taxiprediction-demand.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initial Python environment setup...\n",
    "import numpy as np # linear algebra\n",
    "import pandas as pd # CSV file I/O (e.g. pd.read_csv)\n",
    "import os # reading the input files we have access to"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from haversine import haversine\n",
    "# packages for mapping\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import plotly.figure_factory as ff\n",
    "from plotly.offline import init_notebook_mode, iplot\n",
    "import plotly.graph_objs as go\n",
    "import plotly.plotly as py\n",
    "from plotly import tools\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "key                   object\n",
       "fare_amount          float64\n",
       "pickup_datetime       object\n",
       "pickup_longitude     float64\n",
       "pickup_latitude      float64\n",
       "dropoff_longitude    float64\n",
       "dropoff_latitude     float64\n",
       "passenger_count        int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df =  pd.read_csv('train.csv', nrows = 10_000_00)\n",
    "train_df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>key</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>dropoff_longitude</th>\n",
       "      <th>dropoff_latitude</th>\n",
       "      <th>passenger_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2009-06-15 17:26:21.0000001</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-06-15 17:26:21 UTC</td>\n",
       "      <td>-73.844311</td>\n",
       "      <td>40.721319</td>\n",
       "      <td>-73.841610</td>\n",
       "      <td>40.712278</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2010-01-05 16:52:16.0000002</td>\n",
       "      <td>16.9</td>\n",
       "      <td>2010-01-05 16:52:16 UTC</td>\n",
       "      <td>-74.016048</td>\n",
       "      <td>40.711303</td>\n",
       "      <td>-73.979268</td>\n",
       "      <td>40.782004</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2011-08-18 00:35:00.00000049</td>\n",
       "      <td>5.7</td>\n",
       "      <td>2011-08-18 00:35:00 UTC</td>\n",
       "      <td>-73.982738</td>\n",
       "      <td>40.761270</td>\n",
       "      <td>-73.991242</td>\n",
       "      <td>40.750562</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2012-04-21 04:30:42.0000001</td>\n",
       "      <td>7.7</td>\n",
       "      <td>2012-04-21 04:30:42 UTC</td>\n",
       "      <td>-73.987130</td>\n",
       "      <td>40.733143</td>\n",
       "      <td>-73.991567</td>\n",
       "      <td>40.758092</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2010-03-09 07:51:00.000000135</td>\n",
       "      <td>5.3</td>\n",
       "      <td>2010-03-09 07:51:00 UTC</td>\n",
       "      <td>-73.968095</td>\n",
       "      <td>40.768008</td>\n",
       "      <td>-73.956655</td>\n",
       "      <td>40.783762</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             key  fare_amount          pickup_datetime  \\\n",
       "0    2009-06-15 17:26:21.0000001          4.5  2009-06-15 17:26:21 UTC   \n",
       "1    2010-01-05 16:52:16.0000002         16.9  2010-01-05 16:52:16 UTC   \n",
       "2   2011-08-18 00:35:00.00000049          5.7  2011-08-18 00:35:00 UTC   \n",
       "3    2012-04-21 04:30:42.0000001          7.7  2012-04-21 04:30:42 UTC   \n",
       "4  2010-03-09 07:51:00.000000135          5.3  2010-03-09 07:51:00 UTC   \n",
       "\n",
       "   pickup_longitude  pickup_latitude  dropoff_longitude  dropoff_latitude  \\\n",
       "0        -73.844311        40.721319         -73.841610         40.712278   \n",
       "1        -74.016048        40.711303         -73.979268         40.782004   \n",
       "2        -73.982738        40.761270         -73.991242         40.750562   \n",
       "3        -73.987130        40.733143         -73.991567         40.758092   \n",
       "4        -73.968095        40.768008         -73.956655         40.783762   \n",
       "\n",
       "   passenger_count  \n",
       "0                1  \n",
       "1                1  \n",
       "2                2  \n",
       "3                1  \n",
       "4                1  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# tranform pickup date time to datetimeobject\n",
    "train_df['pickup_datetime']= pd.to_datetime(train_df['pickup_datetime'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    " # Compute the day of week\n",
    "\n",
    "train_df['pickup_day_of_week']=train_df['pickup_datetime'].dt.dayofweek"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    " #Compute the month\n",
    "train_df['pickup_month']=train_df['pickup_datetime'].dt.month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>key</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>dropoff_longitude</th>\n",
       "      <th>dropoff_latitude</th>\n",
       "      <th>passenger_count</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "      <th>pickup_month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>999995</th>\n",
       "      <td>2014-09-13 21:44:38.0000001</td>\n",
       "      <td>7.0</td>\n",
       "      <td>2014-09-13 21:44:38</td>\n",
       "      <td>-73.976676</td>\n",
       "      <td>40.785630</td>\n",
       "      <td>-73.959196</td>\n",
       "      <td>40.775114</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999996</th>\n",
       "      <td>2010-09-20 14:50:37.0000002</td>\n",
       "      <td>7.3</td>\n",
       "      <td>2010-09-20 14:50:37</td>\n",
       "      <td>-73.992103</td>\n",
       "      <td>40.671385</td>\n",
       "      <td>-73.992103</td>\n",
       "      <td>40.671385</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999997</th>\n",
       "      <td>2013-04-26 14:03:00.000000118</td>\n",
       "      <td>10.5</td>\n",
       "      <td>2013-04-26 14:03:00</td>\n",
       "      <td>-73.978423</td>\n",
       "      <td>40.751135</td>\n",
       "      <td>-73.993397</td>\n",
       "      <td>40.762197</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999998</th>\n",
       "      <td>2011-07-08 00:29:00.00000099</td>\n",
       "      <td>6.9</td>\n",
       "      <td>2011-07-08 00:29:00</td>\n",
       "      <td>-73.980317</td>\n",
       "      <td>40.759482</td>\n",
       "      <td>-73.976832</td>\n",
       "      <td>40.743122</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999999</th>\n",
       "      <td>2009-12-31 14:30:00.00000021</td>\n",
       "      <td>4.1</td>\n",
       "      <td>2009-12-31 14:30:00</td>\n",
       "      <td>-74.006635</td>\n",
       "      <td>40.741598</td>\n",
       "      <td>-73.999450</td>\n",
       "      <td>40.745702</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  key  fare_amount     pickup_datetime  \\\n",
       "999995    2014-09-13 21:44:38.0000001          7.0 2014-09-13 21:44:38   \n",
       "999996    2010-09-20 14:50:37.0000002          7.3 2010-09-20 14:50:37   \n",
       "999997  2013-04-26 14:03:00.000000118         10.5 2013-04-26 14:03:00   \n",
       "999998   2011-07-08 00:29:00.00000099          6.9 2011-07-08 00:29:00   \n",
       "999999   2009-12-31 14:30:00.00000021          4.1 2009-12-31 14:30:00   \n",
       "\n",
       "        pickup_longitude  pickup_latitude  dropoff_longitude  \\\n",
       "999995        -73.976676        40.785630         -73.959196   \n",
       "999996        -73.992103        40.671385         -73.992103   \n",
       "999997        -73.978423        40.751135         -73.993397   \n",
       "999998        -73.980317        40.759482         -73.976832   \n",
       "999999        -74.006635        40.741598         -73.999450   \n",
       "\n",
       "        dropoff_latitude  passenger_count  pickup_day_of_week  pickup_month  \n",
       "999995         40.775114                1                   5             9  \n",
       "999996         40.671385                2                   0             9  \n",
       "999997         40.762197                5                   4             4  \n",
       "999998         40.743122                1                   4             7  \n",
       "999999         40.745702                1                   3            12  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df['pickup_date']=train_df['pickup_datetime'].dt.date"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x50e5eb8>"
      ]
     },
     "execution_count": 11,
     "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": [
    "train_df[\"pickup_month\"].value_counts().sort_index().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df.sort_values('pickup_day_of_week', ascending = True, inplace= True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xd49b080>"
      ]
     },
     "execution_count": 13,
     "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": [
    "train_df[\"pickup_day_of_week\"].value_counts().sort_index().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df['pickup_hour']=train_df['pickup_datetime'].dt.hour"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xd5015f8>"
      ]
     },
     "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": [
    "train_df['pickup_hour'].value_counts().sort_index().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df['pickup_year']=train_df['pickup_datetime'].dt.year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xd5607f0>"
      ]
     },
     "execution_count": 17,
     "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": [
    "train_df['pickup_year'].value_counts().sort_index().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xd5cc7f0>"
      ]
     },
     "execution_count": 18,
     "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": [
    "train_df.groupby('pickup_hour')['fare_amount'].mean().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xd610908>"
      ]
     },
     "execution_count": 19,
     "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": [
    "train_df.groupby('pickup_day_of_week')['fare_amount'].mean().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xd67f208>"
      ]
     },
     "execution_count": 20,
     "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": [
    "train_df.groupby('pickup_month')['fare_amount'].mean().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1ad788d0>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_df[\"fare_amount\"].plot.box(figsize=(16,5))  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>key</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>dropoff_longitude</th>\n",
       "      <th>dropoff_latitude</th>\n",
       "      <th>passenger_count</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "      <th>pickup_month</th>\n",
       "      <th>pickup_date</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2009-06-15 17:26:21.0000001</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-06-15 17:26:21</td>\n",
       "      <td>-73.844311</td>\n",
       "      <td>40.721319</td>\n",
       "      <td>-73.841610</td>\n",
       "      <td>40.712278</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>2009-06-15</td>\n",
       "      <td>17</td>\n",
       "      <td>2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>410475</th>\n",
       "      <td>2009-08-03 23:01:00.00000077</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-08-03 23:01:00</td>\n",
       "      <td>-73.992147</td>\n",
       "      <td>40.762532</td>\n",
       "      <td>-73.984915</td>\n",
       "      <td>40.768142</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>2009-08-03</td>\n",
       "      <td>23</td>\n",
       "      <td>2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>410487</th>\n",
       "      <td>2011-07-11 13:08:09.0000001</td>\n",
       "      <td>10.9</td>\n",
       "      <td>2011-07-11 13:08:09</td>\n",
       "      <td>-73.993900</td>\n",
       "      <td>40.750800</td>\n",
       "      <td>-74.014800</td>\n",
       "      <td>40.718000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>2011-07-11</td>\n",
       "      <td>13</td>\n",
       "      <td>2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>410491</th>\n",
       "      <td>2012-12-17 22:58:00.000000163</td>\n",
       "      <td>29.0</td>\n",
       "      <td>2012-12-17 22:58:00</td>\n",
       "      <td>-73.979827</td>\n",
       "      <td>40.766015</td>\n",
       "      <td>-73.962497</td>\n",
       "      <td>40.677382</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>2012-12-17</td>\n",
       "      <td>22</td>\n",
       "      <td>2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>410493</th>\n",
       "      <td>2011-02-28 22:03:00.000000115</td>\n",
       "      <td>6.1</td>\n",
       "      <td>2011-02-28 22:03:00</td>\n",
       "      <td>-73.995742</td>\n",
       "      <td>40.726778</td>\n",
       "      <td>-73.978080</td>\n",
       "      <td>40.721325</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2011-02-28</td>\n",
       "      <td>22</td>\n",
       "      <td>2011</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  key  fare_amount     pickup_datetime  \\\n",
       "0         2009-06-15 17:26:21.0000001          4.5 2009-06-15 17:26:21   \n",
       "410475   2009-08-03 23:01:00.00000077          4.5 2009-08-03 23:01:00   \n",
       "410487    2011-07-11 13:08:09.0000001         10.9 2011-07-11 13:08:09   \n",
       "410491  2012-12-17 22:58:00.000000163         29.0 2012-12-17 22:58:00   \n",
       "410493  2011-02-28 22:03:00.000000115          6.1 2011-02-28 22:03:00   \n",
       "\n",
       "        pickup_longitude  pickup_latitude  dropoff_longitude  \\\n",
       "0             -73.844311        40.721319         -73.841610   \n",
       "410475        -73.992147        40.762532         -73.984915   \n",
       "410487        -73.993900        40.750800         -74.014800   \n",
       "410491        -73.979827        40.766015         -73.962497   \n",
       "410493        -73.995742        40.726778         -73.978080   \n",
       "\n",
       "        dropoff_latitude  passenger_count  pickup_day_of_week  pickup_month  \\\n",
       "0              40.712278                1                   0             6   \n",
       "410475         40.768142                5                   0             8   \n",
       "410487         40.718000                1                   0             7   \n",
       "410491         40.677382                3                   0            12   \n",
       "410493         40.721325                1                   0             2   \n",
       "\n",
       "       pickup_date  pickup_hour  pickup_year  \n",
       "0       2009-06-15           17         2009  \n",
       "410475  2009-08-03           23         2009  \n",
       "410487  2011-07-11           13         2011  \n",
       "410491  2012-12-17           22         2012  \n",
       "410493  2011-02-28           22         2011  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df['id'] = train_df.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 314,
   "metadata": {},
   "outputs": [],
   "source": [
    "new =train_df.groupby(pd.Grouper(key='pickup_datetime', freq='30T')).apply(lambda x: x['id'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 315,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df['id'] = train_df.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pickup_datetime            \n",
       "2009-01-01 00:00:00  496831    880228\n",
       "                     476205    661006\n",
       "                     455668    756707\n",
       "                     463490    745620\n",
       "                     497968    964259\n",
       "                     429612    915988\n",
       "2009-01-01 00:30:00  534793    297860\n",
       "                     551639    124750\n",
       "                     475925    876677\n",
       "                     543219     82469\n",
       "                     553673    124445\n",
       "                     481315    973236\n",
       "                     492985    965480\n",
       "2009-01-01 01:00:00  452273    760438\n",
       "                     473058    894290\n",
       "                     462003    766885\n",
       "                     467958    945037\n",
       "                     501683     76195\n",
       "                     506057    102243\n",
       "                     491811    773755\n",
       "                     413667    542263\n",
       "2009-01-01 01:30:00  470098     12739\n",
       "                     425369    709206\n",
       "                     525187    307980\n",
       "                     416679    674362\n",
       "2009-01-01 02:00:00  445246     17546\n",
       "                     482801     25687\n",
       "                     475710    719299\n",
       "                     414945    542248\n",
       "                     526737    304112\n",
       "                                ...  \n",
       "2015-06-30 22:30:00  265947    990467\n",
       "                     198065    525126\n",
       "                     136498    701682\n",
       "                     223827    392244\n",
       "                     198318    475897\n",
       "                     185718    569474\n",
       "                     259067      1428\n",
       "                     159471    828099\n",
       "                     254095    194700\n",
       "                     268431    114820\n",
       "                     222546    391489\n",
       "                     160395     98082\n",
       "2015-06-30 23:00:00  256779    939905\n",
       "                     137854    337140\n",
       "                     174863    450764\n",
       "                     156421    841675\n",
       "                     139440    350061\n",
       "                     249852    179298\n",
       "                     207301    647613\n",
       "                     208742    362807\n",
       "                     243818    182743\n",
       "2015-06-30 23:30:00  194077    506007\n",
       "                     133754    751978\n",
       "                     160836    240182\n",
       "                     142686    323864\n",
       "                     164262    286276\n",
       "                     234133    955575\n",
       "                     246934    915826\n",
       "                     133822    751350\n",
       "                     172083    785182\n",
       "Name: id, Length: 1000000, dtype: int64"
      ]
     },
     "execution_count": 316,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 317,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TimeGrouper(key='pickup_datetime', freq=<30 * MonthEnds>, axis=0, sort=True, closed='right', label='right', how='mean', convention='e', base=0)"
      ]
     },
     "execution_count": 317,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Grouper(key='pickup_datetime', freq='30m')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DatetimeIndex(['2009-01-01 00:00:00', '2009-01-01 00:00:00',\n",
       "               '2009-01-01 00:00:00', '2009-01-01 00:00:00',\n",
       "               '2009-01-01 00:00:00', '2009-01-01 00:00:00',\n",
       "               '2009-01-01 00:30:00', '2009-01-01 00:30:00',\n",
       "               '2009-01-01 00:30:00', '2009-01-01 00:30:00',\n",
       "               ...\n",
       "               '2015-06-30 23:00:00', '2015-06-30 23:30:00',\n",
       "               '2015-06-30 23:30:00', '2015-06-30 23:30:00',\n",
       "               '2015-06-30 23:30:00', '2015-06-30 23:30:00',\n",
       "               '2015-06-30 23:30:00', '2015-06-30 23:30:00',\n",
       "               '2015-06-30 23:30:00', '2015-06-30 23:30:00'],\n",
       "              dtype='datetime64[ns]', name='pickup_datetime', length=1000000, freq='30T')"
      ]
     },
     "execution_count": 318,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new.index.get_level_values('pickup_datetime')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "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>key</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>dropoff_longitude</th>\n",
       "      <th>dropoff_latitude</th>\n",
       "      <th>passenger_count</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "      <th>pickup_month</th>\n",
       "      <th>pickup_date</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_year</th>\n",
       "      <th>id</th>\n",
       "      <th>interval</th>\n",
       "      <th>interval_time</th>\n",
       "      <th>demand</th>\n",
       "      <th>distance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [key, fare_amount, pickup_datetime, pickup_longitude, pickup_latitude, dropoff_longitude, dropoff_latitude, passenger_count, pickup_day_of_week, pickup_month, pickup_date, pickup_hour, pickup_year, id, interval, interval_time, demand, distance]\n",
       "Index: []"
      ]
     },
     "execution_count": 319,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.loc[train_df[\"key\"] == '799435' ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "key                           object\n",
       "fare_amount                  float64\n",
       "pickup_datetime       datetime64[ns]\n",
       "pickup_longitude             float64\n",
       "pickup_latitude              float64\n",
       "dropoff_longitude            float64\n",
       "dropoff_latitude             float64\n",
       "passenger_count                int64\n",
       "pickup_day_of_week             int64\n",
       "pickup_month                   int64\n",
       "pickup_date                   object\n",
       "pickup_hour                    int64\n",
       "pickup_year                    int64\n",
       "id                             int64\n",
       "interval              datetime64[ns]\n",
       "interval_time                 object\n",
       "demand                         int64\n",
       "distance                     float64\n",
       "dtype: object"
      ]
     },
     "execution_count": 320,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 321,
   "metadata": {},
   "outputs": [],
   "source": [
    "interval_df=pd.DataFrame({'interval':new.index.get_level_values(0), 'id':new.index.get_level_values(1)})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 322,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Int64Index([496831, 476205, 455668, 463490, 497968, 429612, 534793, 551639,\n",
       "            475925, 543219,\n",
       "            ...\n",
       "            243818, 194077, 133754, 160836, 142686, 164262, 234133, 246934,\n",
       "            133822, 172083],\n",
       "           dtype='int64', length=1000000)"
      ]
     },
     "execution_count": 322,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new.index.get_level_values(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 323,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "interval    datetime64[ns]\n",
       "id                   int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 323,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "interval_df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 324,
   "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>interval</th>\n",
       "      <th>id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>999995</th>\n",
       "      <td>2015-06-30 23:30:00</td>\n",
       "      <td>164262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999996</th>\n",
       "      <td>2015-06-30 23:30:00</td>\n",
       "      <td>234133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999997</th>\n",
       "      <td>2015-06-30 23:30:00</td>\n",
       "      <td>246934</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999998</th>\n",
       "      <td>2015-06-30 23:30:00</td>\n",
       "      <td>133822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999999</th>\n",
       "      <td>2015-06-30 23:30:00</td>\n",
       "      <td>172083</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  interval      id\n",
       "999995 2015-06-30 23:30:00  164262\n",
       "999996 2015-06-30 23:30:00  234133\n",
       "999997 2015-06-30 23:30:00  246934\n",
       "999998 2015-06-30 23:30:00  133822\n",
       "999999 2015-06-30 23:30:00  172083"
      ]
     },
     "execution_count": 324,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "interval_df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 325,
   "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>key</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>dropoff_longitude</th>\n",
       "      <th>dropoff_latitude</th>\n",
       "      <th>passenger_count</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "      <th>pickup_month</th>\n",
       "      <th>pickup_date</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_year</th>\n",
       "      <th>id</th>\n",
       "      <th>interval</th>\n",
       "      <th>interval_time</th>\n",
       "      <th>demand</th>\n",
       "      <th>distance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>751350</th>\n",
       "      <td>2015-02-07 10:22:55.0000002</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2015-02-07 10:22:55</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.76</td>\n",
       "      <td>-73.988281</td>\n",
       "      <td>40.767265</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>2015-02-07</td>\n",
       "      <td>10</td>\n",
       "      <td>2015</td>\n",
       "      <td>751350</td>\n",
       "      <td>2015-02-07 10:00:00</td>\n",
       "      <td>10:00:00</td>\n",
       "      <td>2</td>\n",
       "      <td>1.79477</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                key  fare_amount     pickup_datetime  \\\n",
       "751350  2015-02-07 10:22:55.0000002          4.5 2015-02-07 10:22:55   \n",
       "\n",
       "        pickup_longitude  pickup_latitude  dropoff_longitude  \\\n",
       "751350             -74.0            40.76         -73.988281   \n",
       "\n",
       "        dropoff_latitude  passenger_count  pickup_day_of_week  pickup_month  \\\n",
       "751350         40.767265                1                   5             2   \n",
       "\n",
       "       pickup_date  pickup_hour  pickup_year      id            interval  \\\n",
       "751350  2015-02-07           10         2015  751350 2015-02-07 10:00:00   \n",
       "\n",
       "       interval_time  demand  distance  \n",
       "751350      10:00:00       2   1.79477  "
      ]
     },
     "execution_count": 325,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.loc[train_df[\"id\"] == 751350]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 337,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df=pd.merge(train_df, interval_df, left_on = 'id', right_on = 'id', how = 'inner')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 339,
   "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>key</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>dropoff_longitude</th>\n",
       "      <th>dropoff_latitude</th>\n",
       "      <th>passenger_count</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "      <th>pickup_month</th>\n",
       "      <th>pickup_date</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_year</th>\n",
       "      <th>id</th>\n",
       "      <th>interval_x</th>\n",
       "      <th>interval_time</th>\n",
       "      <th>demand</th>\n",
       "      <th>distance</th>\n",
       "      <th>interval_y</th>\n",
       "      <th>interval</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2009-06-15 17:26:21.0000001</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-06-15 17:26:21</td>\n",
       "      <td>-73.8</td>\n",
       "      <td>40.72</td>\n",
       "      <td>-73.841610</td>\n",
       "      <td>40.712278</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>2009-06-15</td>\n",
       "      <td>17</td>\n",
       "      <td>2009</td>\n",
       "      <td>0</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "      <td>17:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>4.365758</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2009-08-03 23:01:00.00000077</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-08-03 23:01:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.76</td>\n",
       "      <td>-73.984915</td>\n",
       "      <td>40.768142</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>2009-08-03</td>\n",
       "      <td>23</td>\n",
       "      <td>2009</td>\n",
       "      <td>1</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "      <td>23:00:00</td>\n",
       "      <td>3</td>\n",
       "      <td>2.175726</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2011-07-11 13:08:09.0000001</td>\n",
       "      <td>10.9</td>\n",
       "      <td>2011-07-11 13:08:09</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.75</td>\n",
       "      <td>-74.014800</td>\n",
       "      <td>40.718000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>2011-07-11</td>\n",
       "      <td>13</td>\n",
       "      <td>2011</td>\n",
       "      <td>2</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "      <td>13:00:00</td>\n",
       "      <td>3</td>\n",
       "      <td>4.805557</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2012-12-17 22:58:00.000000163</td>\n",
       "      <td>29.0</td>\n",
       "      <td>2012-12-17 22:58:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.77</td>\n",
       "      <td>-73.962497</td>\n",
       "      <td>40.677382</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>2012-12-17</td>\n",
       "      <td>22</td>\n",
       "      <td>2012</td>\n",
       "      <td>3</td>\n",
       "      <td>2012-12-17 22:00:00</td>\n",
       "      <td>22:00:00</td>\n",
       "      <td>3</td>\n",
       "      <td>13.461272</td>\n",
       "      <td>2012-12-17 22:30:00</td>\n",
       "      <td>2012-12-17 22:30:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2011-02-28 22:03:00.000000115</td>\n",
       "      <td>6.1</td>\n",
       "      <td>2011-02-28 22:03:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.73</td>\n",
       "      <td>-73.978080</td>\n",
       "      <td>40.721325</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2011-02-28</td>\n",
       "      <td>22</td>\n",
       "      <td>2011</td>\n",
       "      <td>4</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "      <td>22:00:00</td>\n",
       "      <td>6</td>\n",
       "      <td>2.811899</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             key  fare_amount     pickup_datetime  \\\n",
       "0    2009-06-15 17:26:21.0000001          4.5 2009-06-15 17:26:21   \n",
       "1   2009-08-03 23:01:00.00000077          4.5 2009-08-03 23:01:00   \n",
       "2    2011-07-11 13:08:09.0000001         10.9 2011-07-11 13:08:09   \n",
       "3  2012-12-17 22:58:00.000000163         29.0 2012-12-17 22:58:00   \n",
       "4  2011-02-28 22:03:00.000000115          6.1 2011-02-28 22:03:00   \n",
       "\n",
       "   pickup_longitude  pickup_latitude  dropoff_longitude  dropoff_latitude  \\\n",
       "0             -73.8            40.72         -73.841610         40.712278   \n",
       "1             -74.0            40.76         -73.984915         40.768142   \n",
       "2             -74.0            40.75         -74.014800         40.718000   \n",
       "3             -74.0            40.77         -73.962497         40.677382   \n",
       "4             -74.0            40.73         -73.978080         40.721325   \n",
       "\n",
       "   passenger_count  pickup_day_of_week  pickup_month pickup_date  pickup_hour  \\\n",
       "0                1                   0             6  2009-06-15           17   \n",
       "1                5                   0             8  2009-08-03           23   \n",
       "2                1                   0             7  2011-07-11           13   \n",
       "3                3                   0            12  2012-12-17           22   \n",
       "4                1                   0             2  2011-02-28           22   \n",
       "\n",
       "   pickup_year  id          interval_x interval_time  demand   distance  \\\n",
       "0         2009   0 2009-06-15 17:00:00      17:00:00       1   4.365758   \n",
       "1         2009   1 2009-08-03 23:00:00      23:00:00       3   2.175726   \n",
       "2         2011   2 2011-07-11 13:00:00      13:00:00       3   4.805557   \n",
       "3         2012   3 2012-12-17 22:00:00      22:00:00       3  13.461272   \n",
       "4         2011   4 2011-02-28 22:00:00      22:00:00       6   2.811899   \n",
       "\n",
       "           interval_y            interval  \n",
       "0 2009-06-15 17:00:00 2009-06-15 17:00:00  \n",
       "1 2009-08-03 23:00:00 2009-08-03 23:00:00  \n",
       "2 2011-07-11 13:00:00 2011-07-11 13:00:00  \n",
       "3 2012-12-17 22:30:00 2012-12-17 22:30:00  \n",
       "4 2011-02-28 22:00:00 2011-02-28 22:00:00  "
      ]
     },
     "execution_count": 339,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 340,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df['interval_time']=train_df['interval'].dt.time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 342,
   "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>key</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>dropoff_longitude</th>\n",
       "      <th>dropoff_latitude</th>\n",
       "      <th>passenger_count</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "      <th>pickup_month</th>\n",
       "      <th>pickup_date</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_year</th>\n",
       "      <th>id</th>\n",
       "      <th>interval_x</th>\n",
       "      <th>interval_time</th>\n",
       "      <th>demand</th>\n",
       "      <th>distance</th>\n",
       "      <th>interval_y</th>\n",
       "      <th>interval</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2009-06-15 17:26:21.0000001</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-06-15 17:26:21</td>\n",
       "      <td>-73.8</td>\n",
       "      <td>40.72</td>\n",
       "      <td>-73.841610</td>\n",
       "      <td>40.712278</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>2009-06-15</td>\n",
       "      <td>17</td>\n",
       "      <td>2009</td>\n",
       "      <td>0</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "      <td>17:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>4.365758</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2009-08-03 23:01:00.00000077</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-08-03 23:01:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.76</td>\n",
       "      <td>-73.984915</td>\n",
       "      <td>40.768142</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>2009-08-03</td>\n",
       "      <td>23</td>\n",
       "      <td>2009</td>\n",
       "      <td>1</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "      <td>23:00:00</td>\n",
       "      <td>3</td>\n",
       "      <td>2.175726</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2011-07-11 13:08:09.0000001</td>\n",
       "      <td>10.9</td>\n",
       "      <td>2011-07-11 13:08:09</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.75</td>\n",
       "      <td>-74.014800</td>\n",
       "      <td>40.718000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>2011-07-11</td>\n",
       "      <td>13</td>\n",
       "      <td>2011</td>\n",
       "      <td>2</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "      <td>13:00:00</td>\n",
       "      <td>3</td>\n",
       "      <td>4.805557</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2012-12-17 22:58:00.000000163</td>\n",
       "      <td>29.0</td>\n",
       "      <td>2012-12-17 22:58:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.77</td>\n",
       "      <td>-73.962497</td>\n",
       "      <td>40.677382</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>2012-12-17</td>\n",
       "      <td>22</td>\n",
       "      <td>2012</td>\n",
       "      <td>3</td>\n",
       "      <td>2012-12-17 22:00:00</td>\n",
       "      <td>22:30:00</td>\n",
       "      <td>3</td>\n",
       "      <td>13.461272</td>\n",
       "      <td>2012-12-17 22:30:00</td>\n",
       "      <td>2012-12-17 22:30:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2011-02-28 22:03:00.000000115</td>\n",
       "      <td>6.1</td>\n",
       "      <td>2011-02-28 22:03:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.73</td>\n",
       "      <td>-73.978080</td>\n",
       "      <td>40.721325</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2011-02-28</td>\n",
       "      <td>22</td>\n",
       "      <td>2011</td>\n",
       "      <td>4</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "      <td>22:00:00</td>\n",
       "      <td>6</td>\n",
       "      <td>2.811899</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             key  fare_amount     pickup_datetime  \\\n",
       "0    2009-06-15 17:26:21.0000001          4.5 2009-06-15 17:26:21   \n",
       "1   2009-08-03 23:01:00.00000077          4.5 2009-08-03 23:01:00   \n",
       "2    2011-07-11 13:08:09.0000001         10.9 2011-07-11 13:08:09   \n",
       "3  2012-12-17 22:58:00.000000163         29.0 2012-12-17 22:58:00   \n",
       "4  2011-02-28 22:03:00.000000115          6.1 2011-02-28 22:03:00   \n",
       "\n",
       "   pickup_longitude  pickup_latitude  dropoff_longitude  dropoff_latitude  \\\n",
       "0             -73.8            40.72         -73.841610         40.712278   \n",
       "1             -74.0            40.76         -73.984915         40.768142   \n",
       "2             -74.0            40.75         -74.014800         40.718000   \n",
       "3             -74.0            40.77         -73.962497         40.677382   \n",
       "4             -74.0            40.73         -73.978080         40.721325   \n",
       "\n",
       "   passenger_count  pickup_day_of_week  pickup_month pickup_date  pickup_hour  \\\n",
       "0                1                   0             6  2009-06-15           17   \n",
       "1                5                   0             8  2009-08-03           23   \n",
       "2                1                   0             7  2011-07-11           13   \n",
       "3                3                   0            12  2012-12-17           22   \n",
       "4                1                   0             2  2011-02-28           22   \n",
       "\n",
       "   pickup_year  id          interval_x interval_time  demand   distance  \\\n",
       "0         2009   0 2009-06-15 17:00:00      17:00:00       1   4.365758   \n",
       "1         2009   1 2009-08-03 23:00:00      23:00:00       3   2.175726   \n",
       "2         2011   2 2011-07-11 13:00:00      13:00:00       3   4.805557   \n",
       "3         2012   3 2012-12-17 22:00:00      22:30:00       3  13.461272   \n",
       "4         2011   4 2011-02-28 22:00:00      22:00:00       6   2.811899   \n",
       "\n",
       "           interval_y            interval  \n",
       "0 2009-06-15 17:00:00 2009-06-15 17:00:00  \n",
       "1 2009-08-03 23:00:00 2009-08-03 23:00:00  \n",
       "2 2011-07-11 13:00:00 2011-07-11 13:00:00  \n",
       "3 2012-12-17 22:30:00 2012-12-17 22:30:00  \n",
       "4 2011-02-28 22:00:00 2011-02-28 22:00:00  "
      ]
     },
     "execution_count": 342,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 343,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x5ea7a588>"
      ]
     },
     "execution_count": 343,
     "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": [
    "train_df['interval_time'].value_counts().sort_index().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 344,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df[\"pickup_longitude\"] = train_df['pickup_longitude'].apply(lambda x:round(x,1))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 345,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df[\"pickup_latitude\"] = train_df['pickup_latitude'].apply(lambda x:round(x,2))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 346,
   "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>key</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>dropoff_longitude</th>\n",
       "      <th>dropoff_latitude</th>\n",
       "      <th>passenger_count</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "      <th>pickup_month</th>\n",
       "      <th>pickup_date</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_year</th>\n",
       "      <th>id</th>\n",
       "      <th>interval_x</th>\n",
       "      <th>interval_time</th>\n",
       "      <th>demand</th>\n",
       "      <th>distance</th>\n",
       "      <th>interval_y</th>\n",
       "      <th>interval</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2009-06-15 17:26:21.0000001</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-06-15 17:26:21</td>\n",
       "      <td>-73.8</td>\n",
       "      <td>40.72</td>\n",
       "      <td>-73.841610</td>\n",
       "      <td>40.712278</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>2009-06-15</td>\n",
       "      <td>17</td>\n",
       "      <td>2009</td>\n",
       "      <td>0</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "      <td>17:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>4.365758</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2009-08-03 23:01:00.00000077</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-08-03 23:01:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.76</td>\n",
       "      <td>-73.984915</td>\n",
       "      <td>40.768142</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>2009-08-03</td>\n",
       "      <td>23</td>\n",
       "      <td>2009</td>\n",
       "      <td>1</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "      <td>23:00:00</td>\n",
       "      <td>3</td>\n",
       "      <td>2.175726</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2011-07-11 13:08:09.0000001</td>\n",
       "      <td>10.9</td>\n",
       "      <td>2011-07-11 13:08:09</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.75</td>\n",
       "      <td>-74.014800</td>\n",
       "      <td>40.718000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>2011-07-11</td>\n",
       "      <td>13</td>\n",
       "      <td>2011</td>\n",
       "      <td>2</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "      <td>13:00:00</td>\n",
       "      <td>3</td>\n",
       "      <td>4.805557</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2012-12-17 22:58:00.000000163</td>\n",
       "      <td>29.0</td>\n",
       "      <td>2012-12-17 22:58:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.77</td>\n",
       "      <td>-73.962497</td>\n",
       "      <td>40.677382</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>2012-12-17</td>\n",
       "      <td>22</td>\n",
       "      <td>2012</td>\n",
       "      <td>3</td>\n",
       "      <td>2012-12-17 22:00:00</td>\n",
       "      <td>22:30:00</td>\n",
       "      <td>3</td>\n",
       "      <td>13.461272</td>\n",
       "      <td>2012-12-17 22:30:00</td>\n",
       "      <td>2012-12-17 22:30:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2011-02-28 22:03:00.000000115</td>\n",
       "      <td>6.1</td>\n",
       "      <td>2011-02-28 22:03:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.73</td>\n",
       "      <td>-73.978080</td>\n",
       "      <td>40.721325</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2011-02-28</td>\n",
       "      <td>22</td>\n",
       "      <td>2011</td>\n",
       "      <td>4</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "      <td>22:00:00</td>\n",
       "      <td>6</td>\n",
       "      <td>2.811899</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             key  fare_amount     pickup_datetime  \\\n",
       "0    2009-06-15 17:26:21.0000001          4.5 2009-06-15 17:26:21   \n",
       "1   2009-08-03 23:01:00.00000077          4.5 2009-08-03 23:01:00   \n",
       "2    2011-07-11 13:08:09.0000001         10.9 2011-07-11 13:08:09   \n",
       "3  2012-12-17 22:58:00.000000163         29.0 2012-12-17 22:58:00   \n",
       "4  2011-02-28 22:03:00.000000115          6.1 2011-02-28 22:03:00   \n",
       "\n",
       "   pickup_longitude  pickup_latitude  dropoff_longitude  dropoff_latitude  \\\n",
       "0             -73.8            40.72         -73.841610         40.712278   \n",
       "1             -74.0            40.76         -73.984915         40.768142   \n",
       "2             -74.0            40.75         -74.014800         40.718000   \n",
       "3             -74.0            40.77         -73.962497         40.677382   \n",
       "4             -74.0            40.73         -73.978080         40.721325   \n",
       "\n",
       "   passenger_count  pickup_day_of_week  pickup_month pickup_date  pickup_hour  \\\n",
       "0                1                   0             6  2009-06-15           17   \n",
       "1                5                   0             8  2009-08-03           23   \n",
       "2                1                   0             7  2011-07-11           13   \n",
       "3                3                   0            12  2012-12-17           22   \n",
       "4                1                   0             2  2011-02-28           22   \n",
       "\n",
       "   pickup_year  id          interval_x interval_time  demand   distance  \\\n",
       "0         2009   0 2009-06-15 17:00:00      17:00:00       1   4.365758   \n",
       "1         2009   1 2009-08-03 23:00:00      23:00:00       3   2.175726   \n",
       "2         2011   2 2011-07-11 13:00:00      13:00:00       3   4.805557   \n",
       "3         2012   3 2012-12-17 22:00:00      22:30:00       3  13.461272   \n",
       "4         2011   4 2011-02-28 22:00:00      22:00:00       6   2.811899   \n",
       "\n",
       "           interval_y            interval  \n",
       "0 2009-06-15 17:00:00 2009-06-15 17:00:00  \n",
       "1 2009-08-03 23:00:00 2009-08-03 23:00:00  \n",
       "2 2011-07-11 13:00:00 2011-07-11 13:00:00  \n",
       "3 2012-12-17 22:30:00 2012-12-17 22:30:00  \n",
       "4 2011-02-28 22:00:00 2011-02-28 22:00:00  "
      ]
     },
     "execution_count": 346,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 347,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df['demand']=train_df.groupby(['interval','pickup_longitude','pickup_latitude'])['interval'].transform('count')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 348,
   "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>key</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>dropoff_longitude</th>\n",
       "      <th>dropoff_latitude</th>\n",
       "      <th>passenger_count</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "      <th>pickup_month</th>\n",
       "      <th>pickup_date</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_year</th>\n",
       "      <th>id</th>\n",
       "      <th>interval_x</th>\n",
       "      <th>interval_time</th>\n",
       "      <th>demand</th>\n",
       "      <th>distance</th>\n",
       "      <th>interval_y</th>\n",
       "      <th>interval</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2009-06-15 17:26:21.0000001</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-06-15 17:26:21</td>\n",
       "      <td>-73.8</td>\n",
       "      <td>40.72</td>\n",
       "      <td>-73.841610</td>\n",
       "      <td>40.712278</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>2009-06-15</td>\n",
       "      <td>17</td>\n",
       "      <td>2009</td>\n",
       "      <td>0</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "      <td>17:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>4.365758</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "      <td>2009-06-15 17:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2009-08-03 23:01:00.00000077</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-08-03 23:01:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.76</td>\n",
       "      <td>-73.984915</td>\n",
       "      <td>40.768142</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>2009-08-03</td>\n",
       "      <td>23</td>\n",
       "      <td>2009</td>\n",
       "      <td>1</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "      <td>23:00:00</td>\n",
       "      <td>1</td>\n",
       "      <td>2.175726</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "      <td>2009-08-03 23:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2011-07-11 13:08:09.0000001</td>\n",
       "      <td>10.9</td>\n",
       "      <td>2011-07-11 13:08:09</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.75</td>\n",
       "      <td>-74.014800</td>\n",
       "      <td>40.718000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>2011-07-11</td>\n",
       "      <td>13</td>\n",
       "      <td>2011</td>\n",
       "      <td>2</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "      <td>13:00:00</td>\n",
       "      <td>3</td>\n",
       "      <td>4.805557</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "      <td>2011-07-11 13:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2012-12-17 22:58:00.000000163</td>\n",
       "      <td>29.0</td>\n",
       "      <td>2012-12-17 22:58:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.77</td>\n",
       "      <td>-73.962497</td>\n",
       "      <td>40.677382</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>2012-12-17</td>\n",
       "      <td>22</td>\n",
       "      <td>2012</td>\n",
       "      <td>3</td>\n",
       "      <td>2012-12-17 22:00:00</td>\n",
       "      <td>22:30:00</td>\n",
       "      <td>2</td>\n",
       "      <td>13.461272</td>\n",
       "      <td>2012-12-17 22:30:00</td>\n",
       "      <td>2012-12-17 22:30:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2011-02-28 22:03:00.000000115</td>\n",
       "      <td>6.1</td>\n",
       "      <td>2011-02-28 22:03:00</td>\n",
       "      <td>-74.0</td>\n",
       "      <td>40.73</td>\n",
       "      <td>-73.978080</td>\n",
       "      <td>40.721325</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2011-02-28</td>\n",
       "      <td>22</td>\n",
       "      <td>2011</td>\n",
       "      <td>4</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "      <td>22:00:00</td>\n",
       "      <td>4</td>\n",
       "      <td>2.811899</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "      <td>2011-02-28 22:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             key  fare_amount     pickup_datetime  \\\n",
       "0    2009-06-15 17:26:21.0000001          4.5 2009-06-15 17:26:21   \n",
       "1   2009-08-03 23:01:00.00000077          4.5 2009-08-03 23:01:00   \n",
       "2    2011-07-11 13:08:09.0000001         10.9 2011-07-11 13:08:09   \n",
       "3  2012-12-17 22:58:00.000000163         29.0 2012-12-17 22:58:00   \n",
       "4  2011-02-28 22:03:00.000000115          6.1 2011-02-28 22:03:00   \n",
       "\n",
       "   pickup_longitude  pickup_latitude  dropoff_longitude  dropoff_latitude  \\\n",
       "0             -73.8            40.72         -73.841610         40.712278   \n",
       "1             -74.0            40.76         -73.984915         40.768142   \n",
       "2             -74.0            40.75         -74.014800         40.718000   \n",
       "3             -74.0            40.77         -73.962497         40.677382   \n",
       "4             -74.0            40.73         -73.978080         40.721325   \n",
       "\n",
       "   passenger_count  pickup_day_of_week  pickup_month pickup_date  pickup_hour  \\\n",
       "0                1                   0             6  2009-06-15           17   \n",
       "1                5                   0             8  2009-08-03           23   \n",
       "2                1                   0             7  2011-07-11           13   \n",
       "3                3                   0            12  2012-12-17           22   \n",
       "4                1                   0             2  2011-02-28           22   \n",
       "\n",
       "   pickup_year  id          interval_x interval_time  demand   distance  \\\n",
       "0         2009   0 2009-06-15 17:00:00      17:00:00       1   4.365758   \n",
       "1         2009   1 2009-08-03 23:00:00      23:00:00       1   2.175726   \n",
       "2         2011   2 2011-07-11 13:00:00      13:00:00       3   4.805557   \n",
       "3         2012   3 2012-12-17 22:00:00      22:30:00       2  13.461272   \n",
       "4         2011   4 2011-02-28 22:00:00      22:00:00       4   2.811899   \n",
       "\n",
       "           interval_y            interval  \n",
       "0 2009-06-15 17:00:00 2009-06-15 17:00:00  \n",
       "1 2009-08-03 23:00:00 2009-08-03 23:00:00  \n",
       "2 2011-07-11 13:00:00 2011-07-11 13:00:00  \n",
       "3 2012-12-17 22:30:00 2012-12-17 22:30:00  \n",
       "4 2011-02-28 22:00:00 2011-02-28 22:00:00  "
      ]
     },
     "execution_count": 348,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 349,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x35bb7e10>"
      ]
     },
     "execution_count": 349,
     "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": [
    "train_df['demand'].value_counts().sort_index().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 350,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x48bef0f0>"
      ]
     },
     "execution_count": 350,
     "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": [
    "train_df.groupby('demand')['fare_amount'].mean().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 351,
   "metadata": {},
   "outputs": [],
   "source": [
    "def distance(lat1, lon1, lat2, lon2):\n",
    "    \"\"\"\n",
    "    calculates the Manhattan distance between 2 points using their coordinates   \n",
    "    Returns\n",
    "    -------\n",
    "    d: float\n",
    "        The Manhattan distance between the two points in kilometers\n",
    "    \"\"\"\n",
    "    d = haversine((lat1, lon1), (lat2, lon1)) + haversine((lat2, lon1), (lat2, lon2))\n",
    "    return d\n",
    "\n",
    "# The distance is calculated in kilometers\n",
    "train_df[\"distance\"] = train_df.apply(lambda row: distance(row[\"pickup_latitude\"], \n",
    "                                               row[\"pickup_longitude\"], \n",
    "                                               row[\"dropoff_latitude\"], \n",
    "                                               row[\"dropoff_longitude\"]), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 352,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0          4.365758\n",
       "1          2.175726\n",
       "2          4.805557\n",
       "3         13.461272\n",
       "4          2.811899\n",
       "5          2.236704\n",
       "6          2.759051\n",
       "7          3.779220\n",
       "8          7.227713\n",
       "9          0.758220\n",
       "10         0.936794\n",
       "11         6.980522\n",
       "12         0.858484\n",
       "13         4.409194\n",
       "14         5.597084\n",
       "15        20.155024\n",
       "16         2.769499\n",
       "17         3.462153\n",
       "18         4.028002\n",
       "19         2.386084\n",
       "20         2.323553\n",
       "21         5.556529\n",
       "22         7.693177\n",
       "23         3.795712\n",
       "24         5.689749\n",
       "25         1.425140\n",
       "26         2.069988\n",
       "27         4.691100\n",
       "28         3.058164\n",
       "29         0.000000\n",
       "            ...    \n",
       "999970     8.208024\n",
       "999971     4.013119\n",
       "999972     4.445124\n",
       "999973     4.751429\n",
       "999974     3.350710\n",
       "999975     4.375909\n",
       "999976     1.765962\n",
       "999977    33.232364\n",
       "999978     8.190792\n",
       "999979     5.135354\n",
       "999980    21.703501\n",
       "999981    10.432956\n",
       "999982     2.734151\n",
       "999983     1.729356\n",
       "999984     2.445260\n",
       "999985     4.390693\n",
       "999986     4.740688\n",
       "999987    17.242261\n",
       "999988     7.503730\n",
       "999989     5.500275\n",
       "999990     5.495309\n",
       "999991     2.917594\n",
       "999992     5.268885\n",
       "999993    14.528045\n",
       "999994     1.708432\n",
       "999995     3.271881\n",
       "999996     1.904782\n",
       "999997     7.750297\n",
       "999998     0.589235\n",
       "999999     3.043615\n",
       "Name: distance, Length: 1000000, dtype: float64"
      ]
     },
     "execution_count": 352,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df[\"distance\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 353,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x8771c8d0>"
      ]
     },
     "execution_count": 353,
     "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": [
    "train_df.groupby('distance')['fare_amount'].mean().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 354,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0          4.50\n",
       "1          4.50\n",
       "2         10.90\n",
       "3         29.00\n",
       "4          6.10\n",
       "5          6.90\n",
       "6          6.10\n",
       "7          7.70\n",
       "8          8.10\n",
       "9         56.80\n",
       "10         4.50\n",
       "11        22.89\n",
       "12         4.10\n",
       "13        13.50\n",
       "14         6.10\n",
       "15        44.00\n",
       "16         6.90\n",
       "17         7.30\n",
       "18         5.00\n",
       "19         4.50\n",
       "20         4.90\n",
       "21         7.00\n",
       "22        14.00\n",
       "23         6.50\n",
       "24        16.10\n",
       "25         6.50\n",
       "26         7.50\n",
       "27         9.70\n",
       "28        10.50\n",
       "29         4.90\n",
       "          ...  \n",
       "999970    15.50\n",
       "999971    15.00\n",
       "999972    11.00\n",
       "999973     7.00\n",
       "999974     4.50\n",
       "999975     8.90\n",
       "999976     7.30\n",
       "999977    57.33\n",
       "999978    12.10\n",
       "999979     4.90\n",
       "999980    48.33\n",
       "999981    15.70\n",
       "999982     6.50\n",
       "999983     6.50\n",
       "999984     6.50\n",
       "999985     5.30\n",
       "999986    10.00\n",
       "999987    17.30\n",
       "999988    10.90\n",
       "999989    10.10\n",
       "999990    13.70\n",
       "999991     9.50\n",
       "999992    31.83\n",
       "999993    19.30\n",
       "999994     3.70\n",
       "999995    14.90\n",
       "999996    11.50\n",
       "999997    24.00\n",
       "999998     4.90\n",
       "999999     7.50\n",
       "Name: fare_amount, Length: 1000000, dtype: float64"
      ]
     },
     "execution_count": 354,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df[\"fare_amount\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 355,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "22138.03906981074"
      ]
     },
     "execution_count": 355,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df['distance'].max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 356,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df.to_csv(\"train_formatted_demand.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 357,
   "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>fare_amount</th>\n",
       "      <td>1000000.0</td>\n",
       "      <td>11.348079</td>\n",
       "      <td>9.822090</td>\n",
       "      <td>-44.900000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>8.500000</td>\n",
       "      <td>12.500000</td>\n",
       "      <td>500.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pickup_longitude</th>\n",
       "      <td>1000000.0</td>\n",
       "      <td>-72.541486</td>\n",
       "      <td>12.059714</td>\n",
       "      <td>-3377.700000</td>\n",
       "      <td>-74.000000</td>\n",
       "      <td>-74.000000</td>\n",
       "      <td>-74.000000</td>\n",
       "      <td>2522.300000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pickup_latitude</th>\n",
       "      <td>1000000.0</td>\n",
       "      <td>39.928901</td>\n",
       "      <td>7.626147</td>\n",
       "      <td>-3116.290000</td>\n",
       "      <td>40.730000</td>\n",
       "      <td>40.750000</td>\n",
       "      <td>40.770000</td>\n",
       "      <td>2621.630000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dropoff_longitude</th>\n",
       "      <td>999990.0</td>\n",
       "      <td>-72.527860</td>\n",
       "      <td>11.324494</td>\n",
       "      <td>-3383.296608</td>\n",
       "      <td>-73.991385</td>\n",
       "      <td>-73.980135</td>\n",
       "      <td>-73.963654</td>\n",
       "      <td>45.581619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dropoff_latitude</th>\n",
       "      <td>999990.0</td>\n",
       "      <td>39.919954</td>\n",
       "      <td>8.201418</td>\n",
       "      <td>-3114.338567</td>\n",
       "      <td>40.734046</td>\n",
       "      <td>40.753166</td>\n",
       "      <td>40.768129</td>\n",
       "      <td>1651.553433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>passenger_count</th>\n",
       "      <td>1000000.0</td>\n",
       "      <td>1.684924</td>\n",
       "      <td>1.323911</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>208.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pickup_day_of_week</th>\n",
       "      <td>1000000.0</td>\n",
       "      <td>3.039856</td>\n",
       "      <td>1.949970</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pickup_month</th>\n",
       "      <td>1000000.0</td>\n",
       "      <td>6.267875</td>\n",
       "      <td>3.436243</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>12.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pickup_hour</th>\n",
       "      <td>1000000.0</td>\n",
       "      <td>13.509477</td>\n",
       "      <td>6.513840</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>23.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pickup_year</th>\n",
       "      <td>1000000.0</td>\n",
       "      <td>2011.741106</td>\n",
       "      <td>1.860754</td>\n",
       "      <td>2009.000000</td>\n",
       "      <td>2010.000000</td>\n",
       "      <td>2012.000000</td>\n",
       "      <td>2013.000000</td>\n",
       "      <td>2015.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <td>1000000.0</td>\n",
       "      <td>499999.500000</td>\n",
       "      <td>288675.278933</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>249999.750000</td>\n",
       "      <td>499999.500000</td>\n",
       "      <td>749999.250000</td>\n",
       "      <td>999999.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>demand</th>\n",
       "      <td>1000000.0</td>\n",
       "      <td>2.228720</td>\n",
       "      <td>1.384743</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>13.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>distance</th>\n",
       "      <td>999990.0</td>\n",
       "      <td>25.555126</td>\n",
       "      <td>482.917576</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.237175</td>\n",
       "      <td>3.617448</td>\n",
       "      <td>5.619264</td>\n",
       "      <td>22138.039070</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        count           mean            std          min  \\\n",
       "fare_amount         1000000.0      11.348079       9.822090   -44.900000   \n",
       "pickup_longitude    1000000.0     -72.541486      12.059714 -3377.700000   \n",
       "pickup_latitude     1000000.0      39.928901       7.626147 -3116.290000   \n",
       "dropoff_longitude    999990.0     -72.527860      11.324494 -3383.296608   \n",
       "dropoff_latitude     999990.0      39.919954       8.201418 -3114.338567   \n",
       "passenger_count     1000000.0       1.684924       1.323911     0.000000   \n",
       "pickup_day_of_week  1000000.0       3.039856       1.949970     0.000000   \n",
       "pickup_month        1000000.0       6.267875       3.436243     1.000000   \n",
       "pickup_hour         1000000.0      13.509477       6.513840     0.000000   \n",
       "pickup_year         1000000.0    2011.741106       1.860754  2009.000000   \n",
       "id                  1000000.0  499999.500000  288675.278933     0.000000   \n",
       "demand              1000000.0       2.228720       1.384743     1.000000   \n",
       "distance             999990.0      25.555126     482.917576     0.000000   \n",
       "\n",
       "                              25%            50%            75%            max  \n",
       "fare_amount              6.000000       8.500000      12.500000     500.000000  \n",
       "pickup_longitude       -74.000000     -74.000000     -74.000000    2522.300000  \n",
       "pickup_latitude         40.730000      40.750000      40.770000    2621.630000  \n",
       "dropoff_longitude      -73.991385     -73.980135     -73.963654      45.581619  \n",
       "dropoff_latitude        40.734046      40.753166      40.768129    1651.553433  \n",
       "passenger_count          1.000000       1.000000       2.000000     208.000000  \n",
       "pickup_day_of_week       1.000000       3.000000       5.000000       6.000000  \n",
       "pickup_month             3.000000       6.000000       9.000000      12.000000  \n",
       "pickup_hour              9.000000      14.000000      19.000000      23.000000  \n",
       "pickup_year           2010.000000    2012.000000    2013.000000    2015.000000  \n",
       "id                  249999.750000  499999.500000  749999.250000  999999.000000  \n",
       "demand                   1.000000       2.000000       3.000000      13.000000  \n",
       "distance                 2.237175       3.617448       5.619264   22138.039070  "
      ]
     },
     "execution_count": 357,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.describe().transpose()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 358,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        2009-06-15 17:00:00\n",
       "1        2009-08-03 23:00:00\n",
       "2        2011-07-11 13:00:00\n",
       "3        2012-12-17 22:30:00\n",
       "4        2011-02-28 22:00:00\n",
       "5        2010-03-29 13:00:00\n",
       "6        2012-03-26 02:00:00\n",
       "7        2011-04-18 14:00:00\n",
       "8        2012-07-02 22:30:00\n",
       "9        2012-10-01 00:00:00\n",
       "10       2009-09-21 20:30:00\n",
       "11       2012-06-18 21:30:00\n",
       "12       2009-06-15 14:30:00\n",
       "13       2013-05-06 07:30:00\n",
       "14       2010-01-11 17:00:00\n",
       "15       2013-06-17 16:00:00\n",
       "16       2012-03-26 17:00:00\n",
       "17       2009-05-04 15:30:00\n",
       "18       2015-02-02 10:30:00\n",
       "19       2015-06-15 15:30:00\n",
       "20       2011-02-07 09:00:00\n",
       "21       2012-09-10 22:30:00\n",
       "22       2014-10-27 17:00:00\n",
       "23       2009-03-16 10:00:00\n",
       "24       2009-08-31 01:00:00\n",
       "25       2013-02-11 09:00:00\n",
       "26       2015-03-02 13:30:00\n",
       "27       2012-04-02 22:00:00\n",
       "28       2010-02-22 09:30:00\n",
       "29       2009-12-28 20:00:00\n",
       "                 ...        \n",
       "999970   2014-02-09 20:30:00\n",
       "999971   2014-05-11 16:00:00\n",
       "999972   2014-06-01 12:00:00\n",
       "999973   2013-12-08 21:00:00\n",
       "999974   2014-11-02 00:00:00\n",
       "999975   2012-06-03 19:30:00\n",
       "999976   2009-04-19 17:00:00\n",
       "999977   2014-02-23 17:30:00\n",
       "999978   2010-01-03 15:00:00\n",
       "999979   2012-04-29 15:30:00\n",
       "999980   2013-10-13 15:30:00\n",
       "999981   2010-02-28 06:30:00\n",
       "999982   2015-02-08 11:30:00\n",
       "999983   2011-12-25 11:00:00\n",
       "999984   2012-01-29 00:30:00\n",
       "999985   2011-07-31 03:30:00\n",
       "999986   2012-11-18 01:00:00\n",
       "999987   2009-04-19 23:30:00\n",
       "999988   2011-09-18 15:00:00\n",
       "999989   2011-10-09 23:30:00\n",
       "999990   2012-08-12 17:00:00\n",
       "999991   2014-08-03 20:00:00\n",
       "999992   2015-03-08 23:00:00\n",
       "999993   2011-11-20 12:00:00\n",
       "999994   2009-07-26 06:30:00\n",
       "999995   2012-06-24 15:30:00\n",
       "999996   2012-09-30 04:00:00\n",
       "999997   2012-10-07 12:30:00\n",
       "999998   2011-04-17 20:30:00\n",
       "999999   2015-05-10 13:00:00\n",
       "Name: interval, Length: 1000000, dtype: datetime64[ns]"
      ]
     },
     "execution_count": 358,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df[\"interval\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 359,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 1000000 entries, 0 to 999999\n",
      "Data columns (total 20 columns):\n",
      "key                   1000000 non-null object\n",
      "fare_amount           1000000 non-null float64\n",
      "pickup_datetime       1000000 non-null datetime64[ns]\n",
      "pickup_longitude      1000000 non-null float64\n",
      "pickup_latitude       1000000 non-null float64\n",
      "dropoff_longitude     999990 non-null float64\n",
      "dropoff_latitude      999990 non-null float64\n",
      "passenger_count       1000000 non-null int64\n",
      "pickup_day_of_week    1000000 non-null int64\n",
      "pickup_month          1000000 non-null int64\n",
      "pickup_date           1000000 non-null object\n",
      "pickup_hour           1000000 non-null int64\n",
      "pickup_year           1000000 non-null int64\n",
      "id                    1000000 non-null int64\n",
      "interval_x            1000000 non-null datetime64[ns]\n",
      "interval_time         1000000 non-null object\n",
      "demand                1000000 non-null int64\n",
      "distance              999990 non-null float64\n",
      "interval_y            1000000 non-null datetime64[ns]\n",
      "interval              1000000 non-null datetime64[ns]\n",
      "dtypes: datetime64[ns](4), float64(6), int64(7), object(3)\n",
      "memory usage: 160.2+ MB\n"
     ]
    }
   ],
   "source": [
    "train_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('formatted_demand.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "df = df[[\"demand\",\"fare_amount\",\"pickup_longitude\",\"pickup_latitude\",\"pickup_hour\",\"interval\",\"pickup_day_of_week\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.interval = pd.to_datetime(df.interval)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "df =df.loc[df.pickup_longitude<80]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "df=df.loc[df.demand<7]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": true
   },
   "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>demand</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>988663.000000</td>\n",
       "      <td>988663.000000</td>\n",
       "      <td>988663.000000</td>\n",
       "      <td>988663.000000</td>\n",
       "      <td>988663.000000</td>\n",
       "      <td>988663.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2.168129</td>\n",
       "      <td>11.364076</td>\n",
       "      <td>-72.529627</td>\n",
       "      <td>39.915132</td>\n",
       "      <td>13.478199</td>\n",
       "      <td>3.044598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.267691</td>\n",
       "      <td>9.846900</td>\n",
       "      <td>11.632511</td>\n",
       "      <td>7.019797</td>\n",
       "      <td>6.516761</td>\n",
       "      <td>1.951764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>-44.900000</td>\n",
       "      <td>-3377.700000</td>\n",
       "      <td>-3116.290000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>-74.000000</td>\n",
       "      <td>40.730000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>8.500000</td>\n",
       "      <td>-74.000000</td>\n",
       "      <td>40.750000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>12.500000</td>\n",
       "      <td>-74.000000</td>\n",
       "      <td>40.770000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>5.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>6.000000</td>\n",
       "      <td>500.000000</td>\n",
       "      <td>40.900000</td>\n",
       "      <td>424.080000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              demand    fare_amount  pickup_longitude  pickup_latitude  \\\n",
       "count  988663.000000  988663.000000     988663.000000    988663.000000   \n",
       "mean        2.168129      11.364076        -72.529627        39.915132   \n",
       "std         1.267691       9.846900         11.632511         7.019797   \n",
       "min         1.000000     -44.900000      -3377.700000     -3116.290000   \n",
       "25%         1.000000       6.000000        -74.000000        40.730000   \n",
       "50%         2.000000       8.500000        -74.000000        40.750000   \n",
       "75%         3.000000      12.500000        -74.000000        40.770000   \n",
       "max         6.000000     500.000000         40.900000       424.080000   \n",
       "\n",
       "         pickup_hour  pickup_day_of_week  \n",
       "count  988663.000000       988663.000000  \n",
       "mean       13.478199            3.044598  \n",
       "std         6.516761            1.951764  \n",
       "min         0.000000            0.000000  \n",
       "25%         9.000000            1.000000  \n",
       "50%        14.000000            3.000000  \n",
       "75%        19.000000            5.000000  \n",
       "max        23.000000            6.000000  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>demand</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>demand</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.112045</td>\n",
       "      <td>-0.094769</td>\n",
       "      <td>0.089643</td>\n",
       "      <td>0.119948</td>\n",
       "      <td>-0.021290</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fare_amount</th>\n",
       "      <td>-0.112045</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.008863</td>\n",
       "      <td>-0.008412</td>\n",
       "      <td>-0.018264</td>\n",
       "      <td>0.002116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pickup_longitude</th>\n",
       "      <td>-0.094769</td>\n",
       "      <td>0.008863</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.731132</td>\n",
       "      <td>0.001008</td>\n",
       "      <td>0.001039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pickup_latitude</th>\n",
       "      <td>0.089643</td>\n",
       "      <td>-0.008412</td>\n",
       "      <td>-0.731132</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000418</td>\n",
       "      <td>-0.003277</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pickup_hour</th>\n",
       "      <td>0.119948</td>\n",
       "      <td>-0.018264</td>\n",
       "      <td>0.001008</td>\n",
       "      <td>0.000418</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.085874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pickup_day_of_week</th>\n",
       "      <td>-0.021290</td>\n",
       "      <td>0.002116</td>\n",
       "      <td>0.001039</td>\n",
       "      <td>-0.003277</td>\n",
       "      <td>-0.085874</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      demand  fare_amount  pickup_longitude  pickup_latitude  \\\n",
       "demand              1.000000    -0.112045         -0.094769         0.089643   \n",
       "fare_amount        -0.112045     1.000000          0.008863        -0.008412   \n",
       "pickup_longitude   -0.094769     0.008863          1.000000        -0.731132   \n",
       "pickup_latitude     0.089643    -0.008412         -0.731132         1.000000   \n",
       "pickup_hour         0.119948    -0.018264          0.001008         0.000418   \n",
       "pickup_day_of_week -0.021290     0.002116          0.001039        -0.003277   \n",
       "\n",
       "                    pickup_hour  pickup_day_of_week  \n",
       "demand                 0.119948           -0.021290  \n",
       "fare_amount           -0.018264            0.002116  \n",
       "pickup_longitude       0.001008            0.001039  \n",
       "pickup_latitude        0.000418           -0.003277  \n",
       "pickup_hour            1.000000           -0.085874  \n",
       "pickup_day_of_week    -0.085874            1.000000  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.index = df.interval"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.resample(rule='60T', closed='left', label='left', base=17).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[df.columns].apply(pd.to_numeric, errors='coerce')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>demand</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>interval</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2008-12-31 23:17:00</th>\n",
       "      <td>1.333333</td>\n",
       "      <td>11.666667</td>\n",
       "      <td>-74.000000</td>\n",
       "      <td>40.768333</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-01-01 00:17:00</th>\n",
       "      <td>1.400000</td>\n",
       "      <td>9.666667</td>\n",
       "      <td>-73.993333</td>\n",
       "      <td>40.754667</td>\n",
       "      <td>0.533333</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-01-01 01:17:00</th>\n",
       "      <td>1.200000</td>\n",
       "      <td>8.800000</td>\n",
       "      <td>-74.000000</td>\n",
       "      <td>40.729000</td>\n",
       "      <td>1.600000</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-01-01 02:17:00</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>12.811111</td>\n",
       "      <td>-73.988889</td>\n",
       "      <td>40.763333</td>\n",
       "      <td>2.555556</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-01-01 03:17:00</th>\n",
       "      <td>1.400000</td>\n",
       "      <td>8.470000</td>\n",
       "      <td>-74.000000</td>\n",
       "      <td>40.735000</td>\n",
       "      <td>3.600000</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       demand  fare_amount  pickup_longitude  pickup_latitude  \\\n",
       "interval                                                                        \n",
       "2008-12-31 23:17:00  1.333333    11.666667        -74.000000        40.768333   \n",
       "2009-01-01 00:17:00  1.400000     9.666667        -73.993333        40.754667   \n",
       "2009-01-01 01:17:00  1.200000     8.800000        -74.000000        40.729000   \n",
       "2009-01-01 02:17:00  1.000000    12.811111        -73.988889        40.763333   \n",
       "2009-01-01 03:17:00  1.400000     8.470000        -74.000000        40.735000   \n",
       "\n",
       "                     pickup_hour  pickup_day_of_week  \n",
       "interval                                              \n",
       "2008-12-31 23:17:00     0.000000                 3.0  \n",
       "2009-01-01 00:17:00     0.533333                 3.0  \n",
       "2009-01-01 01:17:00     1.600000                 3.0  \n",
       "2009-01-01 02:17:00     2.555556                 3.0  \n",
       "2009-01-01 03:17:00     3.600000                 3.0  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 332,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 332,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().values.any()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<60 * Minutes>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.index.freq"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.fillna(df.mean(), inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "demand                float64\n",
       "fare_amount           float64\n",
       "pickup_longitude      float64\n",
       "pickup_latitude       float64\n",
       "pickup_hour           float64\n",
       "pickup_day_of_week    float64\n",
       "dtype: object"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_size = int(len(df) * 0.80)\n",
    "train, test = df[0:train_size], df[train_size:len(df)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "demand                float64\n",
       "fare_amount           float64\n",
       "pickup_longitude      float64\n",
       "pickup_latitude       float64\n",
       "pickup_hour           float64\n",
       "pickup_day_of_week    float64\n",
       "dtype: object"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x26da6eb8>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.demand.plot(figsize=(15,8), title= 'Daily Ridership', fontsize=14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>demand</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>interval</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-06-30 19:17:00</th>\n",
       "      <td>2.769231</td>\n",
       "      <td>9.653846</td>\n",
       "      <td>-71.153846</td>\n",
       "      <td>39.185000</td>\n",
       "      <td>19.423077</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-06-30 20:17:00</th>\n",
       "      <td>1.875000</td>\n",
       "      <td>12.375000</td>\n",
       "      <td>-69.362500</td>\n",
       "      <td>38.198750</td>\n",
       "      <td>20.500000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-06-30 21:17:00</th>\n",
       "      <td>1.526316</td>\n",
       "      <td>16.556842</td>\n",
       "      <td>-70.084211</td>\n",
       "      <td>38.613158</td>\n",
       "      <td>21.473684</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-06-30 22:17:00</th>\n",
       "      <td>2.583333</td>\n",
       "      <td>14.522500</td>\n",
       "      <td>-70.908333</td>\n",
       "      <td>39.040417</td>\n",
       "      <td>22.375000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-06-30 23:17:00</th>\n",
       "      <td>1.888889</td>\n",
       "      <td>12.722222</td>\n",
       "      <td>-65.777778</td>\n",
       "      <td>36.220000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       demand  fare_amount  pickup_longitude  pickup_latitude  \\\n",
       "interval                                                                        \n",
       "2015-06-30 19:17:00  2.769231     9.653846        -71.153846        39.185000   \n",
       "2015-06-30 20:17:00  1.875000    12.375000        -69.362500        38.198750   \n",
       "2015-06-30 21:17:00  1.526316    16.556842        -70.084211        38.613158   \n",
       "2015-06-30 22:17:00  2.583333    14.522500        -70.908333        39.040417   \n",
       "2015-06-30 23:17:00  1.888889    12.722222        -65.777778        36.220000   \n",
       "\n",
       "                     pickup_hour  pickup_day_of_week  \n",
       "interval                                              \n",
       "2015-06-30 19:17:00    19.423077                 1.0  \n",
       "2015-06-30 20:17:00    20.500000                 1.0  \n",
       "2015-06-30 21:17:00    21.473684                 1.0  \n",
       "2015-06-30 22:17:00    22.375000                 1.0  \n",
       "2015-06-30 23:17:00    23.000000                 1.0  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x14ce2c88>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train.demand.plot(figsize=(15,8), title= 'Daily Ridership', fontsize=14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x26d960f0>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "test.demand.plot(figsize=(15,8), title= 'Daily Ridership', fontsize=14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = train[['pickup_hour','pickup_longitude','pickup_latitude','pickup_day_of_week']]\n",
    "x_test = test[['pickup_hour','pickup_longitude','pickup_latitude','pickup_day_of_week']]\n",
    "y_train = train[['fare_amount']]\n",
    "y_test = test[['fare_amount']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    56929.000000\n",
       "mean         1.916833\n",
       "std          0.614832\n",
       "min          1.000000\n",
       "25%          1.444444\n",
       "50%          1.900000\n",
       "75%          2.333333\n",
       "max          5.000000\n",
       "Name: demand, dtype: float64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.demand.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train= y_train.apply(np.int64)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['pickup_hour', 'pickup_longitude', 'pickup_latitude',\n",
       "       'pickup_day_of_week'], dtype=object)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.columns.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test= y_test.apply(np.int64)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.tsa.vector_ar.var_model import VAR\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<60 * Minutes>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.index.freq"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pickup_hour           float64\n",
       "pickup_longitude      float64\n",
       "pickup_latitude       float64\n",
       "pickup_day_of_week    float64\n",
       "dtype: object"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "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>pickup_hour</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>45543.000000</td>\n",
       "      <td>45543.000000</td>\n",
       "      <td>45543.000000</td>\n",
       "      <td>45543.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>11.570437</td>\n",
       "      <td>-72.520607</td>\n",
       "      <td>39.901661</td>\n",
       "      <td>3.006394</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>6.505984</td>\n",
       "      <td>4.045862</td>\n",
       "      <td>2.445415</td>\n",
       "      <td>1.979397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-397.577778</td>\n",
       "      <td>-62.404839</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>6.333333</td>\n",
       "      <td>-73.996875</td>\n",
       "      <td>39.428871</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>11.560000</td>\n",
       "      <td>-73.986957</td>\n",
       "      <td>40.746111</td>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>17.333333</td>\n",
       "      <td>-71.533333</td>\n",
       "      <td>40.754167</td>\n",
       "      <td>5.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>23.000000</td>\n",
       "      <td>9.680000</td>\n",
       "      <td>77.167000</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        pickup_hour  pickup_longitude  pickup_latitude  pickup_day_of_week\n",
       "count  45543.000000      45543.000000     45543.000000        45543.000000\n",
       "mean      11.570437        -72.520607        39.901661            3.006394\n",
       "std        6.505984          4.045862         2.445415            1.979397\n",
       "min        0.000000       -397.577778       -62.404839            0.000000\n",
       "25%        6.333333        -73.996875        39.428871            1.000000\n",
       "50%       11.560000        -73.986957        40.746111            3.000000\n",
       "75%       17.333333        -71.533333        40.754167            5.000000\n",
       "max       23.000000          9.680000        77.167000            6.000000"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import datasets, linear_model\n",
    "from sklearn.metrics import mean_squared_error, r2_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create linear regression object\n",
    "regr = linear_model.LinearRegression()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regr.fit(x_train, y_train)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_test = x_test[np.isfinite(x_test['pickup_hour'])]\n",
    "y_test = y_test[np.isfinite(y_test['fare_amount'])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = regr.predict(x_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficients: \n",
      " [[-0.07462924  0.00893669 -0.00940529 -0.01715573]]\n"
     ]
    }
   ],
   "source": [
    "# The coefficients\n",
    "print('Coefficients: \\n', regr.coef_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean squared error: 25.14\n",
      "Variance score: -0.21\n"
     ]
    }
   ],
   "source": [
    "print(\"Mean squared error: %.2f\"\n",
    "      % mean_squared_error(y_test, y_pred))\n",
    "# Explained variance score: 1 is perfect prediction\n",
    "print('Variance score: %.2f' % r2_score(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\user\\Anaconda\\lib\\site-packages\\sklearn\\ensemble\\weight_boosting.py:29: DeprecationWarning:\n",
      "\n",
      "numpy.core.umath_tests is an internal NumPy module and should not be imported. It will be removed in a future NumPy release.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "regressor = RandomForestRegressor(n_estimators=20, random_state=0)  \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\user\\Anaconda\\lib\\site-packages\\ipykernel_launcher.py:1: DataConversionWarning:\n",
      "\n",
      "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n",
      "\n"
     ]
    }
   ],
   "source": [
    "regressor.fit(x_train, y_train)  \n",
    "y_pred = regressor.predict(x_test) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean squared error: 0.26\n",
      "Variance score: 0.20\n"
     ]
    }
   ],
   "source": [
    "print(\"Mean squared error: %.2f\"\n",
    "      % mean_squared_error(y_test, y_pred))\n",
    "# Explained variance score: 1 is perfect prediction\n",
    "print('Variance score: %.2f' % r2_score(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "regressor  = DecisionTreeRegressor(max_depth=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "regressor.fit(x_train, y_train)  \n",
    "y_pred = regressor.predict(x_test) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean squared error: 0.25\n",
      "Variance score: 0.23\n"
     ]
    }
   ],
   "source": [
    "print(\"Mean squared error: %.2f\"\n",
    "    % mean_squared_error(y_test, y_pred))\n",
    "# Explained variance score: 1 is perfect prediction\n",
    "print('Variance score: %.2f' % r2_score(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [],
   "source": [
    "import xgboost as xgb\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [],
   "source": [
    "xg_reg = xgb.XGBRegressor(objective ='reg:linear', colsample_bytree = 0.3, learning_rate = 0.2,\n",
    "                max_depth = 5, alpha = 10, n_estimators = 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [],
   "source": [
    "xg_reg.fit(x_train,y_train)\n",
    "preds = xg_reg.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean squared error: 0.25\n",
      "Variance score: 0.23\n"
     ]
    }
   ],
   "source": [
    "print(\"Mean squared error: %.2f\"\n",
    "    % mean_squared_error(y_test, y_pred))\n",
    "# Explained variance score: 1 is perfect prediction\n",
    "print('Variance score: %.2f' % r2_score(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.tsa.ar_model import AR\n",
    "from sklearn.metrics import mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<60 * Minutes>"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.index.freq"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "# train autoregression\n",
    "model = AR(y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_fit = model.fit()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Lag: 55\n"
     ]
    }
   ],
   "source": [
    "print('Lag: %s' % model_fit.k_ar)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficients: const         0.702579\n",
      "L1.demand     0.162802\n",
      "L2.demand     0.095635\n",
      "L3.demand     0.052560\n",
      "L4.demand     0.024558\n",
      "L5.demand     0.004197\n",
      "L6.demand    -0.009174\n",
      "L7.demand    -0.028387\n",
      "L8.demand    -0.028328\n",
      "L9.demand    -0.015319\n",
      "L10.demand    0.012529\n",
      "L11.demand    0.008629\n",
      "L12.demand    0.014961\n",
      "L13.demand    0.003621\n",
      "L14.demand    0.007762\n",
      "L15.demand   -0.018156\n",
      "L16.demand   -0.030276\n",
      "L17.demand   -0.024111\n",
      "L18.demand   -0.018494\n",
      "L19.demand   -0.000753\n",
      "L20.demand    0.005049\n",
      "L21.demand    0.016303\n",
      "L22.demand    0.041006\n",
      "L23.demand    0.064503\n",
      "L24.demand    0.099904\n",
      "L25.demand    0.061250\n",
      "L26.demand    0.036749\n",
      "L27.demand    0.009136\n",
      "L28.demand    0.009229\n",
      "L29.demand    0.003180\n",
      "L30.demand   -0.012268\n",
      "L31.demand   -0.023193\n",
      "L32.demand   -0.027028\n",
      "L33.demand   -0.015128\n",
      "L34.demand    0.001276\n",
      "L35.demand   -0.000301\n",
      "L36.demand    0.003961\n",
      "L37.demand   -0.007007\n",
      "L38.demand   -0.006221\n",
      "L39.demand   -0.002771\n",
      "L40.demand   -0.007749\n",
      "L41.demand   -0.005110\n",
      "L42.demand    0.006452\n",
      "L43.demand    0.001555\n",
      "L44.demand    0.004115\n",
      "L45.demand    0.006520\n",
      "L46.demand    0.000595\n",
      "L47.demand    0.027499\n",
      "L48.demand    0.045227\n",
      "L49.demand    0.009779\n",
      "L50.demand    0.004537\n",
      "L51.demand   -0.005508\n",
      "L52.demand   -0.005599\n",
      "L53.demand   -0.005104\n",
      "L54.demand   -0.005018\n",
      "L55.demand   -0.006923\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print('Coefficients: %s' % model_fit.params)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = model_fit.predict(start=len(y_train), end=len(y_train)+len(y_test)-1, dynamic=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import r2_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "ar_pred = np.array(predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean squared error: 0.33\n",
      "Variance score: -0.01\n"
     ]
    }
   ],
   "source": [
    "print(\"Mean squared error: %.2f\"\n",
    "      % mean_squared_error(y_test, ar_pred))\n",
    "# Explained variance score: 1 is perfect prediction\n",
    "print('Variance score: %.2f' % r2_score(y_test, ar_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib import pyplot\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "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>interval</th>\n",
       "      <th>demand</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2014-03-13 14:17:00</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2014-03-13 15:17:00</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2014-03-13 16:17:00</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2014-03-13 17:17:00</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2014-03-13 18:17:00</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             interval  demand\n",
       "0 2014-03-13 14:17:00       2\n",
       "1 2014-03-13 15:17:00       1\n",
       "2 2014-03-13 16:17:00       2\n",
       "3 2014-03-13 17:17:00       2\n",
       "4 2014-03-13 18:17:00       2"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test.reset_index().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = VAR(endog=train)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_fit = model.fit()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "prediction = model_fit.forecast(model_fit.y, steps=len(y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols=train.columns.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#converting predictions to dataframe\n",
    "pred = pd.DataFrame(index=range(0,len(prediction)),columns=[cols])\n",
    "for j in range(0,5):\n",
    "    for i in range(0, len(prediction)):\n",
    "       pred.iloc[i][j] = prediction[i][j]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "import math"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rmse value for 11385 is :  0.58582727236348\n"
     ]
    }
   ],
   "source": [
    "print('rmse value for', i, 'is : ', math.sqrt(mean_squared_error(pred['demand'], test['demand'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.tsa.api import ExponentialSmoothing, SimpleExpSmoothing, Holt\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Exponential Smoothing\n",
    "fit2 = SimpleExpSmoothing(np.asarray(train['demand'])).fit(smoothing_level=0.6,optimized=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred  = fit2.forecast(len(test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2.18554509, 2.18554509, 2.18554509, ..., 2.18554509, 2.18554509,\n",
       "       2.18554509])"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rmse value for pickup_day_of_week is :  0.6658102745096396\n"
     ]
    }
   ],
   "source": [
    "print('rmse value for', i, 'is : ', math.sqrt(mean_squared_error(pred, test['demand'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\user\\Anaconda\\lib\\site-packages\\mkl_fft\\_numpy_fft.py:1044: FutureWarning:\n",
      "\n",
      "Using a non-tuple sequence for multidimensional indexing is deprecated; use `arr[tuple(seq)]` instead of `arr[seq]`. In the future this will be interpreted as an array index, `arr[np.array(seq)]`, which will result either in an error or a different result.\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import statsmodels.api as sm\n",
    "sm.tsa.seasonal_decompose(train.demand).plot()\n",
    "result = sm.tsa.stattools.adfuller(train.demand)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Holt Linear Trend Method..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "fit1 = Holt(np.asarray(train['demand'])).fit(smoothing_level = 0.3,smoothing_slope = 0.1)\n",
    "pred= fit1.forecast(len(test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rmse value for pickup_day_of_week is :  250.5814058722232\n"
     ]
    }
   ],
   "source": [
    "print('rmse value for', i, 'is : ', math.sqrt(mean_squared_error(pred, test['demand'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "fit1 = ExponentialSmoothing(np.asarray(train['demand']) ,seasonal_periods=7 ,trend='add', seasonal='add',).fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred= fit1.forecast(len(test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rmse value for pickup_day_of_week is :  0.7952281782404124\n"
     ]
    }
   ],
   "source": [
    "print('rmse value for', i, 'is : ', math.sqrt(mean_squared_error(pred, test['demand'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "fit1 = sm.tsa.statespace.SARIMAX(train.demand, order=(2, 1, 4),seasonal_order=(0,1,1,7)).fit()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "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>demand</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>pickup_longitude</th>\n",
       "      <th>pickup_latitude</th>\n",
       "      <th>pickup_hour</th>\n",
       "      <th>pickup_day_of_week</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>interval</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014-03-13 14:17:00</th>\n",
       "      <td>2.100000</td>\n",
       "      <td>10.741500</td>\n",
       "      <td>-73.995000</td>\n",
       "      <td>40.752500</td>\n",
       "      <td>14.650000</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-03-13 15:17:00</th>\n",
       "      <td>1.250000</td>\n",
       "      <td>12.135000</td>\n",
       "      <td>-69.362500</td>\n",
       "      <td>38.209375</td>\n",
       "      <td>15.375000</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-03-13 16:17:00</th>\n",
       "      <td>2.619048</td>\n",
       "      <td>14.221905</td>\n",
       "      <td>-73.995238</td>\n",
       "      <td>40.758095</td>\n",
       "      <td>16.666667</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-03-13 17:17:00</th>\n",
       "      <td>2.652174</td>\n",
       "      <td>10.644783</td>\n",
       "      <td>-73.995652</td>\n",
       "      <td>40.747826</td>\n",
       "      <td>17.391304</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014-03-13 18:17:00</th>\n",
       "      <td>2.314286</td>\n",
       "      <td>13.266571</td>\n",
       "      <td>-69.762857</td>\n",
       "      <td>38.425429</td>\n",
       "      <td>18.600000</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       demand  fare_amount  pickup_longitude  pickup_latitude  \\\n",
       "interval                                                                        \n",
       "2014-03-13 14:17:00  2.100000    10.741500        -73.995000        40.752500   \n",
       "2014-03-13 15:17:00  1.250000    12.135000        -69.362500        38.209375   \n",
       "2014-03-13 16:17:00  2.619048    14.221905        -73.995238        40.758095   \n",
       "2014-03-13 17:17:00  2.652174    10.644783        -73.995652        40.747826   \n",
       "2014-03-13 18:17:00  2.314286    13.266571        -69.762857        38.425429   \n",
       "\n",
       "                     pickup_hour  pickup_day_of_week  \n",
       "interval                                              \n",
       "2014-03-13 14:17:00    14.650000                 3.0  \n",
       "2014-03-13 15:17:00    15.375000                 3.0  \n",
       "2014-03-13 16:17:00    16.666667                 3.0  \n",
       "2014-03-13 17:17:00    17.391304                 3.0  \n",
       "2014-03-13 18:17:00    18.600000                 3.0  "
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\user\\Anaconda\\lib\\site-packages\\statsmodels\\tsa\\statespace\\kalman_filter.py:1740: ValueWarning:\n",
      "\n",
      "Dynamic prediction specified to begin during out-of-sample forecasting period, and so has no effect.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "pred = fit1.predict(start=len(train)+1, end=len(train)+1+len(test)-1, dynamic=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rmse value for pickup_day_of_week is :  0.6024895928111502\n"
     ]
    }
   ],
   "source": [
    "print('rmse value for', i, 'is : ', math.sqrt(mean_squared_error(pred, test['demand'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}