN6fgZZ6Tuq29+kyFDWu4o6Enbvqj81dWllfGrW/7/YJ0MXJKk7L3zDuy7L+ywA1xxRdo78Je/hBNPTLdvtBFssAHQcs9Bv37puFRO6uqahrknT+7oYw1ckpSzhoY0/6NXr3TZ0FB0RTlYuhRuuy3Ny4K0SfPQoWkbnBdegHvugZNPTsfX0LznIIR0Wal77EmtcdK8JOWou23a+y4xpq1vfvELuOmmtHRDdTXMnAm9PQle3df6rDRvD5ck5ailtXsWLmxarbrCXXkl7L13WqH9sMPgzjsNW1IrfFVIUo5a3Vam0s7Ae/HFtN3N+PHwxS+mzZs//vG09tVxx0FVVdEVSmXNwCVJORoypOWNkyviDLzFi+HrX089V1OmpGPbb5+OQzrz8AMfKK4+qYI4pChJOaqoM/DmzYNf/xp++tN0faON0iS0/v3hW9+Cp5+GZ5+Fj3602DqlCmTgkqQclf0ZeM88A9/4BnzwgzBwYCrsxz9Ot4UAzz0Hf/kLDYP/m5qjd6PXBqH7nmkp5cizFCWpJ1mwAP78Z/jIR9I6FWefDddeCyNHwlFHpeO1tavWyIJufqaltB7W5yxFA5ckdXczZqR5WHfcAX/5S1oz67HHYK+90uz9Pn1g0KBWH/7uPeRWc69D9VQuCyFJPdS7FletjjT8b6k76r770mrvF1wAc+fCeeelY7vvnm4fMqTNsAXd6ExLqUCepShJFa6hAerPiCxclPYknD0nUP+Z3rAx1H1sX/jRj9Jw4fvfv17tV/SZllKZsIdLkipZjIw5e/6qsNVk4bIN0+KqG28M55673mELKuxMS6lMFRq4QgijQwjjGhsbiyxDkirLsmUwcWL6PATmvL1Fi3fLasiv7M+0lCqAk+YlqVK89FJKOuPGwcsvw7RpsMsu1FRHZs8Ja93dSe1SPpw0L0nd0Zw5cMIJKUFdfnma8D5hAuy4IwBjvxEc8pPKnIFLksrR22/D9Onp8/794ZFH4LOfTUs83HUXHH30qrWyHPKTyp+BS1LZe9eSBzXdfJXzadPgnHPgfe9LiSlG2HzzNDb43e/C0KEtPqyuLt1l5cp0adiSyovLQkgqa2uucj57droO3SxUTJwIV1wB99+f9jD8xCdS8GrSbOV3SZXHHi5JZW3MmHdvKQPp+pgxxdSTqRdfhLfeSp/PmZO6pr797bRA6Y03wqhRaYxQUsUzcEkqa91ulfMYUy/W8cenyVbXX5+On3wyzJwJX/wiDBhQaImSsmfgklTWWlvNvCJXOf/pT2G33eCgg9IQ4gUXwDHHpNv69HHYUOrGDFySylrFrnK+YEHaMPoHP1h9bPz4tPL79den4cTvfrdTK8BLqhxOmpdU1pomxo8Zk4YRhwxJYassJ8w/8wz87nfwpz+lZRyWLYNNNoEzz0wp8ZZbYMsti65SUgHs4ZJU9spyyYMY4R//gB//GF5/PR275x649FJ45x34/Ofh3nvhtddWd9EZtqQeyx4uSWqvt95KK7zfe2/qxXrxxXT8fe+Dj30MTjstTX530rukNRi4JKk1CxbAX/8KW2wB++yTerJOPjn1VB1yCBx6aPpomoe1+ebF1iupbBm4JKm5v/1tdQ9W0zysT34yBa73vx+mTIFhw9Ky95LUTgYuSWrujDPS9jp77JHmYR16KHzwg6tv33334mqTVLEMXJJ6tn//OwWrH/4Qtt4afvUr2GYb52FJypR94pJ6rkcegZEj4bbb0lAhpB4sw5akjBm4JPU8McJ118EBB6SNoh95BA47rOiqJHVjBi5JPc9VV8FZZ6UzDSdNcl6WpNw5h0tSz1NXB0uWwBe+4P6FkrqEPVySeoaJE9PipMuWpTlaX/yiYUtSlzFwSereYoQrr4TDD4dnn01b7UhSFzNwSeq+3n4bTjgh9WZ9/OPw6KNpGx5J6mIGLkndV10d3Hpr6uH6zW9g002LrkhSD+WkeUndT4wQAlxxBVxwARx8cNEVSerhDFySuo8VK+Cyy9Im09dcA8OHF12RJAEOKUrqLubPh6OPTr1aS5ak8CVJZcIeLkmV78kn05IPL7wA114L9fVpSFGSyoSBS1JlW7QoLfnQuzc88ADss0/RFUnSWgxckirT8uVp4dKNN4bx42HXXeE97ym6KklqkXO4JFWef/87nXl4zTXp+sEHG7YklTUDl6TK8sgjMHJk2nR6882LrkaS2sXAJakyxAjXXQcHHAB9+6bg9clPFl2VJLWLgUtSZXjqKTjrLDj00NS7tfvuRVckSe3mpHlJ5e3VV2HgwBSw7rsP9t8/TZaXpAqSeQ9XCGHnEMK1IYRbQghnZ92+pB5i5Ur47nehuhoeeigdO+ggw5akitSuwBVCuD6E8GoIYeoax48IITwbQpgZQrgYIMY4PcZ4FnACUJt9yZK6vZdfhg9/GC66CI46CnbaqeiKJKlT2tvDdQNwRPMDIYQNgKuBI4FdgJNCCLuUbjsGeBCYmFmlknqGCRPSHogPPQTjxsEtt8BWWxVdlSR1SrsCV4zxAWD+GodHATNjjP+KMS4FbgL+s3T/22OM/wHUZVmspB7gH/+AwYPh8cfhjDPcokdSt9CZOVzbAC80uz4X2CaEcGAI4aoQwnXAna09OIRQH0KYFEKY9Nprr3WiDEkV7+mn4U9/Sp9/4Qvwt785jCipW+nMWYot/dsZY4z3A/ev68ExxnHAOIDa2trYiTokVaoY4cc/TnO1hg5NSz/06gUbbVR0ZZKUqc70cM0Ftm12fTDwUufKkdRjvPoqHH00nH9+WlvrvvtS2JKkbqgzPVx/B7YPIWwHvAicCLjss6R1e+kl2HNPePNN+NGP4JxznKslqVtr77IQ44FHgB1DCHNDCJ+OMS4HzgXuBqYDv40xTsuvVEkVL5ZmDwwaBP/v/8Hf/w7nnmvYktTttfcsxZNijINijH1ijINjjD8rHb8zxrhDjPEDMcax+ZYqqaJNnw4f+lA6CzEEuOIKGDas6KoAaGiAmpo0ollTk65LUpYKnTARQhgdQhjX2NhYZBmS8tS06fTIkfDss2lR0zLS0AD19TB7dip19ux03dAlKUshxuJPEKytrY2TJk0qugxJWZs3Lw0d3nYbHHYY3HhjGk4sIzU1KWStqboaZs3q6mokVYIQwuQYY4d20/GUIEn5+cEP4I474Hvfgz/+sezCFsCcOR07Lknrw8AlKVvLlsHzz6fPv/IVmDQJPv/5sl3yYciQjh2XpPVRnu+AkirTzJnwwQ+mdbUWL4a+fdO+iGVs7Fjo1+/dx/r1S8clKSsGLkmd1tAANQMW0Gv791Mz+RYajv51ClsVoK4u7ZFdXZ1OnqyuTtfr3AlWUoYKnTQfQhgNjB46dOgZM2bMKKwOSeuvoQHq/2sZC5f1WXWsXz9Di6Tua30mzXuWoqROqamOzJ6z9sKlnuUnqbvyLEVJXW7OCy2vEu9ZfpK0moFL0vr517/guOMYss2KFm/2LD9JWs3AJanj5s2DI4+EP/+ZsRe86ll+krQOBi6pm8t8n8DFi+HYY9Py7L//PXVfGORZfpK0Dr2LLkBSfpr2CVy4MF1v2icQ1jMQrVwJp54KDz4Iv/kN7LffqrYMWJLUOnu4pG5szJjVYavJwoXp+Hp55RV49FH4znfghBM6XZ8k9RSF9nA1W4eryDKkbivzfQIHDYInn4T+/de7JknqiQrt4YoxTogx1ldVVRVZhtRtZbZP4B/+AJ/5DCxfDlVVabKWJKndHFKUurFM9gmcPBk+8Ql47DFYsiTT+iSppzBwSd1Yp/cJnDULjj4att469XJtskme5UpSt+VZilI3t95nEL7xBhx1VFoG4r774L3vzbw2Seop7OGS1LKpU+Hf/4bbboOddy66GkmqaPZwSWrZhz6UhhQ9I1GSOs0eLknvdumlcO216XPDliRlwsAlabX//V/42tfSmYkxFl2NJHUbhQauEMLoEMK4xsbGIsuQBPDHP8JZZ8ERR8BPfuJaW5KUIRc+lQRTpsDxx8OwYfDb30KfPkVXJEndikOKkuChh2DLLeGOO2CzzYquRpK6HQOXJDjnnLQMxPveV3QlktQtGbiknmrZMjjhBPjLX9J1e7YkKTcGLqknihHq6+Hmm9NaW5KkXBm4pJ7oa1+DG26Ayy6DU08tuhpJ6vYMXFJPc+ONKWiddhpccknR1UhSj2DgknqaiRPhkEPguutca0uSuoh7KUo9zQ03wKJFsOGGRVciST2GK81LPcGLL8KHP5wmyPfqBZtsUnRFktSjuNK81N3NnAn77QePPAJvvll0NZLUIzmHS+rOpkxJYWvBArjvPhgxouiKJKlHMnBJBWtogJqaNNJXU5OuZ2LSJDjggDRX669/hdrajBqWJHWUgUsqUENDWn909uy0Funs2el6JqHrAx+Aww9P+yTutFMGDUqS1peBSyrQmDGwcOG7jy1cmI6vtz/+ERYvhi22SCvJb7ttp2qUJHWegUsq0Jw5HTu+Tj/8IRx5JFx55XrXJEnKnoFLKtCQIR073qoY4atfhQsugI9+FC66qNO1SZKyY+CSCjR2LPTr9+5j/fql4+22YgV85jNwxRXw6U/Db38LfftmWqckqXMMXFKB6upg3Diork677FRXp+t1dR1oZO7cFLK++EX46U+htxtISFK5CTHGomugtrY2Tpo0qegypMqyZEla8iEEePllGDSo6IokqUcIIUyOMXZorR17uKRKNH8+HHhgGkYEw5YklTkDl1RpXnopLWj6+OOw225FVyNJaodCJ3uEEEYDo4cOHVpkGVLlmDkTDjsMXn89rbd10EFFVyRJagc3r5YqxTvvpJ6tBQvgz382bElSBfF0JqlSbLIJfO97aQNqt+qRpIpi4JLK3YQJ6XL0aDjxxGJrkSStFyfNS+XsF79IK8d/97tpNXlJUkUycEnl6n/+B049NS3/8Ic/pPW2JEkVycAllZumfRE/9zn42Mfgjjtgs82KrkqS1AkGLqkcvfEG/L//l7bs2WijoquRJHWSk+alcrF0aVrUtKYGrroqDSE6jChJ3YI9XFI5WLAgnYW4//7p8169DFuS1I3YwyUV7dFH0/DhM8/AT38Km25adEWSpIzZwyUVZdkyuOAC2HdfePPNNDn+9NOLrkqSlAMDl1SU3r3huefgM5+BadPgiCOKrkiSlBMDl9SVXnsNPv1pmD07zdG6/Xb48Y+hf/+iK5Mk5cjAJXWFGOGXv4Sdd06XDz+cjvd2GqUk9QQGLilvs2al4cJTToEddoAnnoCTTiq6KklSFzJwSXn73vdSj9aPfwwPPgi77lp0RZKkLhZigRvihhBGA6OHDh16xowZMwqrQ8rck0+my913h8ZGeOst2HbbYmuSJGUihDA5xljbkccU2sMVY5wQY6yvqqoqsgwpO4sWwZe/DLW1cOGF6VhVlWFLkno4hxSlrPzlL6lH65vfhJNPhptuKroiSVKZ8BQpKQt33QVHHQXbbQf33guHHlp0RZKkMmIPl9QZr7ySLg89FL7zHXj6acOWJGktBi5pHRoaoKYm7SddU5Ou89JL8PGPwx57pG15+vSBiy6CTTYpuFpJUjlySFFqQ0MD1NfDwoXp+uzZUH/6MtjgEurinXD55YYsSdI6FbosRJPa2to4adKkosuQ1lJTk0LWmqo3+jezpi6AoUO7vCZJUrHWZ1kIe7ikNsyZ08rxpe+Boe/t2mIkSRXLOVxSa+bOZcim81u8aciQ0MXFSJIqmYFLWtPUqXDaabDddoxdcD79Nljyrpv79YOxY4spTZJUmQxcUnOvvw577gk33wznnEPdv65g3I0bUV0NIUB1NYwbB3V1RRcqSaokzuFSz7ZiBfzf/6XNpWbt+ycAAB5pSURBVL//fRgwAH77W/jQh2CrrQCoqzFgSZI6xx4u9UyLFsE118COO8Lxx8OECWmDaYBjj10VtiRJyoKBSz3Pww+nscHPfCYFq1tugX/8A/r3L7oySVI35ZCieobnn0/zs/baC3bZBfbfH84/Pw0dBs84lCTlyx4udW+PPw4nnZQWKD3nnHRs881Tr9b++xu2JEldwsCl7umhh9Im0iNHwh13wBe+kCbHS5JUAIcU1X0sW5bOOuzbF557Dp55Br79bTjzTKiqKro6SVIPZg+XKt/SpfDDH8IHPpDOPAQ4+eQ0b+uLXzRsSZIKZw+XKtuTT8Ipp8BTT6U5Wbvvno736VNsXZIkNWPgUuX62c/g7LNhyy3h97+HY44puiJJklrkkKIq17BhcNxxMG2aYUuSVNYKDVwhhNEhhHGNjY1FlqFKsXIl/OAH6YxDgFGj4Ne/dlV4SVLZKzRwxRgnxBjrq5zUrHX517/goIPg85+HmTNh+fKiK5Ikqd0cUlR5ixGuvRaGD4cpU+DnP4fbboPeTj+UJFUO/2qpvL34YhpC/I//SJPkhwwpuiJJkjrMHi6VnxjhT39Kl4MHw2OPwd13G7YkSRXLwKXy8sor8NGPwmGHwR/+kI7tuiv08ldVklS5/Cum8nHzzSlc/fGP8L3vwVFHFV2RJEmZMHCpPJx/PpxwAmy3HTzxRDobcYMNiq5KkqRMOGlexYoRQkhDiAMHwsUXewaiJKnbsYdLxWhshE9/Gr75zXR99Gj4ylcMW5KkbsnApa43cWJaV+uGG2Dp0qKrkSQpdwYudZ133oHzzoNDD4W+feHhh+Gyy4quSpKk3Bm41HWmT4frroMLLkgT4/feu+iKJEnqEk6YUX5ihL/8JfVkffnLUFub9kF0AVNJUg9jD5ey9/rraR2tnXai4aBx1Hz1U/TqFampgYa/GrYkST2PPVzK1h/+AB//OCxdSsP2l1K/4VdYuDT9ms2eDfX16W51dQXWKElSF7OHS53T1Js1YUK6vs8+cPbZMHUqY5ZetipsNVm4EMaMKaBOSZIKZOBSx8UI998PJ50E22wDF16YtuMBGDAA/ud/YNddmTOn5Ye3dlySpO7KIUV13HHHwa23QlUVnHlmGifcbbe17jZkSBpGbOm4JEk9iT1caltTb9app8Jbb6Vjn/pUWrT0pZfgqqtaDFsAY8dCv37vPtavXzouSVJPYuBSy5qdachBB8Hvfw9PP51uO/bYFMDWTFNrqKuDceOgujptl1hdna47YV6S1NMYuLS2uXNXz80aMGB1b9YHP9jhpurqYNYsWLkyXeYRthoaoKYGevVKlw0N2T+HJEmd4RwuwRtvwM9/DvPmpfG+wYPT5RFHtDpcWC4aGtIUsoUL03WXnpAklaMQYyy6Bmpra+OkSZOKLqNnWrkS9toLHn8cDj4Y7r03dRVViJqalifmV1enHjVJkrIWQpgcY6ztyGMq5y+r8vHrX6ewdcMNMHFiRYUtaH2JCZeekCSVk8r666psLVkCX/0q7LFHOvOwArW2xIRLT0iSyomBqyebNQs22AC+9a2K69lq4tITkqRK4KT5nmzHHWH6dOhdub8GTRPjx4xJw4hDhqSw5YR5SVI5qdy/tOqcBx6AkSNhk02KrqTT6uoMWJKk8laZ40jqnFdegaOOgs99ruhKJEnqEQxcPdHXvw6LF8NFFxVdiSRJPYKBq6f55z/huuvgjDNg++2LrkaSpB7BwNXTfOUrsOGGcMklRVciSVKPYeDqSZYuhX//O83dGjSo6GokSeoxPEuxJ9lwQ7jvPli+vOhKJEnqUezh6immToWXXoIQoE+foquRJKlHybyHK4RwLPARYCBwdYzxnqyfQx20ciWcdhq88w4880wKXZIkqcu0q4crhHB9COHVEMLUNY4fEUJ4NoQwM4RwMUCM8bYY4xnAacAnMq9YHXfLLTB5Mlx8sWFLkqQCtHdI8QbgiOYHQggbAFcDRwK7ACeFEHZpdpevlG5XkZYtS/ve7LYbnHxy0dVIktQjtWtIMcb4QAihZo3Do4CZMcZ/AYQQbgL+M4QwHfgWcFeM8fEMa9X6+NnPYOZMmDAhbVQtSZK6XGcmzW8DvNDs+tzSsfOAQ4HjQghntfbgEEJ9CGFSCGHSa6+91oky1KZ//hMOOAA+8pGiK5EkqcfqTOBqaTJQjDFeFWMcGWM8K8Z4bWsPjjGOizHWxhhrt956606UoTZdeSXce2+rc7caGqCmBnr1SpcNDV1anSRJPUJnAtdcYNtm1wcDL3WuHGVm3jx44on0eSvLQDQ0QH09zJ4NMabL+npDlyRJWetM4Po7sH0IYbsQwobAicDt2ZSlThs7FvbaC158sdW7jBkDCxe++9jChem4JEnKTnuXhRgPPALsGEKYG0L4dIxxOXAucDcwHfhtjHFafqWq3WbPhquvhlNOgW22afVuc+Z07LgkSVo/7T1L8aRWjt8J3JlpReq8Sy9Nc7Yuv7zNuw0ZkrJZS8clSVJ2Ct3aJ4QwOoQwrrGxscgyupenn4Zf/ALOOw+23bbNu44dC/36vftYv37puCRJyk6hgSvGOCHGWF9VVVVkGd3LtGnwvvfBl760zrvW1cG4cVBdnTrEqqvT9bq6LqhTkqQeJMQYi66B2traOGnSpKLL6D6WLoUNNyy6CkmSuqUQwuQYY21HHlNoD5cyFCPcd1+6NGxJklRWDFzdxe23wyGHwO9+V3QlkiRpDQau7mDFCvjyl2GHHeDYY4uuRpIkrcHAVebatfXOL34BzzwD3/gG9G7XSh+SJKkLFTppPoQwGhg9dOjQM2bMmFFYHeWqaeud5qvB9+u3xpmEixalnq33vQ/+9rdW90yUJEnZqLhJ8y4L0bZ2bb0zc2a6/Pa3DVuSJJUpx5/KWLu23hk2DP75T89MlCSpjDmHq4y1tsXOquMPPwxLlhi2JEkqcwauMtbm1jsvvQSHHgoXXVRIbZIkqf0MXGWsza13Lr8cli+HCy4oukxJkrQOzuEqc3V1Lext+Oyz8LOfwWc+A+9/fyF1SZKk9iu0hyuEMDqEMK6xsbHIMirPmDGw8cbwla8UXYkkSWoHl4WoNIsWpdMUL7wQBg4suhpJktQODilWmo03hkcfhWXLiq5EkiS1k5PmK8m0afD662kGvUtBSJJUMezhqhQrV66ePf/EE64qL0lSBTFwVYqbboInn0wbLBq2JEmqKA4pVoKlS9MZiSNGwIknFl2NJEnqIANXJzQ0QE0N9OqVLhsacnqi666D55+Hb30rPZkkSaoo3fqvd56BqKEB6uth9myIMV3W1+cUuqZPh4MPhsMPz6FxSZKUtxBjLO7JQxgNjB46dOgZM2bMyLTtpkC0cOHqY/36Ndsap5NqalLIWlN1Ncya1fn217JkCWy0UQ4NS5KkjgghTI4x1nboMUUGria1tbVx0qRJmbaZdyDq1Sv1bK0phHRCoSRJ6p7WJ3B12yHFOXM6dryjhgzp2HFJktRzddvAlXcgGjs2DVE2169fOi5JktRctw1ceQeiuro0H6y6Og0jVldnNz9MkiR1L9124dOm4DNmTBpGHDIkha0sA1FdnQFLkiStW7cNXGAgkiRJ5aHbDilKkiSVCwOXJElSzgxckiRJOTNwSZIk5azQwBVCGB1CGNfY2FhkGZIkSbkqNHDFGCfEGOurqqqKLEOSJClXDilKkiTlzMAlSZKUMwOXJElSzgxckiRJOTNwSZIk5czAJUmSlDMDlyRJUs4MXJIkSTkzcEmSJOXMwCVJkpQzA5ckSVLO3LxakiQpZ25eLUmSlDOHFCVJknJm4JIkScqZgUuSJClnBi5JkqScGbgkSZJyZuCSJEnKmYFLkiQpZwYuSZKknBm4JEmScmbgkiRJypmBS5IkKWcGLkmSpJwZuCRJknJm4JIkScpZoYErhDA6hDCusbGxyDIkSZJyVWjgijFOiDHWV1VVFVmGJElSrhxSlCRJypmBS5IkKWcGLkmSpJwZuCRJknJm4JIkScqZgUuSJClnBi5JkqScGbgkSZJyZuCSJEnKmYFLkiQpZwYuSZKknBm4JEmScmbgkiRJypmBS5IkKWcGLkmSpJwZuCRJknJm4JIkScqZgUuSJClnBi5JkqScGbgkSZJyZuCSJEnKWaGBK4QwOoQwrrGxscgyJEmSclVo4IoxTogx1ldVVRVZhiRJUq4cUpQkScqZgUuSJClnBi5JkqScGbgkSZJyZuCSJEnKmYFLkiQpZwYuSZKknBm4JEmScmbgkiRJypmBS5IkKWcGLkmSpJwZuCRJknJm4JIkScqZgUuSJClnBi5JkqScGbgkSZJyZuCSJEnKmYFLkiQpZwYuSZKknBm4JEmScmbgkiRJypmBS5IkKWcGLkmSpJwZuCRJknJm4JIkScqZgUuSJClnBi5JkqScGbgkSZJyZuCSJEnKmYFLkiQpZwYuSZKknBm4JEmScmbgkiRJypmBS5IkKWcGLkmSpJxlHrhCCO8PIfwshHBL1m1LkiRVonYFrhDC9SGEV0MIU9c4fkQI4dkQwswQwsUAMcZ/xRg/nUexkiRJlai9PVw3AEc0PxBC2AC4GjgS2AU4KYSwS6bVSZIkdQPtClwxxgeA+WscHgXMLPVoLQVuAv6zvU8cQqgPIUwKIUx67bXX2l2wJElSpenMHK5tgBeaXZ8LbBNC2CqEcC2wRwjhS609OMY4LsZYG2Os3XrrrTtRhiRJUnnr3YnHhhaOxRjjPOCsTrQrSZLUrXSmh2susG2z64OBlzpXjiRJUvfTmcD1d2D7EMJ2IYQNgROB27MpS5Ikqfto77IQ44FHgB1DCHNDCJ+OMS4HzgXuBqYDv40xTsuvVEmSpMrUrjlcMcaTWjl+J3Dn+j55CGE0MHro0KHr24QkSVLZK3RrnxjjhBhjfVVVVZFlSJIk5cq9FCVJknJm4JIkScqZgUuSJClnBi5JkqScGbgkSZJyVmjgCiGMDiGMa2xsLLIMSZKkXLkshCRJUs4cUpQkScqZgUuSJClnBi5JkqScGbgkSZJyZuCSJEnKmYFLkiQpZ67DJUmSlDPX4ZIkScqZQ4qSJEk5M3BJkiTlzMAlSZKUMwOXJElSzgxckiRJOTNwSZIk5czAJUmSlDMXPpUkScqZC59KkiTlzCFFSZKknBm4JEmScmbgkiRJypmBS5IkKWcGLkmSpJwZuCRJknJm4JIkScqZgUuSJClnrjQvSZKUM1ealyRJyplDipIkSTkzcEmSJOXMwCVJkpQzA5ckSVLODFySJEk5M3BJkiTlzMAlSZKUMwOXJElSzsoicE2eDDU10NBQdCWSJEnZK4vABTB7NtTXG7okSVL3UzaBC2DhQhgzpugqJEmSslUWm1c3PzZnTlHVSJIk5aMsNq9ufmzIkKKqkSRJykdZDSn26wdjxxZdhSRJUrbKJnBVV8O4cVBXV3QlkiRJ2epddAEAI0fCpElFVyFJkpSPsunhkiRJ6q4MXJIkSTkzcEmSJOXMwCVJkpQzA5ckSVLODFySJEk5M3BJkiTlzMAlSZKUMwOXJElSzgxckiRJOTNwSZIk5azQwBVCGB1CGNfY2FhkGZIkSbkqNHDFGCfEGOurqqqKLEOSJClXDilKkiTlzMAlSZKUsxBjLLoGQghvA8/m+BRVQJ4TxfJuvyueo9Lb74rnsP3in8P2i3+OvNsfALyeY/v+DLp/+13xHDvGGDfr0CNijIV/AJNybn9cJbffHb4Gv0fdv/3u8DVUevvd4Wuo9L8H3eRnUNHtd9HX0OHf054ypDihwtvviueo9Pa74jlsv/jnsP3in6MrvoY8+TPo/u131XN0SLkMKU6KMdYWXYckqVj+PVAlWJ/f03Lp4RpXdAGSpLLg3wNVgg7/npZF4IoxZvYCCyFcH0J4NYQwtdmx34QQppQ+ZoUQpmTc/ogQwt9K7U8KIYzKuP3dQwiPhBCeDiFMCCH070T724YQ/hxCmB5CmBZC+Gzp+JYhhHtDCDNKl1vk8BzHl66vDCGs93+wbbR/ZQjhHyGEp0II/xdC2Dzj9r9eantKCOGeEML7smy/2e0XhhBiCGHA+rS/jq/hshDCi81eD0dl/TWEEM4LITxbOv6djOvP5LXcRvtZvpZbe45MXs8hhL4hhMdCCE+W2r+8dDyT13Ib7WfyOm5Nln8PAEIIR5R+H2eGEC4uHcvkvaKN9jN5r2it/Wa3ZfFe0VL9mbxPrOtryOK9oo2vIbO/+y1Zr9/TvCeudfUHsD+wJzC1ldu/B1ySZfvAPcCRpc+PAu7PuP2/AweUPj8d+Hon2h8E7Fn6fDPgOWAX4DvAxaXjFwPfzuE5dgZ2BO4HanNo/3Cgd+n4t9f3a2ij/f7N7nM+cG2W7ZeubwvcDcwGBuTwPboMuHB9221H+wcBfwI2Kt02MOvvUbP7rPdruY36s3wtt/YcmbyegQBsWvq8D/AosE9Wr+U22s/kddwVH8AGwD+B9wMbAk9m/F7RWvtZvVe02H7ptk6/V7RRfybvE+t4jqzeK1r9HjW7T6f+7mf10eU9XKGFHpzS8UySbozxAWB+K88dgBOA8Rm3H4Gm/1KrgJcybn9H4IHS5/cCH+9E+y/HGB8vff42MB3YBvhP4MbS3W4Ejs36OWKM02OMnV7+o43274kxLi/d7W/A4Izbf6vZ3TYh/dwza7908w+AL65v2+18jk5ro/2zgW/FGJeUbns14/aBzr+W22g/y9dya8+Ryes5JgtKV/uUPiIZvZZbaz+r1zG02jORZQ/aKGBmjPFfMcalwE3Af2b1XtFG+5m8V7TWfum2LN4r2mo/K609RybvFW20D2Tzd7+V39MO94YXMaR4A3BE8wMhhINI36DhMcZdge/m9NwfAl6JMc7IuN0LgCtDCC+Qav9Sxu1PBY4pfX486T+bTgsh1AB7kP5zfU+M8WVIfyiAgTk8R+baaP904K6s2w8hjC39nOuAS7JsP4RwDPBijPHJzrbb2nOUDp1bGu64fn2Hm9pofwfgQyGER0MIfwkh7JVx+00yey2v0X4ur+U1niOz13MIYYPSUMmrwL0xxkxfy620n4kQwgbA1cCRpB6Pk0IIu5C+Px9jdSjtjG2AF5pdn8va/3h05r2i1fYzeq9osf0M3yva+v5k9T7R2nNk9V6xrp9xp94r2vg9/Q5weYxxBOnnu86Ooi4PXK304GSVdNflJDqRcttwNvC5GOO2wOeAn2Xc/unAOSGEyaShiaWdbTCEsCnwO+CCNf4by0zez9Fa+yGEMcByoCHr9mOMY0o/5wbg3KzaL9U7hgxCXGvPUfoargE+AIwAXiZ1tWfZfm9gC9LQ00XAb0v/YWbVfpNMXssttJ/5a7mF58js9RxjXFF6wx8MjAoh7NbZeruw/dZ6hzLrQSMNi65pVY9QBu8Vrbaf0XtFS+1vRHbvFa3Vn+X7RGvPkdV7RZs/Yzr/XtFaD1qHe8PLYtI8OfxXvKYQQm/Sf02/ybpt4FTg1tLnN5N+QJmJMf4jxnh4jHEk6Rfnn51pL4TQh/QHoCHG2FT3KyGEQaXbB5H+o836OTLTWvshhFOBo4G6GON6d7W3o/5f04mh3Rba/wCwHfBkCGEW6Q/c4yGE92b4HMQYXyn9EV0J/JRO/K628j2aC9xaGo56DFhJWjk8q/Yzey230n6mr+VWfgaZvp5Lbb5JmlN1BBm/lltoPyvt6X3qrLm8uwdxMKU/jBm9V7TafjOdea9oqf05ZPde0WL9Wb5PtPYcZPde0dbPOIv3itZ+TzvcG14ugSvT/4pbcSjwjxjj3IzbhfTDPaD0+cFApkOWIYSBpctewFeAazvRViD91z49xvj9ZjfdTvpjQ+ny9zk8RyZaaz+EcATw38AxMcaFObS/fbO7HQP8I6v2Y4xPxxgHxhhrYow1pBf1njHGf2f8NQxqdrePkoZvMmsfuI30GiCEsANpEmuHt2lZx+9Qp1/LbbSf2Wu5jZ9BJq/nEMLWoXR2XQhhY0rfFzJ6LbfRflbW1TORhb8D24cQtgshbAicCNye1XtFG+1n8l7RSvu3Zvhe0Vr9mbxPtPUcZPRe0Ub7kM3f/dZ+TzveGx4LmKkP1PDus/D+CBzY7Po/ga3Xs+3xpC7QZaRfxE+Xjt8AnJVB7Wu1D+wHTCadHfEoMDLj9j9LOsPpOeBbkBasXc/29yv9sjwFTCl9HAVsBUwk/YGZCGyZw3N8tPQ1LQFeAe7OuP2ZpP9Emo6t75lBrbX/O9Ibz1OkVYy3ybL9Ne4zi86dpdja1/BL4OnS8duBQRm3vyHwq9L36XHg4Ky/R1m8ltuoP8vXcmvPkcnrGRgOPFFqfyqls7Cyei230X5Wr+N9mz+W1EPwpWbX7yeDsyBL3/PnSH9XxpSOZfJe0Ub7mbxXtNb+Grd39r2ipfozeZ9Yx3Nk8l7R1vcoo/eKFn9PSfs0Ni0eH4C31tVWISvNlyaQ/iHGuFvp+lnA+2KMl5SS7kRgSCyiOElS7krDPc8BhwAvknoqPhljnFa6/X7S0gSTCitSPV5rv6fALcDZMcb7QwiHAN+JaZpAq3rnXeyaQgjjgQOBASGEucClwPXA9SEtFbEUONWwJUndV4xxeQjhXNJaUhsA18cYp4UQPgr8CNgauCOEMCXG+OEia1XP1cbv6RnAD0uBbDFQv662ymIvRUmSpO6sXCbNS5IkdVsGLkmSpJwZuCRJknJm4JIkdZkQwoqQ9p9r+qhp474HhhD+0HXVSfnp8rMUJUk92qKYtguSehR7uCRJhQppk+wrQwh/D2nD5DOb3dw/hPB/IYRnQgjXllbolyqOPVySpK60cQhhSunz52OMHyXtqNEYY9wrhLAR8FAI4Z7SfUYBuwCzSbuSfIy06KRUUQxckqSu1NKQ4uHA8BDCcaXrVcD2pIWwH4sx/gtWLZy9HwYuVSADlySpaAE4L8Z497sOhnAga29o7WrdqkiOhUuSinY3cHYIoQ9ACGGHEMImpdtGhRC2K83d+gTwYFFFSp1hD5ckqWj/C9QAj4cQAvAacGzptkeAbwHDgAeA/yuiQKmz3EtRkiQpZw4pSpIk5czAJUmSlDMDlyQpNyGEbUMIfw4hTA8hTAshfLZ0fMsQwr0hhBmlyy1Kxw8LIUwOITxdujy4WVsjS8dnhhCuKs33kiqCgUuSlKflwBdijDsD+wDnhBB2AS4GJsYYtwcmlq4DvA6MjjEOA04FftmsrWuAetIaXdsDR3TNlyB1noFLkpSbGOPLMcbHS5+/DUwHtgH+E7ixdLcbKZ2VGGN8Isb4Uun4NKBvCGGjEMIgoH+M8ZGYzvb6BavPZJTKnoFLktQlQgg1wB7Ao8B7YowvQwplwMAWHvJx4IkY4xJSSJvb7La5pWNSRXAdLklS7kIImwK/Ay6IMb61rulXIYRdgW+Ttv2BtBr9mlzXSBXDHi5JUq5KK8j/DmiIMd5aOvxKaZiQ0uWrze4/mLTA6Skxxn+WDs8FBjdrdjDwElKFMHBJknJTOpPwZ8D0GOP3m910O2lSPKXL35fuvzlwB/ClGONDTXcuDTu+HULYp9TmKU2PkSqBK81LknITQtgP+CvwNLCydPjLpHlcvwWGAHOA42OM80MIXwG+BMxo1szhMcZXQwi1wA3AxsBdpA2v/SOmimDgkiRJyplDipIkSTkzcEmSJOXMwCVJkpQzA5ckSVLODFySJEk5M3BJkiTlzMAlqSKEEGaFEBaFEN4OIbwZQng4hHBWCGGd72MhhJoQQgwhuJ2ZpEIYuCRVktExxs2AauBbwH+TVjGXpLJm4JJUcWKMjTHG24FPAKeGEHYLIXwkhPBECOGtEMILIYTLmj3kgdLlmyGEBSGEfQFCCKeHEKaHEN4IIdwdQqju4i9FUg9h4JJUsWKMj5E2Nf4Q8A5pf73NgY8AZ4cQji3ddf/S5eYxxk1jjI+Ubvsy8DFga9L2M+O7sn5JPYeBS1KlewnYMsZ4f4zx6RjjyhjjU6TwdEAbjzsT+GaMcXqMcTnwDWCEvVyS8mDgklTptgHmhxD2DiH8OYTwWgihETgLGNDG46qBH5Ym4L8JzAdCqT1JypSBS1LFCiHsRQpIDwK/Bm4Hto0xVgHXkgIUQGzh4S8AZ8YYN2/2sXGM8eGuqF1Sz2LgklRxQgj9QwhHAzcBv4oxPg1sBsyPMS4OIYwCPtnsIa8BK4H3Nzt2LfClEMKupTarQgjHd81XIKmncU0aSZVkQghhOSk8PQN8nxScAD4DfC+E8GPgL8BvSRPoiTEuDCGMBR4KIfQBjogx/l8IYVPgptK8rUbgXuDmLv2KJPUIIcaWetolSZKUFYcUJUmScmbgkiRJypmBS5IkKWcGLkmSpJwZuCRJknJm4JIkScqZgUuSJClnBi5JkqScGbgkSZJy9v8B79mYA+EnDnYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "initime=df.index[look_back]\n",
    "predict_new=recover_to_time_series(trainPredict,initime)\n",
    "predict_new['predicted']=np.exp(predict_new['predicted'])\n",
    "fig = plt.figure(figsize=(10, 8))\n",
    "predict_new['predicted'].plot(color='r',style='--',label='Modeled',legend=True)\n",
    "df.plot(color='b',linestyle='',marker='o',label='Reported',legend=True)\n",
    "\n",
    "#plt.title(paralabel)\n",
    "#plt.ylabel(paralabel,fontsize=12)\n",
    "#plt.xlim('2018-01-01','2019-01-01')\n",
    "plt.yscale('log')\n",
    "\n",
    "#plt.yscale('log')\n",
    "plt.xlabel('Date',fontsize=12)\n",
    "plt.legend(loc='best',ncol=3)\n",
    "plt.show()    \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### using the data at the three dates (2-6,2-7, and 2-8) to Predict increased determined cases in 2-9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# convert an array of values into a dataset matrix\n",
    "def create_dataset_pr(dataset, look_back=1):\n",
    "    dataX = []\n",
    "    for i in range(len(dataset)-look_back+1): #-1):\n",
    "        a = dataset[i:(i+look_back), 0]\n",
    "        dataX.append(a)\n",
    "        print(a)\n",
    "    return np.array(dataX)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "def prediction (pred,model):\n",
    "    dataset1=np.log(pred['Inc_deter'].values)\n",
    "    dataset2=dataset1.reshape(-1,1)\n",
    "    sc1= MinMaxScaler(feature_range=(0,1))\n",
    "    dataset3 = sc.fit_transform(dataset2)\n",
    "    pred_X = create_dataset_pr(dataset3, look_back)\n",
    "    pred_X = np.reshape(pred_X, (pred_X.shape[0], 1, pred_X.shape[1]))\n",
    "    Pred_Y = model.predict(pred_X)\n",
    "    Pred_Y = sc.inverse_transform(Pred_Y)\n",
    "    return np.exp(Pred_Y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "file='predict0208.csv'\n",
    "pred=pd.read_csv(file)\n",
    "pr0208=prediction(pred,model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "pr0208"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.69073166 1.         0.        ]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[3152.1936]], dtype=float32)"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file='predict0209.csv'\n",
    "pred=pd.read_csv(file)\n",
    "pr0209=prediction(pred,model)\n",
    "pr0209"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}