LSTM_Model_Latest.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import time\n",
    "import numpy\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas\n",
    "import math\n",
    "import os\n",
    "from keras.models import Sequential\n",
    "from keras.callbacks import TensorBoard\n",
    "from keras.layers import Dense ,Activation\n",
    "from keras.layers import LSTM\n",
    "from keras.layers import GRU\n",
    "from keras.layers import SimpleRNN\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from keras import metrics\n",
    "from IPython.core.debugger import Tracer\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# fix random seed for reproducibility\n",
    "numpy.random.seed(10)\n",
    "cwd = os.getcwd()\n",
    "os.chdir(\"C:/Users/mukul.virmani/\")\n",
    "path = './Desktop/CPI_US_Medical.csv'\n",
    "scaler = MinMaxScaler()\n",
    "#get Transformed data after \n",
    "def getTansformedData(path):\n",
    "    frame = pandas.read_csv(path,engine='python',skiprows=0)\n",
    "    scaler = MinMaxScaler(feature_range=(0, 1))\n",
    "    frame['value'] = scaler.fit_transform(frame.value.values.reshape(-1,1))\n",
    "    return frame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "series_id    0\n",
       "year         0\n",
       "period       0\n",
       "value        0\n",
       "Month        0\n",
       "Date         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data=getTansformedData(path)\n",
    "def num_missing(x):\n",
    "  return sum(x.isnull())\n",
    "data.apply(num_missing, axis=0)\n",
    "#get_null=num_missing(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>series_id</th>\n",
       "      <th>year</th>\n",
       "      <th>period</th>\n",
       "      <th>value</th>\n",
       "      <th>Month</th>\n",
       "      <th>Date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M01</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>1/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M02</td>\n",
       "      <td>0.000216</td>\n",
       "      <td>2</td>\n",
       "      <td>2/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M03</td>\n",
       "      <td>0.000216</td>\n",
       "      <td>3</td>\n",
       "      <td>3/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M04</td>\n",
       "      <td>0.000432</td>\n",
       "      <td>4</td>\n",
       "      <td>4/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M05</td>\n",
       "      <td>0.000648</td>\n",
       "      <td>5</td>\n",
       "      <td>5/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M06</td>\n",
       "      <td>0.000648</td>\n",
       "      <td>6</td>\n",
       "      <td>6/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M07</td>\n",
       "      <td>0.000648</td>\n",
       "      <td>7</td>\n",
       "      <td>7/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M08</td>\n",
       "      <td>0.000864</td>\n",
       "      <td>8</td>\n",
       "      <td>8/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M09</td>\n",
       "      <td>0.001079</td>\n",
       "      <td>9</td>\n",
       "      <td>9/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M10</td>\n",
       "      <td>0.001295</td>\n",
       "      <td>10</td>\n",
       "      <td>10/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M11</td>\n",
       "      <td>0.001295</td>\n",
       "      <td>11</td>\n",
       "      <td>11/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1947</td>\n",
       "      <td>M12</td>\n",
       "      <td>0.001511</td>\n",
       "      <td>12</td>\n",
       "      <td>12/1/1947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M01</td>\n",
       "      <td>0.001727</td>\n",
       "      <td>1</td>\n",
       "      <td>1/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M02</td>\n",
       "      <td>0.001727</td>\n",
       "      <td>2</td>\n",
       "      <td>2/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M03</td>\n",
       "      <td>0.001943</td>\n",
       "      <td>3</td>\n",
       "      <td>3/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M04</td>\n",
       "      <td>0.002375</td>\n",
       "      <td>4</td>\n",
       "      <td>4/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M05</td>\n",
       "      <td>0.002375</td>\n",
       "      <td>5</td>\n",
       "      <td>5/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M06</td>\n",
       "      <td>0.002591</td>\n",
       "      <td>6</td>\n",
       "      <td>6/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M07</td>\n",
       "      <td>0.002807</td>\n",
       "      <td>7</td>\n",
       "      <td>7/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M08</td>\n",
       "      <td>0.003023</td>\n",
       "      <td>8</td>\n",
       "      <td>8/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M09</td>\n",
       "      <td>0.002807</td>\n",
       "      <td>9</td>\n",
       "      <td>9/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M10</td>\n",
       "      <td>0.003023</td>\n",
       "      <td>10</td>\n",
       "      <td>10/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M11</td>\n",
       "      <td>0.003238</td>\n",
       "      <td>11</td>\n",
       "      <td>11/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1948</td>\n",
       "      <td>M12</td>\n",
       "      <td>0.003238</td>\n",
       "      <td>12</td>\n",
       "      <td>12/1/1948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1949</td>\n",
       "      <td>M01</td>\n",
       "      <td>0.003238</td>\n",
       "      <td>1</td>\n",
       "      <td>1/1/1949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1949</td>\n",
       "      <td>M02</td>\n",
       "      <td>0.003454</td>\n",
       "      <td>2</td>\n",
       "      <td>2/1/1949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1949</td>\n",
       "      <td>M03</td>\n",
       "      <td>0.003454</td>\n",
       "      <td>3</td>\n",
       "      <td>3/1/1949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1949</td>\n",
       "      <td>M04</td>\n",
       "      <td>0.003454</td>\n",
       "      <td>4</td>\n",
       "      <td>4/1/1949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1949</td>\n",
       "      <td>M05</td>\n",
       "      <td>0.003454</td>\n",
       "      <td>5</td>\n",
       "      <td>5/1/1949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>1949</td>\n",
       "      <td>M06</td>\n",
       "      <td>0.003454</td>\n",
       "      <td>6</td>\n",
       "      <td>6/1/1949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>817</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2015</td>\n",
       "      <td>M02</td>\n",
       "      <td>0.923795</td>\n",
       "      <td>2</td>\n",
       "      <td>2/1/2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>818</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2015</td>\n",
       "      <td>M03</td>\n",
       "      <td>0.927739</td>\n",
       "      <td>3</td>\n",
       "      <td>3/1/2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>819</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2015</td>\n",
       "      <td>M04</td>\n",
       "      <td>0.933946</td>\n",
       "      <td>4</td>\n",
       "      <td>4/1/2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>820</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2015</td>\n",
       "      <td>M05</td>\n",
       "      <td>0.935354</td>\n",
       "      <td>5</td>\n",
       "      <td>5/1/2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>821</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2015</td>\n",
       "      <td>M06</td>\n",
       "      <td>0.934974</td>\n",
       "      <td>6</td>\n",
       "      <td>6/1/2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>822</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2015</td>\n",
       "      <td>M07</td>\n",
       "      <td>0.936507</td>\n",
       "      <td>7</td>\n",
       "      <td>7/1/2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>823</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2015</td>\n",
       "      <td>M08</td>\n",
       "      <td>0.935904</td>\n",
       "      <td>8</td>\n",
       "      <td>8/1/2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>824</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2015</td>\n",
       "      <td>M09</td>\n",
       "      <td>0.937636</td>\n",
       "      <td>9</td>\n",
       "      <td>9/1/2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>825</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2015</td>\n",
       "      <td>M10</td>\n",
       "      <td>0.944724</td>\n",
       "      <td>10</td>\n",
       "      <td>10/1/2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>826</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2015</td>\n",
       "      <td>M11</td>\n",
       "      <td>0.947891</td>\n",
       "      <td>11</td>\n",
       "      <td>11/1/2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>827</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2015</td>\n",
       "      <td>M12</td>\n",
       "      <td>0.949294</td>\n",
       "      <td>12</td>\n",
       "      <td>12/1/2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>828</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M01</td>\n",
       "      <td>0.953198</td>\n",
       "      <td>1</td>\n",
       "      <td>1/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>829</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M02</td>\n",
       "      <td>0.957172</td>\n",
       "      <td>2</td>\n",
       "      <td>2/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>830</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M03</td>\n",
       "      <td>0.959256</td>\n",
       "      <td>3</td>\n",
       "      <td>3/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>831</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M04</td>\n",
       "      <td>0.962693</td>\n",
       "      <td>4</td>\n",
       "      <td>4/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>832</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M05</td>\n",
       "      <td>0.965389</td>\n",
       "      <td>5</td>\n",
       "      <td>5/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>833</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M06</td>\n",
       "      <td>0.969129</td>\n",
       "      <td>6</td>\n",
       "      <td>6/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>834</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M07</td>\n",
       "      <td>0.974164</td>\n",
       "      <td>7</td>\n",
       "      <td>7/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>835</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M08</td>\n",
       "      <td>0.983035</td>\n",
       "      <td>8</td>\n",
       "      <td>8/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>836</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M09</td>\n",
       "      <td>0.984835</td>\n",
       "      <td>9</td>\n",
       "      <td>9/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>837</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M10</td>\n",
       "      <td>0.986221</td>\n",
       "      <td>10</td>\n",
       "      <td>10/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>838</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M11</td>\n",
       "      <td>0.986746</td>\n",
       "      <td>11</td>\n",
       "      <td>11/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>839</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2016</td>\n",
       "      <td>M12</td>\n",
       "      <td>0.989106</td>\n",
       "      <td>12</td>\n",
       "      <td>12/1/2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>840</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2017</td>\n",
       "      <td>M01</td>\n",
       "      <td>0.991079</td>\n",
       "      <td>1</td>\n",
       "      <td>1/1/2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>841</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2017</td>\n",
       "      <td>M02</td>\n",
       "      <td>0.992131</td>\n",
       "      <td>2</td>\n",
       "      <td>2/1/2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>842</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2017</td>\n",
       "      <td>M03</td>\n",
       "      <td>0.993592</td>\n",
       "      <td>3</td>\n",
       "      <td>3/1/2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>843</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2017</td>\n",
       "      <td>M04</td>\n",
       "      <td>0.991971</td>\n",
       "      <td>4</td>\n",
       "      <td>4/1/2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>844</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2017</td>\n",
       "      <td>M05</td>\n",
       "      <td>0.991863</td>\n",
       "      <td>5</td>\n",
       "      <td>5/1/2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>845</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2017</td>\n",
       "      <td>M06</td>\n",
       "      <td>0.995650</td>\n",
       "      <td>6</td>\n",
       "      <td>6/1/2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>846</th>\n",
       "      <td>CUSR0000SAM</td>\n",
       "      <td>2017</td>\n",
       "      <td>M07</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>7</td>\n",
       "      <td>7/1/2017</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>847 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             series_id  year period     value  Month       Date\n",
       "0    CUSR0000SAM        1947    M01  0.000000      1   1/1/1947\n",
       "1    CUSR0000SAM        1947    M02  0.000216      2   2/1/1947\n",
       "2    CUSR0000SAM        1947    M03  0.000216      3   3/1/1947\n",
       "3    CUSR0000SAM        1947    M04  0.000432      4   4/1/1947\n",
       "4    CUSR0000SAM        1947    M05  0.000648      5   5/1/1947\n",
       "5    CUSR0000SAM        1947    M06  0.000648      6   6/1/1947\n",
       "6    CUSR0000SAM        1947    M07  0.000648      7   7/1/1947\n",
       "7    CUSR0000SAM        1947    M08  0.000864      8   8/1/1947\n",
       "8    CUSR0000SAM        1947    M09  0.001079      9   9/1/1947\n",
       "9    CUSR0000SAM        1947    M10  0.001295     10  10/1/1947\n",
       "10   CUSR0000SAM        1947    M11  0.001295     11  11/1/1947\n",
       "11   CUSR0000SAM        1947    M12  0.001511     12  12/1/1947\n",
       "12   CUSR0000SAM        1948    M01  0.001727      1   1/1/1948\n",
       "13   CUSR0000SAM        1948    M02  0.001727      2   2/1/1948\n",
       "14   CUSR0000SAM        1948    M03  0.001943      3   3/1/1948\n",
       "15   CUSR0000SAM        1948    M04  0.002375      4   4/1/1948\n",
       "16   CUSR0000SAM        1948    M05  0.002375      5   5/1/1948\n",
       "17   CUSR0000SAM        1948    M06  0.002591      6   6/1/1948\n",
       "18   CUSR0000SAM        1948    M07  0.002807      7   7/1/1948\n",
       "19   CUSR0000SAM        1948    M08  0.003023      8   8/1/1948\n",
       "20   CUSR0000SAM        1948    M09  0.002807      9   9/1/1948\n",
       "21   CUSR0000SAM        1948    M10  0.003023     10  10/1/1948\n",
       "22   CUSR0000SAM        1948    M11  0.003238     11  11/1/1948\n",
       "23   CUSR0000SAM        1948    M12  0.003238     12  12/1/1948\n",
       "24   CUSR0000SAM        1949    M01  0.003238      1   1/1/1949\n",
       "25   CUSR0000SAM        1949    M02  0.003454      2   2/1/1949\n",
       "26   CUSR0000SAM        1949    M03  0.003454      3   3/1/1949\n",
       "27   CUSR0000SAM        1949    M04  0.003454      4   4/1/1949\n",
       "28   CUSR0000SAM        1949    M05  0.003454      5   5/1/1949\n",
       "29   CUSR0000SAM        1949    M06  0.003454      6   6/1/1949\n",
       "..                 ...   ...    ...       ...    ...        ...\n",
       "817  CUSR0000SAM        2015    M02  0.923795      2   2/1/2015\n",
       "818  CUSR0000SAM        2015    M03  0.927739      3   3/1/2015\n",
       "819  CUSR0000SAM        2015    M04  0.933946      4   4/1/2015\n",
       "820  CUSR0000SAM        2015    M05  0.935354      5   5/1/2015\n",
       "821  CUSR0000SAM        2015    M06  0.934974      6   6/1/2015\n",
       "822  CUSR0000SAM        2015    M07  0.936507      7   7/1/2015\n",
       "823  CUSR0000SAM        2015    M08  0.935904      8   8/1/2015\n",
       "824  CUSR0000SAM        2015    M09  0.937636      9   9/1/2015\n",
       "825  CUSR0000SAM        2015    M10  0.944724     10  10/1/2015\n",
       "826  CUSR0000SAM        2015    M11  0.947891     11  11/1/2015\n",
       "827  CUSR0000SAM        2015    M12  0.949294     12  12/1/2015\n",
       "828  CUSR0000SAM        2016    M01  0.953198      1   1/1/2016\n",
       "829  CUSR0000SAM        2016    M02  0.957172      2   2/1/2016\n",
       "830  CUSR0000SAM        2016    M03  0.959256      3   3/1/2016\n",
       "831  CUSR0000SAM        2016    M04  0.962693      4   4/1/2016\n",
       "832  CUSR0000SAM        2016    M05  0.965389      5   5/1/2016\n",
       "833  CUSR0000SAM        2016    M06  0.969129      6   6/1/2016\n",
       "834  CUSR0000SAM        2016    M07  0.974164      7   7/1/2016\n",
       "835  CUSR0000SAM        2016    M08  0.983035      8   8/1/2016\n",
       "836  CUSR0000SAM        2016    M09  0.984835      9   9/1/2016\n",
       "837  CUSR0000SAM        2016    M10  0.986221     10  10/1/2016\n",
       "838  CUSR0000SAM        2016    M11  0.986746     11  11/1/2016\n",
       "839  CUSR0000SAM        2016    M12  0.989106     12  12/1/2016\n",
       "840  CUSR0000SAM        2017    M01  0.991079      1   1/1/2017\n",
       "841  CUSR0000SAM        2017    M02  0.992131      2   2/1/2017\n",
       "842  CUSR0000SAM        2017    M03  0.993592      3   3/1/2017\n",
       "843  CUSR0000SAM        2017    M04  0.991971      4   4/1/2017\n",
       "844  CUSR0000SAM        2017    M05  0.991863      5   5/1/2017\n",
       "845  CUSR0000SAM        2017    M06  0.995650      6   6/1/2017\n",
       "846  CUSR0000SAM        2017    M07  1.000000      7   7/1/2017\n",
       "\n",
       "[847 rows x 6 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "splitfactor,prev=.80,4\n",
    "# convert an array of values into a dataset matrix and create sliding window approach.\n",
    "def Get_rolling_window_data(d1,prev):\n",
    "    x,y,val=[],[],len(d1)-prev\n",
    "    for z in range(val-1):\n",
    "        x.append(d1.values[z:(z + prev),3 ]), y.append(d1.values[z + prev, 3])\n",
    "    return numpy.array(x), numpy.array(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "672 165\n",
      "[[[0.0]\n",
      "  [0.00021589738828929964]\n",
      "  [0.00021589738828929964]\n",
      "  [0.00043179477657859233]]\n",
      "\n",
      " [[0.00021589738828929964]\n",
      "  [0.00021589738828929964]\n",
      "  [0.00043179477657859233]\n",
      "  [0.000647692164867885]]\n",
      "\n",
      " [[0.00021589738828929964]\n",
      "  [0.00043179477657859233]\n",
      "  [0.000647692164867885]\n",
      "  [0.000647692164867885]]\n",
      "\n",
      " ..., \n",
      " [[0.5971721760081867]\n",
      "  [0.6001947394442367]\n",
      "  [0.6021378159388405]\n",
      "  [0.6034332002685762]]\n",
      "\n",
      " [[0.6001947394442367]\n",
      "  [0.6021378159388405]\n",
      "  [0.6034332002685762]\n",
      "  [0.6040808924334441]]\n",
      "\n",
      " [[0.6021378159388405]\n",
      "  [0.6034332002685762]\n",
      "  [0.6040808924334441]\n",
      "  [0.6049444819866012]]] [[[0.6107737114704123]\n",
      "  [0.613580377518173]\n",
      "  [0.6155234540127766]\n",
      "  [0.6187619148371161]]\n",
      "\n",
      " [[0.613580377518173]\n",
      "  [0.6155234540127766]\n",
      "  [0.6187619148371161]\n",
      "  [0.6200572991668518]]\n",
      "\n",
      " [[0.6155234540127766]\n",
      "  [0.6187619148371161]\n",
      "  [0.6200572991668518]\n",
      "  [0.6222162730497447]]\n",
      "\n",
      " [[0.6187619148371161]\n",
      "  [0.6200572991668518]\n",
      "  [0.6222162730497447]\n",
      "  [0.6256706312623734]]\n",
      "\n",
      " [[0.6200572991668518]\n",
      "  [0.6222162730497447]\n",
      "  [0.6256706312623734]\n",
      "  [0.6273978103686878]]\n",
      "\n",
      " [[0.6222162730497447]\n",
      "  [0.6256706312623734]\n",
      "  [0.6273978103686878]\n",
      "  [0.6304203738047379]]\n",
      "\n",
      " [[0.6256706312623734]\n",
      "  [0.6273978103686878]\n",
      "  [0.6304203738047379]\n",
      "  [0.6332270398524987]]\n",
      "\n",
      " [[0.6273978103686878]\n",
      "  [0.6304203738047379]\n",
      "  [0.6332270398524987]\n",
      "  [0.6360337059002596]]\n",
      "\n",
      " [[0.6304203738047379]\n",
      "  [0.6332270398524987]\n",
      "  [0.6360337059002596]\n",
      "  [0.637760885006574]]\n",
      "\n",
      " [[0.6332270398524987]\n",
      "  [0.6360337059002596]\n",
      "  [0.637760885006574]\n",
      "  [0.6403516536660455]]\n",
      "\n",
      " [[0.6360337059002596]\n",
      "  [0.637760885006574]\n",
      "  [0.6403516536660455]\n",
      "  [0.6425106275489384]]\n",
      "\n",
      " [[0.637760885006574]\n",
      "  [0.6403516536660455]\n",
      "  [0.6425106275489384]\n",
      "  [0.6446696014318314]]\n",
      "\n",
      " [[0.6403516536660455]\n",
      "  [0.6425106275489384]\n",
      "  [0.6446696014318314]\n",
      "  [0.6470444727030135]]\n",
      "\n",
      " [[0.6425106275489384]\n",
      "  [0.6446696014318314]\n",
      "  [0.6470444727030135]\n",
      "  [0.6494193439741958]]\n",
      "\n",
      " [[0.6446696014318314]\n",
      "  [0.6470444727030135]\n",
      "  [0.6494193439741958]\n",
      "  [0.6509306256922208]]\n",
      "\n",
      " [[0.6470444727030135]\n",
      "  [0.6494193439741958]\n",
      "  [0.6509306256922208]\n",
      "  [0.6533054969634031]]\n",
      "\n",
      " [[0.6494193439741958]\n",
      "  [0.6509306256922208]\n",
      "  [0.6533054969634031]\n",
      "  [0.6556803682345853]]\n",
      "\n",
      " [[0.6509306256922208]\n",
      "  [0.6533054969634031]\n",
      "  [0.6556803682345853]\n",
      "  [0.6589188290589246]]\n",
      "\n",
      " [[0.6533054969634031]\n",
      "  [0.6556803682345853]\n",
      "  [0.6589188290589246]\n",
      "  [0.6617254951066855]]\n",
      "\n",
      " [[0.6556803682345853]\n",
      "  [0.6589188290589246]\n",
      "  [0.6617254951066855]\n",
      "  [0.6638844689895784]]\n",
      "\n",
      " [[0.6589188290589246]\n",
      "  [0.6617254951066855]\n",
      "  [0.6638844689895784]\n",
      "  [0.6662593402607607]]\n",
      "\n",
      " [[0.6617254951066855]\n",
      "  [0.6638844689895784]\n",
      "  [0.6662593402607607]\n",
      "  [0.6684183141436537]]\n",
      "\n",
      " [[0.6638844689895784]\n",
      "  [0.6662593402607607]\n",
      "  [0.6684183141436537]\n",
      "  [0.6712249801914145]]\n",
      "\n",
      " [[0.6662593402607607]\n",
      "  [0.6684183141436537]\n",
      "  [0.6712249801914145]\n",
      "  [0.6714408775797037]]\n",
      "\n",
      " [[0.6684183141436537]\n",
      "  [0.6712249801914145]\n",
      "  [0.6714408775797037]\n",
      "  [0.673815748850886]]\n",
      "\n",
      " [[0.6712249801914145]\n",
      "  [0.6714408775797037]\n",
      "  [0.673815748850886]\n",
      "  [0.6772701070635146]]\n",
      "\n",
      " [[0.6714408775797037]\n",
      "  [0.673815748850886]\n",
      "  [0.6772701070635146]\n",
      "  [0.681156260052722]]\n",
      "\n",
      " [[0.673815748850886]\n",
      "  [0.6772701070635146]\n",
      "  [0.681156260052722]\n",
      "  [0.6822357469941684]]\n",
      "\n",
      " [[0.6772701070635146]\n",
      "  [0.681156260052722]\n",
      "  [0.6822357469941684]\n",
      "  [0.6830993365473257]]\n",
      "\n",
      " [[0.681156260052722]\n",
      "  [0.6822357469941684]\n",
      "  [0.6830993365473257]\n",
      "  [0.6865536947599543]]\n",
      "\n",
      " [[0.6822357469941684]\n",
      "  [0.6830993365473257]\n",
      "  [0.6865536947599543]\n",
      "  [0.6895762581960045]]\n",
      "\n",
      " [[0.6830993365473257]\n",
      "  [0.6865536947599543]\n",
      "  [0.6895762581960045]\n",
      "  [0.692167026855476]]\n",
      "\n",
      " [[0.6865536947599543]\n",
      "  [0.6895762581960045]\n",
      "  [0.692167026855476]\n",
      "  [0.6947577955149475]]\n",
      "\n",
      " [[0.6895762581960045]\n",
      "  [0.692167026855476]\n",
      "  [0.6947577955149475]\n",
      "  [0.6969167693978404]]\n",
      "\n",
      " [[0.692167026855476]\n",
      "  [0.6947577955149475]\n",
      "  [0.6969167693978404]\n",
      "  [0.698859845892444]]\n",
      "\n",
      " [[0.6947577955149475]\n",
      "  [0.6969167693978404]\n",
      "  [0.698859845892444]\n",
      "  [0.7014506145519156]]\n",
      "\n",
      " [[0.6969167693978404]\n",
      "  [0.698859845892444]\n",
      "  [0.7014506145519156]\n",
      "  [0.7036095884348086]]\n",
      "\n",
      " [[0.698859845892444]\n",
      "  [0.7014506145519156]\n",
      "  [0.7036095884348086]\n",
      "  [0.7055526649294122]]\n",
      "\n",
      " [[0.7014506145519156]\n",
      "  [0.7036095884348086]\n",
      "  [0.7055526649294122]\n",
      "  [0.7072798440357265]]\n",
      "\n",
      " [[0.7036095884348086]\n",
      "  [0.7055526649294122]\n",
      "  [0.7072798440357265]\n",
      "  [0.7081434335888837]]\n",
      "\n",
      " [[0.7055526649294122]\n",
      "  [0.7072798440357265]\n",
      "  [0.7081434335888837]\n",
      "  [0.713279632456286]]\n",
      "\n",
      " [[0.7072798440357265]\n",
      "  [0.7081434335888837]\n",
      "  [0.713279632456286]\n",
      "  [0.7171031752028895]]\n",
      "\n",
      " [[0.7081434335888837]\n",
      "  [0.713279632456286]\n",
      "  [0.7171031752028895]\n",
      "  [0.7182884518645977]]\n",
      "\n",
      " [[0.713279632456286]\n",
      "  [0.7171031752028895]\n",
      "  [0.7182884518645977]\n",
      "  [0.7213563537521884]]\n",
      "\n",
      " [[0.7171031752028895]\n",
      "  [0.7182884518645977]\n",
      "  [0.7213563537521884]\n",
      "  [0.7238845121690561]]\n",
      "\n",
      " [[0.7182884518645977]\n",
      "  [0.7213563537521884]\n",
      "  [0.7238845121690561]\n",
      "  [0.7260909834773728]]\n",
      "\n",
      " [[0.7213563537521884]\n",
      "  [0.7238845121690561]\n",
      "  [0.7260909834773728]\n",
      "  [0.7308536798630345]]\n",
      "\n",
      " [[0.7238845121690561]\n",
      "  [0.7260909834773728]\n",
      "  [0.7308536798630345]\n",
      "  [0.734431099586988]]\n",
      "\n",
      " [[0.7260909834773728]\n",
      "  [0.7308536798630345]\n",
      "  [0.734431099586988]\n",
      "  [0.7369247144217295]]\n",
      "\n",
      " [[0.7308536798630345]\n",
      "  [0.734431099586988]\n",
      "  [0.7369247144217295]\n",
      "  [0.7408367750975315]]\n",
      "\n",
      " [[0.734431099586988]\n",
      "  [0.7369247144217295]\n",
      "  [0.7408367750975315]\n",
      "  [0.7438744513507618]]\n",
      "\n",
      " [[0.7369247144217295]\n",
      "  [0.7408367750975315]\n",
      "  [0.7438744513507618]\n",
      "  [0.7463745431071519]]\n",
      "\n",
      " [[0.7408367750975315]\n",
      "  [0.7438744513507618]\n",
      "  [0.7463745431071519]\n",
      "  [0.7497317474950504]]\n",
      "\n",
      " [[0.7438744513507618]\n",
      "  [0.7463745431071519]\n",
      "  [0.7497317474950504]\n",
      "  [0.7503513729994407]]\n",
      "\n",
      " [[0.7463745431071519]\n",
      "  [0.7497317474950504]\n",
      "  [0.7503513729994407]\n",
      "  [0.7520418495497458]]\n",
      "\n",
      " [[0.7497317474950504]\n",
      "  [0.7503513729994407]\n",
      "  [0.7520418495497458]\n",
      "  [0.753205536472625]]\n",
      "\n",
      " [[0.7503513729994407]\n",
      "  [0.7520418495497458]\n",
      "  [0.753205536472625]\n",
      "  [0.7544706951680004]]\n",
      "\n",
      " [[0.7520418495497458]\n",
      "  [0.753205536472625]\n",
      "  [0.7544706951680004]\n",
      "  [0.756657735711371]]\n",
      "\n",
      " [[0.753205536472625]\n",
      "  [0.7544706951680004]\n",
      "  [0.756657735711371]\n",
      "  [0.757704838044574]]\n",
      "\n",
      " [[0.7544706951680004]\n",
      "  [0.756657735711371]\n",
      "  [0.757704838044574]\n",
      "  [0.759477355602429]]\n",
      "\n",
      " [[0.756657735711371]\n",
      "  [0.757704838044574]\n",
      "  [0.759477355602429]\n",
      "  [0.7613427090372487]]\n",
      "\n",
      " [[0.757704838044574]\n",
      "  [0.759477355602429]\n",
      "  [0.7613427090372487]\n",
      "  [0.7628302420425618]]\n",
      "\n",
      " [[0.759477355602429]\n",
      "  [0.7613427090372487]\n",
      "  [0.7628302420425618]\n",
      "  [0.7648359287797694]]\n",
      "\n",
      " [[0.7613427090372487]\n",
      "  [0.7628302420425618]\n",
      "  [0.7648359287797694]\n",
      "  [0.7676188461148185]]\n",
      "\n",
      " [[0.7628302420425618]\n",
      "  [0.7648359287797694]\n",
      "  [0.7676188461148185]\n",
      "  [0.7699419020128112]]\n",
      "\n",
      " [[0.7648359287797694]\n",
      "  [0.7676188461148185]\n",
      "  [0.7699419020128112]\n",
      "  [0.772163486138308]]\n",
      "\n",
      " [[0.7676188461148185]\n",
      "  [0.7699419020128112]\n",
      "  [0.772163486138308]\n",
      "  [0.7740029318865328]]\n",
      "\n",
      " [[0.7699419020128112]\n",
      "  [0.772163486138308]\n",
      "  [0.7740029318865328]\n",
      "  [0.7769650440538619]]\n",
      "\n",
      " [[0.772163486138308]\n",
      "  [0.7740029318865328]\n",
      "  [0.7769650440538619]\n",
      "  [0.7797328485717308]]\n",
      "\n",
      " [[0.7740029318865328]\n",
      "  [0.7769650440538619]\n",
      "  [0.7797328485717308]\n",
      "  [0.781498889207937]]\n",
      "\n",
      " [[0.7769650440538619]\n",
      "  [0.7797328485717308]\n",
      "  [0.781498889207937]\n",
      "  [0.7832778836874408]]\n",
      "\n",
      " [[0.7797328485717308]\n",
      "  [0.781498889207937]\n",
      "  [0.7832778836874408]\n",
      "  [0.7856311652197941]]\n",
      "\n",
      " [[0.781498889207937]\n",
      "  [0.7832778836874408]\n",
      "  [0.7856311652197941]\n",
      "  [0.7883687441033025]]\n",
      "\n",
      " [[0.7832778836874408]\n",
      "  [0.7856311652197941]\n",
      "  [0.7883687441033025]\n",
      "  [0.7903809077621586]]\n",
      "\n",
      " [[0.7856311652197941]\n",
      "  [0.7883687441033025]\n",
      "  [0.7903809077621586]\n",
      "  [0.7928831584924315]]\n",
      "\n",
      " [[0.7883687441033025]\n",
      "  [0.7903809077621586]\n",
      "  [0.7928831584924315]\n",
      "  [0.7943102402290237]]\n",
      "\n",
      " [[0.7903809077621586]\n",
      "  [0.7928831584924315]\n",
      "  [0.7943102402290237]\n",
      "  [0.7978315266320221]]\n",
      "\n",
      " [[0.7928831584924315]\n",
      "  [0.7943102402290237]\n",
      "  [0.7978315266320221]\n",
      "  [0.8012124797326327]]\n",
      "\n",
      " [[0.7943102402290237]\n",
      "  [0.7978315266320221]\n",
      "  [0.8012124797326327]\n",
      "  [0.8042436790642141]]\n",
      "\n",
      " [[0.7978315266320221]\n",
      "  [0.8012124797326327]\n",
      "  [0.8042436790642141]\n",
      "  [0.8061370991595113]]\n",
      "\n",
      " [[0.8012124797326327]\n",
      "  [0.8042436790642141]\n",
      "  [0.8061370991595113]\n",
      "  [0.8072273809703722]]\n",
      "\n",
      " [[0.8042436790642141]\n",
      "  [0.8061370991595113]\n",
      "  [0.8072273809703722]\n",
      "  [0.8095892983982572]]\n",
      "\n",
      " [[0.8061370991595113]\n",
      "  [0.8072273809703722]\n",
      "  [0.8095892983982572]\n",
      "  [0.8094489650958692]]\n",
      "\n",
      " [[0.8072273809703722]\n",
      "  [0.8095892983982572]\n",
      "  [0.8094489650958692]\n",
      "  [0.8114697646502569]]\n",
      "\n",
      " [[0.8095892983982572]\n",
      "  [0.8094489650958692]\n",
      "  [0.8114697646502569]\n",
      "  [0.8162821174352253]]\n",
      "\n",
      " [[0.8094489650958692]\n",
      "  [0.8114697646502569]\n",
      "  [0.8162821174352253]\n",
      "  [0.8176595427725109]]\n",
      "\n",
      " [[0.8114697646502569]\n",
      "  [0.8162821174352253]\n",
      "  [0.8176595427725109]\n",
      "  [0.8189743578671927]]\n",
      "\n",
      " [[0.8162821174352253]\n",
      "  [0.8176595427725109]\n",
      "  [0.8189743578671927]\n",
      "  [0.8210879932985449]]\n",
      "\n",
      " [[0.8176595427725109]\n",
      "  [0.8189743578671927]\n",
      "  [0.8210879932985449]\n",
      "  [0.822292700725199]]\n",
      "\n",
      " [[0.8189743578671927]\n",
      "  [0.8210879932985449]\n",
      "  [0.822292700725199]\n",
      "  [0.8254145769598624]]\n",
      "\n",
      " [[0.8210879932985449]\n",
      "  [0.822292700725199]\n",
      "  [0.8254145769598624]\n",
      "  [0.8270813047974557]]\n",
      "\n",
      " [[0.822292700725199]\n",
      "  [0.8254145769598624]\n",
      "  [0.8270813047974557]\n",
      "  [0.8301276169462175]]\n",
      "\n",
      " [[0.8254145769598624]\n",
      "  [0.8270813047974557]\n",
      "  [0.8301276169462175]\n",
      "  [0.8323297703067685]]\n",
      "\n",
      " [[0.8270813047974557]\n",
      "  [0.8301276169462175]\n",
      "  [0.8323297703067685]\n",
      "  [0.833830257155379]]\n",
      "\n",
      " [[0.8301276169462175]\n",
      "  [0.8323297703067685]\n",
      "  [0.833830257155379]\n",
      "  [0.8362547848258678]]\n",
      "\n",
      " [[0.8323297703067685]\n",
      "  [0.833830257155379]\n",
      "  [0.8362547848258678]\n",
      "  [0.8380855946785609]]\n",
      "\n",
      " [[0.833830257155379]\n",
      "  [0.8362547848258678]\n",
      "  [0.8380855946785609]\n",
      "  [0.8398451583931187]]\n",
      "\n",
      " [[0.8362547848258678]\n",
      "  [0.8380855946785609]\n",
      "  [0.8398451583931187]\n",
      "  [0.843873803658597]]\n",
      "\n",
      " [[0.8380855946785609]\n",
      "  [0.8398451583931187]\n",
      "  [0.843873803658597]\n",
      "  [0.8475073567035057]]\n",
      "\n",
      " [[0.8398451583931187]\n",
      "  [0.843873803658597]\n",
      "  [0.8475073567035057]\n",
      "  [0.8507004790763044]]\n",
      "\n",
      " [[0.843873803658597]\n",
      "  [0.8475073567035057]\n",
      "  [0.8507004790763044]\n",
      "  [0.8532135246759918]]\n",
      "\n",
      " [[0.8475073567035057]\n",
      "  [0.8507004790763044]\n",
      "  [0.8532135246759918]\n",
      "  [0.8542843757219066]]\n",
      "\n",
      " [[0.8507004790763044]\n",
      "  [0.8532135246759918]\n",
      "  [0.8542843757219066]\n",
      "  [0.8572356930198213]]\n",
      "\n",
      " [[0.8532135246759918]\n",
      "  [0.8542843757219066]\n",
      "  [0.8572356930198213]\n",
      "  [0.859865323209185]]\n",
      "\n",
      " [[0.8542843757219066]\n",
      "  [0.8572356930198213]\n",
      "  [0.859865323209185]\n",
      "  [0.8628641379325233]]\n",
      "\n",
      " [[0.8572356930198213]\n",
      "  [0.859865323209185]\n",
      "  [0.8628641379325233]\n",
      "  [0.8680996495985386]]\n",
      "\n",
      " [[0.859865323209185]\n",
      "  [0.8628641379325233]\n",
      "  [0.8680996495985386]\n",
      "  [0.8718022898077]]\n",
      "\n",
      " [[0.8628641379325233]\n",
      "  [0.8680996495985386]\n",
      "  [0.8718022898077]\n",
      "  [0.872991884417174]]\n",
      "\n",
      " [[0.8680996495985386]\n",
      "  [0.8718022898077]\n",
      "  [0.872991884417174]\n",
      "  [0.8749306429640119]]\n",
      "\n",
      " [[0.8718022898077]\n",
      "  [0.872991884417174]\n",
      "  [0.8749306429640119]\n",
      "  [0.8762627298497568]]\n",
      "\n",
      " [[0.872991884417174]\n",
      "  [0.8749306429640119]\n",
      "  [0.8762627298497568]\n",
      "  [0.8772105193843468]]\n",
      "\n",
      " [[0.8749306429640119]\n",
      "  [0.8762627298497568]\n",
      "  [0.8772105193843468]\n",
      "  [0.8789636061772558]]\n",
      "\n",
      " [[0.8762627298497568]\n",
      "  [0.8772105193843468]\n",
      "  [0.8789636061772558]\n",
      "  [0.8805806776155427]]\n",
      "\n",
      " [[0.8772105193843468]\n",
      "  [0.8789636061772558]\n",
      "  [0.8805806776155427]\n",
      "  [0.8816817542958181]]\n",
      "\n",
      " [[0.8789636061772558]\n",
      "  [0.8805806776155427]\n",
      "  [0.8816817542958181]\n",
      "  [0.8845790972466604]]\n",
      "\n",
      " [[0.8805806776155427]\n",
      "  [0.8816817542958181]\n",
      "  [0.8845790972466604]\n",
      "  [0.8844301280487408]]\n",
      "\n",
      " [[0.8816817542958181]\n",
      "  [0.8845790972466604]\n",
      "  [0.8844301280487408]\n",
      "  [0.8828389642970487]]\n",
      "\n",
      " [[0.8845790972466604]\n",
      "  [0.8844301280487408]\n",
      "  [0.8828389642970487]\n",
      "  [0.887428942772079]]\n",
      "\n",
      " [[0.8844301280487408]\n",
      "  [0.8828389642970487]\n",
      "  [0.887428942772079]\n",
      "  [0.8891539629045105]]\n",
      "\n",
      " [[0.8828389642970487]\n",
      "  [0.887428942772079]\n",
      "  [0.8891539629045105]\n",
      "  [0.8938195054654422]]\n",
      "\n",
      " [[0.887428942772079]\n",
      "  [0.8891539629045105]\n",
      "  [0.8938195054654422]\n",
      "  [0.8963411869606611]]\n",
      "\n",
      " [[0.8891539629045105]\n",
      "  [0.8938195054654422]\n",
      "  [0.8963411869606611]\n",
      "  [0.897191822670521]]\n",
      "\n",
      " [[0.8938195054654422]\n",
      "  [0.8963411869606611]\n",
      "  [0.897191822670521]\n",
      "  [0.8967924125021858]]\n",
      "\n",
      " [[0.8963411869606611]\n",
      "  [0.897191822670521]\n",
      "  [0.8967924125021858]\n",
      "  [0.8973127252079629]]\n",
      "\n",
      " [[0.897191822670521]\n",
      "  [0.8967924125021858]\n",
      "  [0.8973127252079629]\n",
      "  [0.8999812169272186]]\n",
      "\n",
      " [[0.8967924125021858]\n",
      "  [0.8973127252079629]\n",
      "  [0.8999812169272186]\n",
      "  [0.9022265497654273]]\n",
      "\n",
      " [[0.8973127252079629]\n",
      "  [0.8999812169272186]\n",
      "  [0.9022265497654273]\n",
      "  [0.9046381235926187]]\n",
      "\n",
      " [[0.8999812169272186]\n",
      "  [0.9022265497654273]\n",
      "  [0.9046381235926187]\n",
      "  [0.9066502872514749]]\n",
      "\n",
      " [[0.9022265497654273]\n",
      "  [0.9046381235926187]\n",
      "  [0.9066502872514749]\n",
      "  [0.90876608165671]]\n",
      "\n",
      " [[0.9046381235926187]\n",
      "  [0.9066502872514749]\n",
      "  [0.90876608165671]\n",
      "  [0.911387075950542]]\n",
      "\n",
      " [[0.9066502872514749]\n",
      "  [0.90876608165671]\n",
      "  [0.911387075950542]\n",
      "  [0.9130797114747302]]\n",
      "\n",
      " [[0.90876608165671]\n",
      "  [0.911387075950542]\n",
      "  [0.9130797114747302]\n",
      "  [0.9128487012692604]]\n",
      "\n",
      " [[0.911387075950542]\n",
      "  [0.9130797114747302]\n",
      "  [0.9128487012692604]\n",
      "  [0.9146622393308906]]\n",
      "\n",
      " [[0.9130797114747302]\n",
      "  [0.9128487012692604]\n",
      "  [0.9146622393308906]\n",
      "  [0.9165189568701786]]\n",
      "\n",
      " [[0.9128487012692604]\n",
      "  [0.9146622393308906]\n",
      "  [0.9165189568701786]\n",
      "  [0.9199171817618521]]\n",
      "\n",
      " [[0.9146622393308906]\n",
      "  [0.9165189568701786]\n",
      "  [0.9199171817618521]\n",
      "  [0.9246604473825677]]\n",
      "\n",
      " [[0.9165189568701786]\n",
      "  [0.9199171817618521]\n",
      "  [0.9246604473825677]\n",
      "  [0.924615108931027]]\n",
      "\n",
      " [[0.9199171817618521]\n",
      "  [0.9246604473825677]\n",
      "  [0.924615108931027]\n",
      "  [0.9237946988555278]]\n",
      "\n",
      " [[0.9246604473825677]\n",
      "  [0.924615108931027]\n",
      "  [0.9237946988555278]\n",
      "  [0.9277391441395731]]\n",
      "\n",
      " [[0.924615108931027]\n",
      "  [0.9237946988555278]\n",
      "  [0.9277391441395731]\n",
      "  [0.9339461940528904]]\n",
      "\n",
      " [[0.9237946988555278]\n",
      "  [0.9277391441395731]\n",
      "  [0.9339461940528904]\n",
      "  [0.9353538450245366]]\n",
      "\n",
      " [[0.9277391441395731]\n",
      "  [0.9339461940528904]\n",
      "  [0.9353538450245366]\n",
      "  [0.9349738656211474]]\n",
      "\n",
      " [[0.9339461940528904]\n",
      "  [0.9353538450245366]\n",
      "  [0.9349738656211474]\n",
      "  [0.9365067370780014]]\n",
      "\n",
      " [[0.9353538450245366]\n",
      "  [0.9349738656211474]\n",
      "  [0.9365067370780014]\n",
      "  [0.9359043833646742]]\n",
      "\n",
      " [[0.9349738656211474]\n",
      "  [0.9365067370780014]\n",
      "  [0.9359043833646742]\n",
      "  [0.9376358804187542]]\n",
      "\n",
      " [[0.9365067370780014]\n",
      "  [0.9359043833646742]\n",
      "  [0.9376358804187542]\n",
      "  [0.9447237916762918]]\n",
      "\n",
      " [[0.9359043833646742]\n",
      "  [0.9376358804187542]\n",
      "  [0.9447237916762918]\n",
      "  [0.9478910063624957]]\n",
      "\n",
      " [[0.9376358804187542]\n",
      "  [0.9447237916762918]\n",
      "  [0.9478910063624957]\n",
      "  [0.9492943393863762]]\n",
      "\n",
      " [[0.9447237916762918]\n",
      "  [0.9478910063624957]\n",
      "  [0.9492943393863762]\n",
      "  [0.9531977641666466]]\n",
      "\n",
      " [[0.9478910063624957]\n",
      "  [0.9492943393863762]\n",
      "  [0.9531977641666466]\n",
      "  [0.9571724350850526]]\n",
      "\n",
      " [[0.9492943393863762]\n",
      "  [0.9531977641666466]\n",
      "  [0.9571724350850526]\n",
      "  [0.9592558448820442]]\n",
      "\n",
      " [[0.9531977641666466]\n",
      "  [0.9571724350850526]\n",
      "  [0.9592558448820442]\n",
      "  [0.9626929313036098]]\n",
      "\n",
      " [[0.9571724350850526]\n",
      "  [0.9592558448820442]\n",
      "  [0.9626929313036098]\n",
      "  [0.965389489683343]]\n",
      "\n",
      " [[0.9592558448820442]\n",
      "  [0.9626929313036098]\n",
      "  [0.965389489683343]\n",
      "  [0.9691288324485137]]\n",
      "\n",
      " [[0.9626929313036098]\n",
      "  [0.965389489683343]\n",
      "  [0.9691288324485137]\n",
      "  [0.97416355954342]]\n",
      "\n",
      " [[0.965389489683343]\n",
      "  [0.9691288324485137]\n",
      "  [0.97416355954342]\n",
      "  [0.983034783228227]]\n",
      "\n",
      " [[0.9691288324485137]\n",
      "  [0.97416355954342]\n",
      "  [0.983034783228227]\n",
      "  [0.9848353674465599]]\n",
      "\n",
      " [[0.97416355954342]\n",
      "  [0.983034783228227]\n",
      "  [0.9848353674465599]\n",
      "  [0.986221428679377]]\n",
      "\n",
      " [[0.983034783228227]\n",
      "  [0.9848353674465599]\n",
      "  [0.986221428679377]\n",
      "  [0.9867460593329201]]\n",
      "\n",
      " [[0.9848353674465599]\n",
      "  [0.986221428679377]\n",
      "  [0.9867460593329201]\n",
      "  [0.989105817786922]]\n",
      "\n",
      " [[0.986221428679377]\n",
      "  [0.9867460593329201]\n",
      "  [0.989105817786922]\n",
      "  [0.991079119915886]]\n",
      "\n",
      " [[0.9867460593329201]\n",
      "  [0.989105817786922]\n",
      "  [0.991079119915886]\n",
      "  [0.9921305401968551]]\n",
      "\n",
      " [[0.989105817786922]\n",
      "  [0.991079119915886]\n",
      "  [0.9921305401968551]\n",
      "  [0.9935921655155736]]\n",
      "\n",
      " [[0.991079119915886]\n",
      "  [0.9921305401968551]\n",
      "  [0.9935921655155736]\n",
      "  [0.991970776129521]]\n",
      "\n",
      " [[0.9921305401968551]\n",
      "  [0.9935921655155736]\n",
      "  [0.991970776129521]\n",
      "  [0.9918628274353763]]]\n"
     ]
    }
   ],
   "source": [
    "def data_reshape(d1, splitfactor,prev):\n",
    "    # split into train and test sets\n",
    "    data_train,data_test = int(len(d1) *splitfactor ),int(len(d1) *(1-splitfactor))\n",
    "    train, test = d1.iloc[0:data_train,:], d1.iloc[data_train:len(d1),:]\n",
    "\n",
    "    # reshape into X=t and Y=t+1\n",
    "    train_X, train_Y = Get_rolling_window_data(train, prev)\n",
    "    testX, testY=Get_rolling_window_data(test, prev)\n",
    "   \n",
    "    # reshape input to be [samples, time steps, features]\n",
    "    train_X = numpy.reshape(train_X, (train_X.shape[0], train_X.shape[1], 1))\n",
    "    testX = numpy.reshape(testX, (testX.shape[0], testX.shape[1], 1))\n",
    "    return train_X,testX, train_Y, testY\n",
    "\n",
    "train_X,testX, train_Y, testY=data_reshape(data,splitfactor,prev)\n",
    "print (len(train_X),len(testX))\n",
    "print(train_X,testX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "6s - loss: 4.6788e-05\n",
      "Epoch 2/20\n",
      "4s - loss: 0.0011\n",
      "Epoch 3/20\n",
      "4s - loss: 0.0013\n",
      "Epoch 4/20\n",
      "4s - loss: 6.0478e-04\n",
      "Epoch 5/20\n",
      "4s - loss: 1.5957e-04\n",
      "Epoch 6/20\n",
      "4s - loss: 1.2154e-04\n",
      "Epoch 7/20\n",
      "3s - loss: 8.6188e-05\n",
      "Epoch 8/20\n",
      "3s - loss: 5.1209e-05\n",
      "Epoch 9/20\n",
      "3s - loss: 1.6221e-04\n",
      "Epoch 10/20\n",
      "3s - loss: 9.6188e-05\n",
      "Epoch 11/20\n",
      "3s - loss: 7.9686e-05\n",
      "Epoch 12/20\n",
      "3s - loss: 7.2339e-05\n",
      "Epoch 13/20\n",
      "3s - loss: 3.1394e-05\n",
      "Epoch 14/20\n",
      "3s - loss: 3.0618e-05\n",
      "Epoch 15/20\n",
      "4s - loss: 3.4925e-05\n",
      "Epoch 16/20\n",
      "3s - loss: 3.6282e-05\n",
      "Epoch 17/20\n",
      "4s - loss: 1.4610e-05\n",
      "Epoch 18/20\n",
      "3s - loss: 1.6627e-05\n",
      "Epoch 19/20\n",
      "4s - loss: 4.9784e-05\n",
      "Epoch 20/20\n",
      "4s - loss: 4.8547e-05\n"
     ]
    }
   ],
   "source": [
    "def build_model_FINAL(train_X, train_Y,prev):\n",
    "    model = Sequential()\n",
    "    batch_size=1\n",
    "    model.add(SimpleRNN(4, batch_input_shape=(batch_size, prev, 1), stateful=True, return_sequences=True))\n",
    "    model.add(LSTM(4, batch_input_shape=(batch_size,prev, 1), stateful=True, return_sequences=True))\n",
    "    model.add(LSTM(6,batch_input_shape=(batch_size,prev,1),stateful=True))\n",
    "    model.add(Dense(1))\n",
    "    model.compile(loss='mean_squared_error', optimizer='adam')\n",
    "    for i in range(1):\n",
    "        model.fit(train_X, train_Y, epochs=20, batch_size=batch_size, verbose=2,shuffle=False)\n",
    "        model.reset_states()\n",
    "    return model\n",
    "\n",
    "model=build_model_FINAL(train_X, train_Y,prev)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "batch_size=1\n",
    "def Get_Inverse_transform(x):\n",
    "    fit_1=scaler.fit(x)\n",
    "    return scaler.inverse_transform(x)\n",
    "# make predictions\n",
    "predtr = Get_Inverse_transform(model.predict(train_X,batch_size))\n",
    "predtr[predtr < 0] = 0\n",
    "train_Y =  Get_Inverse_transform([train_Y])\n",
    "predte,testY = Get_Inverse_transform(model.predict(testX, batch_size)),Get_Inverse_transform([testY])\n",
    "predte[predte < 0] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Score: 0.32 RMSE\n",
      "Test Score: 1.02 RMSE\n",
      "Train Score: 0.21 MAE\n",
      "Test Score: 1.00 MAE\n"
     ]
    }
   ],
   "source": [
    "def Find_RMSE_MAE(x,x1,i):\n",
    "    if i==1:\n",
    "        return math.sqrt(mean_squared_error(x, x1))\n",
    "    else:\n",
    "        return mean_absolute_error(x, x1)\n",
    "# calculate root mean squared error\n",
    "trerr= Find_RMSE_MAE(train_Y[0],predtr[:,0],1)\n",
    "teerr =Find_RMSE_MAE(testY[0], predte[:,0],1)\n",
    "print('Train Score: %.2f RMSE' %(trerr))\n",
    "print ('Test Score: %.2f RMSE' % (teerr))\n",
    "# calculate mean absloute error\n",
    "meanerrtr = Find_RMSE_MAE(train_Y[0],predtr[:,0],0)\n",
    "meanerrte = Find_RMSE_MAE(testY[0], predte[:,0],0)\n",
    "print ('Train Score: %.2f MAE' %(meanerrtr))\n",
    "print ('Test Score: %.2f MAE' % (meanerrte))"
   ]
  }
 ],
 "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": "3.6.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}