Linear_Regression.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "import pandas as pd\n",
    "import pyspark as spark\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from sklearn.metrics import mean_squared_error\n",
    "sc = spark.SparkContext()\n",
    "sql = spark.SQLContext(sc)\n",
    "from pyspark.sql.types import *\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import LSTM\n",
    "from elephas.utils.rdd_utils import to_simple_rdd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>\n",
    "<span style=\"color:blue\">\n",
    "Below function is to transform dataframes to include time lags in the feature vector<br>\n",
    "</span>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):\n",
    "    n_vars = 1 if type(data) is list else data.shape[1]\n",
    "    df = pd.DataFrame(data)\n",
    "    cols, names = list(), list()\n",
    "    # input sequence (t-n, ... t-1)\n",
    "    for i in range(n_in, 0, -1):\n",
    "        cols.append(df.shift(i))\n",
    "        names += [('var%d(t-%d)' % (j+1, i)) for j in range(n_vars)]\n",
    "    # forecast sequence (t, t+1, ... t+n)\n",
    "    for i in range(0, n_out):\n",
    "        cols.append(df.shift(-i))\n",
    "        if i == 0:\n",
    "            names += [('var%d(t)' % (j+1)) for j in range(n_vars)]\n",
    "        else:\n",
    "            names += [('var%d(t+%d)' % (j+1, i)) for j in range(n_vars)]\n",
    "    # put it all together\n",
    "    agg = pd.concat(cols, axis=1)\n",
    "    agg.columns = names\n",
    "    # drop rows with NaN values\n",
    "    if dropnan:\n",
    "        agg.dropna(inplace=True)\n",
    "    return agg"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Linear Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>DateTime</th>\n",
       "      <th>Negative</th>\n",
       "      <th>Neutral</th>\n",
       "      <th>Positive</th>\n",
       "      <th>Compound</th>\n",
       "      <th>Price</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2017-10-31T05:00:00.000-04:00</td>\n",
       "      <td>0.032105</td>\n",
       "      <td>0.888699</td>\n",
       "      <td>0.076390</td>\n",
       "      <td>0.095283</td>\n",
       "      <td>6158.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2017-10-31T06:00:00.000-04:00</td>\n",
       "      <td>0.035658</td>\n",
       "      <td>0.862046</td>\n",
       "      <td>0.101670</td>\n",
       "      <td>0.130807</td>\n",
       "      <td>6105.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2017-10-31T07:00:00.000-04:00</td>\n",
       "      <td>0.034356</td>\n",
       "      <td>0.864839</td>\n",
       "      <td>0.099413</td>\n",
       "      <td>0.130343</td>\n",
       "      <td>6094.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2017-10-31T08:00:00.000-04:00</td>\n",
       "      <td>0.036202</td>\n",
       "      <td>0.879061</td>\n",
       "      <td>0.084736</td>\n",
       "      <td>0.110576</td>\n",
       "      <td>6125.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2017-10-31T09:00:00.000-04:00</td>\n",
       "      <td>0.043838</td>\n",
       "      <td>0.862753</td>\n",
       "      <td>0.092850</td>\n",
       "      <td>0.108861</td>\n",
       "      <td>6165.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        DateTime  Negative   Neutral  Positive  Compound  \\\n",
       "0  2017-10-31T05:00:00.000-04:00  0.032105  0.888699  0.076390  0.095283   \n",
       "1  2017-10-31T06:00:00.000-04:00  0.035658  0.862046  0.101670  0.130807   \n",
       "2  2017-10-31T07:00:00.000-04:00  0.034356  0.864839  0.099413  0.130343   \n",
       "3  2017-10-31T08:00:00.000-04:00  0.036202  0.879061  0.084736  0.110576   \n",
       "4  2017-10-31T09:00:00.000-04:00  0.043838  0.862753  0.092850  0.108861   \n",
       "\n",
       "     Price  \n",
       "0  6158.76  \n",
       "1  6105.90  \n",
       "2  6094.36  \n",
       "3  6125.13  \n",
       "4  6165.00  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1 = pd.read_csv('/Users/harishpuvvada/Desktop/PBDA/IPynbSpark/results_235.csv',header=None)\n",
    "df1.columns = ['DateTime','Negative','Neutral','Positive','Compound','Price']\n",
    "df1.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exploratory analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x122f5da20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "values = df1.values\n",
    "# specify columns to plot\n",
    "groups = [ 1,2,3,4,5]\n",
    "i = 1\n",
    "# plot each column\n",
    "plt.figure()\n",
    "for group in groups:\n",
    "    plt.subplot(len(groups), 1, i)\n",
    "    plt.plot(values[:, group])\n",
    "    plt.title(df1.columns[group], y=0.5, loc='right')\n",
    "    i += 1\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Price</th>\n",
       "      <th>Negative</th>\n",
       "      <th>Neutral</th>\n",
       "      <th>Positive</th>\n",
       "      <th>Compound</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DateTime</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2017-10-31T05:00:00.000-04:00</th>\n",
       "      <td>6158.76</td>\n",
       "      <td>0.032105</td>\n",
       "      <td>0.888699</td>\n",
       "      <td>0.076390</td>\n",
       "      <td>0.095283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-31T06:00:00.000-04:00</th>\n",
       "      <td>6105.90</td>\n",
       "      <td>0.035658</td>\n",
       "      <td>0.862046</td>\n",
       "      <td>0.101670</td>\n",
       "      <td>0.130807</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-31T07:00:00.000-04:00</th>\n",
       "      <td>6094.36</td>\n",
       "      <td>0.034356</td>\n",
       "      <td>0.864839</td>\n",
       "      <td>0.099413</td>\n",
       "      <td>0.130343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-31T08:00:00.000-04:00</th>\n",
       "      <td>6125.13</td>\n",
       "      <td>0.036202</td>\n",
       "      <td>0.879061</td>\n",
       "      <td>0.084736</td>\n",
       "      <td>0.110576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-10-31T09:00:00.000-04:00</th>\n",
       "      <td>6165.00</td>\n",
       "      <td>0.043838</td>\n",
       "      <td>0.862753</td>\n",
       "      <td>0.092850</td>\n",
       "      <td>0.108861</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 Price  Negative   Neutral  Positive  Compound\n",
       "DateTime                                                                      \n",
       "2017-10-31T05:00:00.000-04:00  6158.76  0.032105  0.888699  0.076390  0.095283\n",
       "2017-10-31T06:00:00.000-04:00  6105.90  0.035658  0.862046  0.101670  0.130807\n",
       "2017-10-31T07:00:00.000-04:00  6094.36  0.034356  0.864839  0.099413  0.130343\n",
       "2017-10-31T08:00:00.000-04:00  6125.13  0.036202  0.879061  0.084736  0.110576\n",
       "2017-10-31T09:00:00.000-04:00  6165.00  0.043838  0.862753  0.092850  0.108861"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.set_index('DateTime',inplace=True)  #making DateTime as index. Because we are doing Time series Analysis\n",
    "cols = df1.columns.tolist()\n",
    "cols = cols[-1:] + cols[:-1]   #just making the price as the first column\n",
    "df1 = df1[cols]\n",
    "df1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>var1(t-3)</th>\n",
       "      <th>var2(t-3)</th>\n",
       "      <th>var3(t-3)</th>\n",
       "      <th>var4(t-3)</th>\n",
       "      <th>var5(t-3)</th>\n",
       "      <th>var1(t-2)</th>\n",
       "      <th>var2(t-2)</th>\n",
       "      <th>var3(t-2)</th>\n",
       "      <th>var4(t-2)</th>\n",
       "      <th>var5(t-2)</th>\n",
       "      <th>var1(t-1)</th>\n",
       "      <th>var2(t-1)</th>\n",
       "      <th>var3(t-1)</th>\n",
       "      <th>var4(t-1)</th>\n",
       "      <th>var5(t-1)</th>\n",
       "      <th>var1(t)</th>\n",
       "      <th>var2(t)</th>\n",
       "      <th>var3(t)</th>\n",
       "      <th>var4(t)</th>\n",
       "      <th>var5(t)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.039567</td>\n",
       "      <td>0.078779</td>\n",
       "      <td>0.831199</td>\n",
       "      <td>0.195240</td>\n",
       "      <td>0.380713</td>\n",
       "      <td>0.007090</td>\n",
       "      <td>0.128849</td>\n",
       "      <td>0.579299</td>\n",
       "      <td>0.503813</td>\n",
       "      <td>0.514548</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.110492</td>\n",
       "      <td>0.605697</td>\n",
       "      <td>0.476259</td>\n",
       "      <td>0.512798</td>\n",
       "      <td>0.018905</td>\n",
       "      <td>0.136519</td>\n",
       "      <td>0.740108</td>\n",
       "      <td>0.297115</td>\n",
       "      <td>0.438328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.007090</td>\n",
       "      <td>0.128849</td>\n",
       "      <td>0.579299</td>\n",
       "      <td>0.503813</td>\n",
       "      <td>0.514548</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.110492</td>\n",
       "      <td>0.605697</td>\n",
       "      <td>0.476259</td>\n",
       "      <td>0.512798</td>\n",
       "      <td>0.018905</td>\n",
       "      <td>0.136519</td>\n",
       "      <td>0.740108</td>\n",
       "      <td>0.297115</td>\n",
       "      <td>0.438328</td>\n",
       "      <td>0.043401</td>\n",
       "      <td>0.244122</td>\n",
       "      <td>0.585976</td>\n",
       "      <td>0.396158</td>\n",
       "      <td>0.431867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.110492</td>\n",
       "      <td>0.605697</td>\n",
       "      <td>0.476259</td>\n",
       "      <td>0.512798</td>\n",
       "      <td>0.018905</td>\n",
       "      <td>0.136519</td>\n",
       "      <td>0.740108</td>\n",
       "      <td>0.297115</td>\n",
       "      <td>0.438328</td>\n",
       "      <td>0.043401</td>\n",
       "      <td>0.244122</td>\n",
       "      <td>0.585976</td>\n",
       "      <td>0.396158</td>\n",
       "      <td>0.431867</td>\n",
       "      <td>0.046946</td>\n",
       "      <td>0.186121</td>\n",
       "      <td>0.513127</td>\n",
       "      <td>0.540516</td>\n",
       "      <td>0.501238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.018905</td>\n",
       "      <td>0.136519</td>\n",
       "      <td>0.740108</td>\n",
       "      <td>0.297115</td>\n",
       "      <td>0.438328</td>\n",
       "      <td>0.043401</td>\n",
       "      <td>0.244122</td>\n",
       "      <td>0.585976</td>\n",
       "      <td>0.396158</td>\n",
       "      <td>0.431867</td>\n",
       "      <td>0.046946</td>\n",
       "      <td>0.186121</td>\n",
       "      <td>0.513127</td>\n",
       "      <td>0.540516</td>\n",
       "      <td>0.501238</td>\n",
       "      <td>0.085634</td>\n",
       "      <td>0.117236</td>\n",
       "      <td>0.544642</td>\n",
       "      <td>0.553859</td>\n",
       "      <td>0.596469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.043401</td>\n",
       "      <td>0.244122</td>\n",
       "      <td>0.585976</td>\n",
       "      <td>0.396158</td>\n",
       "      <td>0.431867</td>\n",
       "      <td>0.046946</td>\n",
       "      <td>0.186121</td>\n",
       "      <td>0.513127</td>\n",
       "      <td>0.540516</td>\n",
       "      <td>0.501238</td>\n",
       "      <td>0.085634</td>\n",
       "      <td>0.117236</td>\n",
       "      <td>0.544642</td>\n",
       "      <td>0.553859</td>\n",
       "      <td>0.596469</td>\n",
       "      <td>0.065537</td>\n",
       "      <td>0.100541</td>\n",
       "      <td>0.701571</td>\n",
       "      <td>0.372117</td>\n",
       "      <td>0.466619</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   var1(t-3)  var2(t-3)  var3(t-3)  var4(t-3)  var5(t-3)  var1(t-2)  \\\n",
       "3   0.039567   0.078779   0.831199   0.195240   0.380713   0.007090   \n",
       "4   0.007090   0.128849   0.579299   0.503813   0.514548   0.000000   \n",
       "5   0.000000   0.110492   0.605697   0.476259   0.512798   0.018905   \n",
       "6   0.018905   0.136519   0.740108   0.297115   0.438328   0.043401   \n",
       "7   0.043401   0.244122   0.585976   0.396158   0.431867   0.046946   \n",
       "\n",
       "   var2(t-2)  var3(t-2)  var4(t-2)  var5(t-2)  var1(t-1)  var2(t-1)  \\\n",
       "3   0.128849   0.579299   0.503813   0.514548   0.000000   0.110492   \n",
       "4   0.110492   0.605697   0.476259   0.512798   0.018905   0.136519   \n",
       "5   0.136519   0.740108   0.297115   0.438328   0.043401   0.244122   \n",
       "6   0.244122   0.585976   0.396158   0.431867   0.046946   0.186121   \n",
       "7   0.186121   0.513127   0.540516   0.501238   0.085634   0.117236   \n",
       "\n",
       "   var3(t-1)  var4(t-1)  var5(t-1)   var1(t)   var2(t)   var3(t)   var4(t)  \\\n",
       "3   0.605697   0.476259   0.512798  0.018905  0.136519  0.740108  0.297115   \n",
       "4   0.740108   0.297115   0.438328  0.043401  0.244122  0.585976  0.396158   \n",
       "5   0.585976   0.396158   0.431867  0.046946  0.186121  0.513127  0.540516   \n",
       "6   0.513127   0.540516   0.501238  0.085634  0.117236  0.544642  0.553859   \n",
       "7   0.544642   0.553859   0.596469  0.065537  0.100541  0.701571  0.372117   \n",
       "\n",
       "    var5(t)  \n",
       "3  0.438328  \n",
       "4  0.431867  \n",
       "5  0.501238  \n",
       "6  0.596469  \n",
       "7  0.466619  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = MinMaxScaler(feature_range=(0, 1)) #scaling values from 0 to 1\n",
    "scaled = scaler.fit_transform(df1.values)\n",
    "\n",
    "n_hours = 3\n",
    "n_features = 5              \n",
    "n_obs = n_hours*n_features\n",
    "\n",
    "reframed = series_to_supervised(scaled, n_hours, 1)  #t-3,t-2,t-1 values of all the columns. you can change n_hours for this\n",
    "reframed.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>var1(t-3)</th>\n",
       "      <th>var2(t-3)</th>\n",
       "      <th>var3(t-3)</th>\n",
       "      <th>var4(t-3)</th>\n",
       "      <th>var5(t-3)</th>\n",
       "      <th>var1(t-2)</th>\n",
       "      <th>var2(t-2)</th>\n",
       "      <th>var3(t-2)</th>\n",
       "      <th>var4(t-2)</th>\n",
       "      <th>var5(t-2)</th>\n",
       "      <th>var1(t-1)</th>\n",
       "      <th>var2(t-1)</th>\n",
       "      <th>var3(t-1)</th>\n",
       "      <th>var4(t-1)</th>\n",
       "      <th>var5(t-1)</th>\n",
       "      <th>var1(t)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.039567</td>\n",
       "      <td>0.078779</td>\n",
       "      <td>0.831199</td>\n",
       "      <td>0.195240</td>\n",
       "      <td>0.380713</td>\n",
       "      <td>0.007090</td>\n",
       "      <td>0.128849</td>\n",
       "      <td>0.579299</td>\n",
       "      <td>0.503813</td>\n",
       "      <td>0.514548</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.110492</td>\n",
       "      <td>0.605697</td>\n",
       "      <td>0.476259</td>\n",
       "      <td>0.512798</td>\n",
       "      <td>0.018905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.007090</td>\n",
       "      <td>0.128849</td>\n",
       "      <td>0.579299</td>\n",
       "      <td>0.503813</td>\n",
       "      <td>0.514548</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.110492</td>\n",
       "      <td>0.605697</td>\n",
       "      <td>0.476259</td>\n",
       "      <td>0.512798</td>\n",
       "      <td>0.018905</td>\n",
       "      <td>0.136519</td>\n",
       "      <td>0.740108</td>\n",
       "      <td>0.297115</td>\n",
       "      <td>0.438328</td>\n",
       "      <td>0.043401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.110492</td>\n",
       "      <td>0.605697</td>\n",
       "      <td>0.476259</td>\n",
       "      <td>0.512798</td>\n",
       "      <td>0.018905</td>\n",
       "      <td>0.136519</td>\n",
       "      <td>0.740108</td>\n",
       "      <td>0.297115</td>\n",
       "      <td>0.438328</td>\n",
       "      <td>0.043401</td>\n",
       "      <td>0.244122</td>\n",
       "      <td>0.585976</td>\n",
       "      <td>0.396158</td>\n",
       "      <td>0.431867</td>\n",
       "      <td>0.046946</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.018905</td>\n",
       "      <td>0.136519</td>\n",
       "      <td>0.740108</td>\n",
       "      <td>0.297115</td>\n",
       "      <td>0.438328</td>\n",
       "      <td>0.043401</td>\n",
       "      <td>0.244122</td>\n",
       "      <td>0.585976</td>\n",
       "      <td>0.396158</td>\n",
       "      <td>0.431867</td>\n",
       "      <td>0.046946</td>\n",
       "      <td>0.186121</td>\n",
       "      <td>0.513127</td>\n",
       "      <td>0.540516</td>\n",
       "      <td>0.501238</td>\n",
       "      <td>0.085634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.043401</td>\n",
       "      <td>0.244122</td>\n",
       "      <td>0.585976</td>\n",
       "      <td>0.396158</td>\n",
       "      <td>0.431867</td>\n",
       "      <td>0.046946</td>\n",
       "      <td>0.186121</td>\n",
       "      <td>0.513127</td>\n",
       "      <td>0.540516</td>\n",
       "      <td>0.501238</td>\n",
       "      <td>0.085634</td>\n",
       "      <td>0.117236</td>\n",
       "      <td>0.544642</td>\n",
       "      <td>0.553859</td>\n",
       "      <td>0.596469</td>\n",
       "      <td>0.065537</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   var1(t-3)  var2(t-3)  var3(t-3)  var4(t-3)  var5(t-3)  var1(t-2)  \\\n",
       "3   0.039567   0.078779   0.831199   0.195240   0.380713   0.007090   \n",
       "4   0.007090   0.128849   0.579299   0.503813   0.514548   0.000000   \n",
       "5   0.000000   0.110492   0.605697   0.476259   0.512798   0.018905   \n",
       "6   0.018905   0.136519   0.740108   0.297115   0.438328   0.043401   \n",
       "7   0.043401   0.244122   0.585976   0.396158   0.431867   0.046946   \n",
       "\n",
       "   var2(t-2)  var3(t-2)  var4(t-2)  var5(t-2)  var1(t-1)  var2(t-1)  \\\n",
       "3   0.128849   0.579299   0.503813   0.514548   0.000000   0.110492   \n",
       "4   0.110492   0.605697   0.476259   0.512798   0.018905   0.136519   \n",
       "5   0.136519   0.740108   0.297115   0.438328   0.043401   0.244122   \n",
       "6   0.244122   0.585976   0.396158   0.431867   0.046946   0.186121   \n",
       "7   0.186121   0.513127   0.540516   0.501238   0.085634   0.117236   \n",
       "\n",
       "   var3(t-1)  var4(t-1)  var5(t-1)   var1(t)  \n",
       "3   0.605697   0.476259   0.512798  0.018905  \n",
       "4   0.740108   0.297115   0.438328  0.043401  \n",
       "5   0.585976   0.396158   0.431867  0.046946  \n",
       "6   0.513127   0.540516   0.501238  0.085634  \n",
       "7   0.544642   0.553859   0.596469  0.065537  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reframed = reframed.drop(reframed.columns[-4:], axis=1)  #dropping all the present values except present price\n",
    "reframed.head()  #Note: here last column is present price which is our target variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>var1(t)</th>\n",
       "      <th>var1(t-3)</th>\n",
       "      <th>var2(t-3)</th>\n",
       "      <th>var3(t-3)</th>\n",
       "      <th>var4(t-3)</th>\n",
       "      <th>var5(t-3)</th>\n",
       "      <th>var1(t-2)</th>\n",
       "      <th>var2(t-2)</th>\n",
       "      <th>var3(t-2)</th>\n",
       "      <th>var4(t-2)</th>\n",
       "      <th>var5(t-2)</th>\n",
       "      <th>var1(t-1)</th>\n",
       "      <th>var2(t-1)</th>\n",
       "      <th>var3(t-1)</th>\n",
       "      <th>var4(t-1)</th>\n",
       "      <th>var5(t-1)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.018905</td>\n",
       "      <td>0.039567</td>\n",
       "      <td>0.078779</td>\n",
       "      <td>0.831199</td>\n",
       "      <td>0.195240</td>\n",
       "      <td>0.380713</td>\n",
       "      <td>0.007090</td>\n",
       "      <td>0.128849</td>\n",
       "      <td>0.579299</td>\n",
       "      <td>0.503813</td>\n",
       "      <td>0.514548</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.110492</td>\n",
       "      <td>0.605697</td>\n",
       "      <td>0.476259</td>\n",
       "      <td>0.512798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.043401</td>\n",
       "      <td>0.007090</td>\n",
       "      <td>0.128849</td>\n",
       "      <td>0.579299</td>\n",
       "      <td>0.503813</td>\n",
       "      <td>0.514548</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.110492</td>\n",
       "      <td>0.605697</td>\n",
       "      <td>0.476259</td>\n",
       "      <td>0.512798</td>\n",
       "      <td>0.018905</td>\n",
       "      <td>0.136519</td>\n",
       "      <td>0.740108</td>\n",
       "      <td>0.297115</td>\n",
       "      <td>0.438328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.046946</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.110492</td>\n",
       "      <td>0.605697</td>\n",
       "      <td>0.476259</td>\n",
       "      <td>0.512798</td>\n",
       "      <td>0.018905</td>\n",
       "      <td>0.136519</td>\n",
       "      <td>0.740108</td>\n",
       "      <td>0.297115</td>\n",
       "      <td>0.438328</td>\n",
       "      <td>0.043401</td>\n",
       "      <td>0.244122</td>\n",
       "      <td>0.585976</td>\n",
       "      <td>0.396158</td>\n",
       "      <td>0.431867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.085634</td>\n",
       "      <td>0.018905</td>\n",
       "      <td>0.136519</td>\n",
       "      <td>0.740108</td>\n",
       "      <td>0.297115</td>\n",
       "      <td>0.438328</td>\n",
       "      <td>0.043401</td>\n",
       "      <td>0.244122</td>\n",
       "      <td>0.585976</td>\n",
       "      <td>0.396158</td>\n",
       "      <td>0.431867</td>\n",
       "      <td>0.046946</td>\n",
       "      <td>0.186121</td>\n",
       "      <td>0.513127</td>\n",
       "      <td>0.540516</td>\n",
       "      <td>0.501238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.065537</td>\n",
       "      <td>0.043401</td>\n",
       "      <td>0.244122</td>\n",
       "      <td>0.585976</td>\n",
       "      <td>0.396158</td>\n",
       "      <td>0.431867</td>\n",
       "      <td>0.046946</td>\n",
       "      <td>0.186121</td>\n",
       "      <td>0.513127</td>\n",
       "      <td>0.540516</td>\n",
       "      <td>0.501238</td>\n",
       "      <td>0.085634</td>\n",
       "      <td>0.117236</td>\n",
       "      <td>0.544642</td>\n",
       "      <td>0.553859</td>\n",
       "      <td>0.596469</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    var1(t)  var1(t-3)  var2(t-3)  var3(t-3)  var4(t-3)  var5(t-3)  var1(t-2)  \\\n",
       "3  0.018905   0.039567   0.078779   0.831199   0.195240   0.380713   0.007090   \n",
       "4  0.043401   0.007090   0.128849   0.579299   0.503813   0.514548   0.000000   \n",
       "5  0.046946   0.000000   0.110492   0.605697   0.476259   0.512798   0.018905   \n",
       "6  0.085634   0.018905   0.136519   0.740108   0.297115   0.438328   0.043401   \n",
       "7  0.065537   0.043401   0.244122   0.585976   0.396158   0.431867   0.046946   \n",
       "\n",
       "   var2(t-2)  var3(t-2)  var4(t-2)  var5(t-2)  var1(t-1)  var2(t-1)  \\\n",
       "3   0.128849   0.579299   0.503813   0.514548   0.000000   0.110492   \n",
       "4   0.110492   0.605697   0.476259   0.512798   0.018905   0.136519   \n",
       "5   0.136519   0.740108   0.297115   0.438328   0.043401   0.244122   \n",
       "6   0.244122   0.585976   0.396158   0.431867   0.046946   0.186121   \n",
       "7   0.186121   0.513127   0.540516   0.501238   0.085634   0.117236   \n",
       "\n",
       "   var3(t-1)  var4(t-1)  var5(t-1)  \n",
       "3   0.605697   0.476259   0.512798  \n",
       "4   0.740108   0.297115   0.438328  \n",
       "5   0.585976   0.396158   0.431867  \n",
       "6   0.513127   0.540516   0.501238  \n",
       "7   0.544642   0.553859   0.596469  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols = reframed.columns.tolist()\n",
    "cols = cols[-1:] + cols[:-1]   #just making the price as the first column\n",
    "reframed = reframed[cols]\n",
    "reframed.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------------------+-------------------+\n",
      "|            features|              label|\n",
      "+--------------------+-------------------+\n",
      "|[0.03956697509246...|0.01890490409309331|\n",
      "|[0.00709010702743...|0.04340079379707795|\n",
      "|[0.0,0.1104924045...|0.04694584731079798|\n",
      "|[0.01890490409309...|0.08563423895012345|\n",
      "|[0.04340079379707...| 0.0655374104520714|\n",
      "+--------------------+-------------------+\n",
      "only showing top 5 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from pyspark.ml.linalg import Vectors\n",
    "reg_DF =sql.createDataFrame(reframed)\n",
    "ad_df = reg_DF.rdd.map(lambda x: [Vectors.dense(x[1:]), x[0]]).toDF(['features', 'label'])\n",
    "ad_df.show(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from pyspark.ml.regression import LinearRegression\n",
    "lr = LinearRegression(featuresCol = 'features', labelCol = 'label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "lr_model = lr.fit(ad_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------------------+-------------------+\n",
      "|          prediction|              label|\n",
      "+--------------------+-------------------+\n",
      "|0.027451487572489858|0.01890490409309331|\n",
      "|0.030811449850603023|0.04340079379707795|\n",
      "| 0.05275280252795525|0.04694584731079798|\n",
      "| 0.06934986071359109|0.08563423895012345|\n",
      "| 0.09380546836196066| 0.0655374104520714|\n",
      "+--------------------+-------------------+\n",
      "only showing top 5 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "predictions = lr_model.transform(ad_df)\n",
    "someDF = predictions.select(\"prediction\", \"label\").show(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "predictions = predictions.toPandas().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1264e0e10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(predictions['label'])\n",
    "plt.plot(predictions['prediction'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficients: [0.0708530377608,-0.0126451811991,-0.0330874284728,-0.0176069385021,-0.0270199822808,0.118903567785,-0.00458866120068,-0.0897364477308,-0.126405117384,0.0779264547102,0.77808962732,-0.152564351238,-0.0517402220063,0.0415864062402,-0.140393355615]\n",
      "Intercept: 0.24257460855969606\n"
     ]
    }
   ],
   "source": [
    "print(\"Coefficients: %s\" % str(lr_model.coefficients))\n",
    "print(\"Intercept: %s\" % str(lr_model.intercept))"
   ]
  }
 ],
 "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}