Solution Part 1_1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#!conda install seaborn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "# import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### To-do\n",
    "&nbsp;&nbsp; 1. Pull in the data (This includes distance calculation) <br>\n",
    "&nbsp;&nbsp; 2. Run around with the categorical data and extract data from timestamped details <br>\n",
    "&nbsp;&nbsp; 3. Do basic exploratory analysis <br>\n",
    "&nbsp;&nbsp; 4. Create categorical data matrix and continuous data matrix <br>\n",
    "&nbsp;&nbsp; 5. Create tensors and embedding sizes for categorical data <br>\n",
    "&nbsp;&nbsp; 6. Create model, optimizer and criterion <br>\n",
    "&nbsp;&nbsp; 7. Build train model <br>\n",
    "&nbsp;&nbsp; 8. Run test <br>\n",
    "&nbsp;&nbsp; 9. Save model, load model <br>\n",
    "&nbsp;&nbsp;10. Create program to directly run data transformation and through model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1. Pull in the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df=pd.read_csv('..//Data//NYCTaxiFares.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>fare_class</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>2010-04-19 08:17:56 UTC</td>\n",
       "      <td>6.5</td>\n",
       "      <td>0</td>\n",
       "      <td>-73.992365</td>\n",
       "      <td>40.730521</td>\n",
       "      <td>-73.975499</td>\n",
       "      <td>40.744746</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2010-04-17 15:43:53 UTC</td>\n",
       "      <td>6.9</td>\n",
       "      <td>0</td>\n",
       "      <td>-73.990078</td>\n",
       "      <td>40.740558</td>\n",
       "      <td>-73.974232</td>\n",
       "      <td>40.744114</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2010-04-17 11:23:26 UTC</td>\n",
       "      <td>10.1</td>\n",
       "      <td>1</td>\n",
       "      <td>-73.994149</td>\n",
       "      <td>40.751118</td>\n",
       "      <td>-73.960064</td>\n",
       "      <td>40.766235</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2010-04-11 21:25:03 UTC</td>\n",
       "      <td>8.9</td>\n",
       "      <td>0</td>\n",
       "      <td>-73.990485</td>\n",
       "      <td>40.756422</td>\n",
       "      <td>-73.971205</td>\n",
       "      <td>40.748192</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2010-04-17 02:19:01 UTC</td>\n",
       "      <td>19.7</td>\n",
       "      <td>1</td>\n",
       "      <td>-73.990976</td>\n",
       "      <td>40.734202</td>\n",
       "      <td>-73.905956</td>\n",
       "      <td>40.743115</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           pickup_datetime  fare_amount  fare_class  pickup_longitude  \\\n",
       "0  2010-04-19 08:17:56 UTC          6.5           0        -73.992365   \n",
       "1  2010-04-17 15:43:53 UTC          6.9           0        -73.990078   \n",
       "2  2010-04-17 11:23:26 UTC         10.1           1        -73.994149   \n",
       "3  2010-04-11 21:25:03 UTC          8.9           0        -73.990485   \n",
       "4  2010-04-17 02:19:01 UTC         19.7           1        -73.990976   \n",
       "\n",
       "   pickup_latitude  dropoff_longitude  dropoff_latitude  passenger_count  \n",
       "0        40.730521         -73.975499         40.744746                1  \n",
       "1        40.740558         -73.974232         40.744114                1  \n",
       "2        40.751118         -73.960064         40.766235                2  \n",
       "3        40.756422         -73.971205         40.748192                1  \n",
       "4        40.734202         -73.905956         40.743115                1  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def haversine_distance(df, lat1, long1, lat2, long2):\n",
    "    \"\"\"\n",
    "    Calculates the haversine distance between 2 sets of GPS coordinates in df\n",
    "    \"\"\"\n",
    "    r = 6371  # average radius of Earth in kilometers\n",
    "       \n",
    "    phi1 = np.radians(df[lat1])\n",
    "    phi2 = np.radians(df[lat2])\n",
    "    \n",
    "    delta_phi = np.radians(df[lat2]-df[lat1])\n",
    "    delta_lambda = np.radians(df[long2]-df[long1])\n",
    "     \n",
    "    a = np.sin(delta_phi/2)**2 + np.cos(phi1) * np.cos(phi2) * np.sin(delta_lambda/2)**2\n",
    "    c = 2 * np.arctan2(np.sqrt(a), np.sqrt(1-a))\n",
    "    d = (r * c) # in kilometers\n",
    "\n",
    "    return d"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['distance_km']=haversine_distance(df,'pickup_latitude','pickup_longitude','dropoff_latitude','dropoff_longitude')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>fare_class</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>distance_km</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2010-04-19 08:17:56 UTC</td>\n",
       "      <td>6.5</td>\n",
       "      <td>0</td>\n",
       "      <td>-73.992365</td>\n",
       "      <td>40.730521</td>\n",
       "      <td>-73.975499</td>\n",
       "      <td>40.744746</td>\n",
       "      <td>1</td>\n",
       "      <td>2.126312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2010-04-17 15:43:53 UTC</td>\n",
       "      <td>6.9</td>\n",
       "      <td>0</td>\n",
       "      <td>-73.990078</td>\n",
       "      <td>40.740558</td>\n",
       "      <td>-73.974232</td>\n",
       "      <td>40.744114</td>\n",
       "      <td>1</td>\n",
       "      <td>1.392307</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2010-04-17 11:23:26 UTC</td>\n",
       "      <td>10.1</td>\n",
       "      <td>1</td>\n",
       "      <td>-73.994149</td>\n",
       "      <td>40.751118</td>\n",
       "      <td>-73.960064</td>\n",
       "      <td>40.766235</td>\n",
       "      <td>2</td>\n",
       "      <td>3.326763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2010-04-11 21:25:03 UTC</td>\n",
       "      <td>8.9</td>\n",
       "      <td>0</td>\n",
       "      <td>-73.990485</td>\n",
       "      <td>40.756422</td>\n",
       "      <td>-73.971205</td>\n",
       "      <td>40.748192</td>\n",
       "      <td>1</td>\n",
       "      <td>1.864129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2010-04-17 02:19:01 UTC</td>\n",
       "      <td>19.7</td>\n",
       "      <td>1</td>\n",
       "      <td>-73.990976</td>\n",
       "      <td>40.734202</td>\n",
       "      <td>-73.905956</td>\n",
       "      <td>40.743115</td>\n",
       "      <td>1</td>\n",
       "      <td>7.231321</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           pickup_datetime  fare_amount  fare_class  pickup_longitude  \\\n",
       "0  2010-04-19 08:17:56 UTC          6.5           0        -73.992365   \n",
       "1  2010-04-17 15:43:53 UTC          6.9           0        -73.990078   \n",
       "2  2010-04-17 11:23:26 UTC         10.1           1        -73.994149   \n",
       "3  2010-04-11 21:25:03 UTC          8.9           0        -73.990485   \n",
       "4  2010-04-17 02:19:01 UTC         19.7           1        -73.990976   \n",
       "\n",
       "   pickup_latitude  dropoff_longitude  dropoff_latitude  passenger_count  \\\n",
       "0        40.730521         -73.975499         40.744746                1   \n",
       "1        40.740558         -73.974232         40.744114                1   \n",
       "2        40.751118         -73.960064         40.766235                2   \n",
       "3        40.756422         -73.971205         40.748192                1   \n",
       "4        40.734202         -73.905956         40.743115                1   \n",
       "\n",
       "   distance_km  \n",
       "0     2.126312  \n",
       "1     1.392307  \n",
       "2     3.326763  \n",
       "3     1.864129  \n",
       "4     7.231321  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2. Run around with the categorical data and extract data from timestamped details"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['pickup_datetime']=pd.to_datetime(df['pickup_datetime'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0   2010-04-19 08:17:56+00:00\n",
       "1   2010-04-17 15:43:53+00:00\n",
       "2   2010-04-17 11:23:26+00:00\n",
       "3   2010-04-11 21:25:03+00:00\n",
       "4   2010-04-17 02:19:01+00:00\n",
       "Name: pickup_datetime, dtype: datetime64[ns, UTC]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['pickup_datetime'][:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#converting EST to NewYork time since all the data is taken from NYC\n",
    "df['date_timeEST']=df['pickup_datetime']-pd.Timedelta(hours=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Creating various columns that are extracted data from the timestamp\n",
    "df['hours']=df['date_timeEST'].dt.hour\n",
    "df['weekday']=df['date_timeEST'].dt.strftime('%a')\n",
    "df['AM_PM']=np.where(df['hours']<12,\"AM\",\"PM\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pickup_datetime</th>\n",
       "      <th>fare_amount</th>\n",
       "      <th>fare_class</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>distance_km</th>\n",
       "      <th>date_timeEST</th>\n",
       "      <th>hours</th>\n",
       "      <th>weekday</th>\n",
       "      <th>AM_PM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2010-04-19 08:17:56+00:00</td>\n",
       "      <td>6.5</td>\n",
       "      <td>0</td>\n",
       "      <td>-73.992365</td>\n",
       "      <td>40.730521</td>\n",
       "      <td>-73.975499</td>\n",
       "      <td>40.744746</td>\n",
       "      <td>1</td>\n",
       "      <td>2.126312</td>\n",
       "      <td>2010-04-19 04:17:56+00:00</td>\n",
       "      <td>4</td>\n",
       "      <td>Mon</td>\n",
       "      <td>AM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2010-04-17 15:43:53+00:00</td>\n",
       "      <td>6.9</td>\n",
       "      <td>0</td>\n",
       "      <td>-73.990078</td>\n",
       "      <td>40.740558</td>\n",
       "      <td>-73.974232</td>\n",
       "      <td>40.744114</td>\n",
       "      <td>1</td>\n",
       "      <td>1.392307</td>\n",
       "      <td>2010-04-17 11:43:53+00:00</td>\n",
       "      <td>11</td>\n",
       "      <td>Sat</td>\n",
       "      <td>AM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2010-04-17 11:23:26+00:00</td>\n",
       "      <td>10.1</td>\n",
       "      <td>1</td>\n",
       "      <td>-73.994149</td>\n",
       "      <td>40.751118</td>\n",
       "      <td>-73.960064</td>\n",
       "      <td>40.766235</td>\n",
       "      <td>2</td>\n",
       "      <td>3.326763</td>\n",
       "      <td>2010-04-17 07:23:26+00:00</td>\n",
       "      <td>7</td>\n",
       "      <td>Sat</td>\n",
       "      <td>AM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2010-04-11 21:25:03+00:00</td>\n",
       "      <td>8.9</td>\n",
       "      <td>0</td>\n",
       "      <td>-73.990485</td>\n",
       "      <td>40.756422</td>\n",
       "      <td>-73.971205</td>\n",
       "      <td>40.748192</td>\n",
       "      <td>1</td>\n",
       "      <td>1.864129</td>\n",
       "      <td>2010-04-11 17:25:03+00:00</td>\n",
       "      <td>17</td>\n",
       "      <td>Sun</td>\n",
       "      <td>PM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2010-04-17 02:19:01+00:00</td>\n",
       "      <td>19.7</td>\n",
       "      <td>1</td>\n",
       "      <td>-73.990976</td>\n",
       "      <td>40.734202</td>\n",
       "      <td>-73.905956</td>\n",
       "      <td>40.743115</td>\n",
       "      <td>1</td>\n",
       "      <td>7.231321</td>\n",
       "      <td>2010-04-16 22:19:01+00:00</td>\n",
       "      <td>22</td>\n",
       "      <td>Fri</td>\n",
       "      <td>PM</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            pickup_datetime  fare_amount  fare_class  pickup_longitude  \\\n",
       "0 2010-04-19 08:17:56+00:00          6.5           0        -73.992365   \n",
       "1 2010-04-17 15:43:53+00:00          6.9           0        -73.990078   \n",
       "2 2010-04-17 11:23:26+00:00         10.1           1        -73.994149   \n",
       "3 2010-04-11 21:25:03+00:00          8.9           0        -73.990485   \n",
       "4 2010-04-17 02:19:01+00:00         19.7           1        -73.990976   \n",
       "\n",
       "   pickup_latitude  dropoff_longitude  dropoff_latitude  passenger_count  \\\n",
       "0        40.730521         -73.975499         40.744746                1   \n",
       "1        40.740558         -73.974232         40.744114                1   \n",
       "2        40.751118         -73.960064         40.766235                2   \n",
       "3        40.756422         -73.971205         40.748192                1   \n",
       "4        40.734202         -73.905956         40.743115                1   \n",
       "\n",
       "   distance_km              date_timeEST  hours weekday AM_PM  \n",
       "0     2.126312 2010-04-19 04:17:56+00:00      4     Mon    AM  \n",
       "1     1.392307 2010-04-17 11:43:53+00:00     11     Sat    AM  \n",
       "2     3.326763 2010-04-17 07:23:26+00:00      7     Sat    AM  \n",
       "3     1.864129 2010-04-11 17:25:03+00:00     17     Sun    PM  \n",
       "4     7.231321 2010-04-16 22:19:01+00:00     22     Fri    PM  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3. Data Exploration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cont=df.copy()\n",
    "df_cont.drop(['pickup_datetime','pickup_longitude','pickup_latitude','dropoff_longitude','dropoff_latitude','fare_class','hours','weekday','AM_PM'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>fare_amount</th>\n",
       "      <th>passenger_count</th>\n",
       "      <th>distance_km</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>fare_amount</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.034072</td>\n",
       "      <td>0.896157</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>passenger_count</th>\n",
       "      <td>0.034072</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.023100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>distance_km</th>\n",
       "      <td>0.896157</td>\n",
       "      <td>0.023100</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 fare_amount  passenger_count  distance_km\n",
       "fare_amount         1.000000         0.034072     0.896157\n",
       "passenger_count     0.034072         1.000000     0.023100\n",
       "distance_km         0.896157         0.023100     1.000000"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#correlation of all significant continuous variables\n",
    "df_cont.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEWCAYAAABMoxE0AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAYRElEQVR4nO3df7DddX3n8efLQBH5/SNQSChhl2xXYBWXLDKD9RdWUlFhdqCNjiU7m5qWpSNOdW2ws/6axeJ0C4VpwTLiEFGBSIvEIsVMkFW2LHhRNAZkTSVCDCUREOIPqMH3/nE+dzy5nNzc+M25l5v7fMycOd/v+3w/3+/nc2ZyX/l8v+d8T6oKSZJ+VS+a6g5IkqY3g0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEs1YSX4zyTeSbEnyrqnujzRdGSSayd4H3FFV+1XV5VPdmReKJK9NsmGq+6HpwyDRTHY0sHZnGyXZYwh9kaYtg0QzUpLbgdcBf53kx0kuaKe5nk7ySJIP9W07L0klWZLkYeD2Vv+vSR5I8mSS25IcPYHjXtb2/3SSe5P8Vt9rH0ryuSSfbqfb1iT5d0kuTLKptXtj3/ZHJlmZ5Ikk65K8s++1a5L8z771bWYZSdYneW+SbyV5KskNSV6cZB/gVuDI9r78OMmRv+r7rJnBINGMVFWvB74K/HFV7Qt8EzgXOBA4AzgvyVljmr0GeClwenvt/cB/Bma3fV03gUN/DTgROBj4LPC5JC/ue/0twLXAQcA3gNvo/TudA3wE+Nu+ba8DNgBHAmcDH01y2gT6MOp3gYXAMcDLgP9SVT8BfgfYWFX7tsfGndinZiCDRAKq6o6qWlNVv6iqb9H7I/2aMZt9qKp+UlU/A/4Q+POqeqCqtgIfBU7c0aykqj5dVY9X1daq+ktgL+A3+zb5alXd1vb5OXohdXFV/Ry4HpiX5MAkRwGvAv60qp6pqvuATwC/vxPDvryqNlbVE8AX6AWctNMMEglI8sokX06yOclTwB8Bh47Z7JG+5aOBy5L8KMmPgCeA0Js5jHec97TTYU+1dgeMOc5jfcs/A35YVc/1rQPsS28W8kRVbenb/vs7Ov4Y/9K3/NO2X2mnGSRSz2eBlcBRVXUA8HF6wdCv/zcXHgH+sKoO7HvsXVX/tL0DtOshf0rvlNJBVXUg8NSA40zERuDgJPv11X4D+EFb/gnwkr7Xfn0n9u1vS2inGCRSz370/of/TJKTgbfvYPuPAxcmOR4gyQFJzpnAMbYCm4E9knwA2P9X6WxVPQL8E/Dn7SL5y4AlwGfaJvcBb0pycJJfB969E7t/DDgkyQG/St808xgkUs9/Az6SZAvwAWDFeBtX1U3Ax4DrkzwNfJveRerx3EbvE1H/j95pqGfY9nTZznobMI/e7OQm4INVtaq9di29DxCsB74E3DDRnVbVd+hdI/peO3Xnp7Y0rvgLiZKkLpyRSJI68Ru60i7ULqjfOui19n0VabfjqS1JUiczbkZy6KGH1rx586a6G5I0rdx7770/rKrZg16bcUEyb948RkZGprobkjStJPn+9l7zYrskqRODRJLUiUEiSerEIJEkdWKQSJI6MUgkSZ0YJJKkTgwSSVInBokkqZMZ9832LuYtu2XKjr3+4jOm7NiSNB5nJJKkTgwSSVInQw2SJOuTrElyX5KRVjs4yaok323PB/Vtf2GSdUkeTHJ6X/2ktp91SS5PklbfK8kNrX53knnDHI8k6fkmY0byuqo6saoWtPVlwOqqmg+sbuskOQ5YBBwPLASuSDKrtbkSWArMb4+Frb4EeLKqjgUupfcb2pKkSTQVp7bOBJa35eXAWX3166vq2ap6CFgHnJzkCGD/qrqrer/C9akxbUb3dSNw2uhsRZI0OYYdJAV8Kcm9SZa22uFV9ShAez6s1ecAj/S13dBqc9ry2Po2bapqK/AUcMjYTiRZmmQkycjmzZt3ycAkST3D/vjvqVW1MclhwKok3xln20EziRqnPl6bbQtVVwFXASxYsMDfFpakXWioM5Kq2tieNwE3AScDj7XTVbTnTW3zDcBRfc3nAhtbfe6A+jZtkuwBHAA8MYyxSJIGG1qQJNknyX6jy8AbgW8DK4HFbbPFwM1teSWwqH0S6xh6F9Xvaae/tiQ5pV3/OHdMm9F9nQ3c3q6jSJImyTBPbR0O3NSufe8BfLaq/jHJ14AVSZYADwPnAFTV2iQrgPuBrcD5VfVc29d5wDXA3sCt7QFwNXBtknX0ZiKLhjgeSdIAQwuSqvoe8PIB9ceB07bT5iLgogH1EeCEAfVnaEEkSZoafrNdktSJQSJJ6sQgkSR1YpBIkjoxSCRJnRgkkqRODBJJUicGiSSpE4NEktSJQSJJ6sQgkSR1YpBIkjoxSCRJnRgkkqRODBJJUicGiSSpE4NEktSJQSJJ6sQgkSR1YpBIkjoxSCRJnRgkkqRODBJJUicGiSSpE4NEktSJQSJJ6sQgkSR1YpBIkjoxSCRJnRgkkqRODBJJUid7DPsASWYBI8APqurNSQ4GbgDmAeuB362qJ9u2FwJLgOeAd1XVba1+EnANsDfwReCCqqokewGfAk4CHgd+r6rWD3tMU2Heslum5LjrLz5jSo4rafqYjBnJBcADfevLgNVVNR9Y3dZJchywCDgeWAhc0UII4EpgKTC/PRa2+hLgyao6FrgU+NhwhyJJGmuoQZJkLnAG8Im+8pnA8ra8HDirr359VT1bVQ8B64CTkxwB7F9Vd1VV0ZuBnDVgXzcCpyXJkIYjSRpg2DOSvwLeB/yir3Z4VT0K0J4Pa/U5wCN9221otTlteWx9mzZVtRV4Cjhkl45AkjSuoQVJkjcDm6rq3ok2GVCrcerjtRnbl6VJRpKMbN68eYLdkSRNxDBnJKcCb02yHrgeeH2STwOPtdNVtOdNbfsNwFF97ecCG1t97oD6Nm2S7AEcADwxtiNVdVVVLaiqBbNnz941o5MkAUMMkqq6sKrmVtU8ehfRb6+qdwArgcVts8XAzW15JbAoyV5JjqF3Uf2edvprS5JT2vWPc8e0Gd3X2e0Yz5uRSJKGZ+gf/x3gYmBFkiXAw8A5AFW1NskK4H5gK3B+VT3X2pzHLz/+e2t7AFwNXJtkHb2ZyKLJGoQkqWdSgqSq7gDuaMuPA6dtZ7uLgIsG1EeAEwbUn6EFkSRpavjNdklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHVikEiSOjFIJEmdDC1Ikrw4yT1JvplkbZIPt/rBSVYl+W57PqivzYVJ1iV5MMnpffWTkqxpr12eJK2+V5IbWv3uJPOGNR5J0mDDnJE8C7y+ql4OnAgsTHIKsAxYXVXzgdVtnSTHAYuA44GFwBVJZrV9XQksBea3x8JWXwI8WVXHApcCHxvieCRJA+wwSJKcmmSftvyOJJckOXpH7arnx211z/Yo4ExgeasvB85qy2cC11fVs1X1ELAOODnJEcD+VXVXVRXwqTFtRvd1I3Da6GxFkjQ5JjIjuRL4aZKXA+8Dvk/vj/kOJZmV5D5gE7Cqqu4GDq+qRwHa82Ft8znAI33NN7TanLY8tr5Nm6raCjwFHDKRvkmSdo2JBMnWNhM4E7isqi4D9pvIzqvquao6EZhLb3ZxwjibD5pJ1Dj18dpsu+NkaZKRJCObN2/eQa8lSTtjIkGyJcmFwDuAW9p1iz135iBV9SPgDnrXNh5rp6toz5vaZhuAo/qazQU2tvrcAfVt2iTZAzgAeGLA8a+qqgVVtWD27Nk703VJ0g5MJEh+j96F8yVV9S/0Tif9xY4aJZmd5MC2vDfwBuA7wEpgcdtsMXBzW14JLGqfxDqG3kX1e9rpry1JTmnXP84d02Z0X2cDt7fZkyRpkuyxow1aeFzSt/4wE7tGcgSwvM1gXgSsqKp/SHIXsCLJEuBh4Jy237VJVgD3A1uB86vqubav84BrgL2BW9sD4Grg2iTr6M1EFk2gX5KkXWi7QZLkzqp6VZItbHvdIfQ+lLX/eDuuqm8BrxhQfxw4bTttLgIuGlAfAZ53faWqnqEFkSRpamw3SKrqVe15QhfWJUkz00S+R/KGAbXFg7aVJM08E7nY/oEkVybZJ8nhSb4AvGXYHZMkTQ8TCZLXAP8M3AfcCXy2qs4eZqckSdPHRILkIOCV9MLkWeBob0MiSRo1kSD5v8CtVbUQ+E/AkcD/GWqvJEnTxg6/RwK8oX13hKr6GfCuJK8ebrckSdPFRL6Q+PDocpJ/C7yN3hf/xrtvliRphpjIx3+PSPLuJPcAa4FZ9MJEkqTtB0mSdya5HfjfwKHAHwCPVtWHq2rNZHVQkvTCNt6prb8B7gLe3m5RQhJviChJ2sZ4QXIkvftYXZLkcGAFO3n7eEnS7m+7p7aq6odVdWVVvZreTRafAjYleSDJRyeth5KkF7SJfI+EqtpQVf+rqk6i93vpzw61V5KkaWNCQTIqyder6sGq+vCwOiRJml7G+9TWF5PMG1sebnckSdPNeDOSa4AvJfmzJKMX2W8ZfpckSdPJeBfbV9D7hcP9gZEk7wWeSPInSf5ksjooSXph29EtUn4O/ATYC9gP+MXQeyRJmlbG+832hcAlwErgP1bVTyetV5KkaWO8GcmfAedU1drJ6owkafrZbpBU1W9NZkckSdPTTn2PRJKksQwSSVInBokkqRODRJLUiUEiSerEIJEkdWKQSJI6MUgkSZ0YJJKkTgwSSVInQwuSJEcl+XL7jfe1SS5o9YOTrEry3fZ8UF+bC5OsS/JgktP76iclWdNeuzxJWn2vJDe0+t0DfohLkjRkw5yRbAXeU1UvBU4Bzk9yHLAMWF1V84HVbZ322iLgeGAhcEWSWW1fVwJLgfntsbDVlwBPVtWxwKXAx4Y4HknSAEMLkqp6tKq+3pa3AA8Ac4AzgeVts+XAWW35TOD6qnq2qh4C1gEnJzkC2L+q7qqqAj41ps3ovm4EThudrUiSJsekXCNpp5xeAdwNHF5Vj0IvbIDD2mZzgEf6mm1otTlteWx9mzZVtRV4CjhkwPGXJhlJMrJ58+ZdNCpJEkxCkCTZF/g74N1V9fR4mw6o1Tj18dpsW6i6qqoWVNWC2bNn76jLkqSdMNQgSbInvRD5TFX9fSs/1k5X0Z43tfoG4Ki+5nOBja0+d0B9mzZJ9gAOAJ7Y9SORJG3PMD+1FeBq4IGquqTvpZXA4ra8GLi5r76ofRLrGHoX1e9pp7+2JDml7fPcMW1G93U2cHu7jiJJmiTj/dRuV6cCvw+sSXJfq70fuBhYkWQJ8DBwDkBVrU2yArif3ie+zq+q51q784BrgL2BW9sDekF1bZJ19GYii4Y4HknSAEMLkqq6k8HXMABO206bi4CLBtRHgBMG1J+hBZEkaWr4zXZJUicGiSSpE4NEktSJQSJJ6sQgkSR1YpBIkjoxSCRJnRgkkqRODBJJUicGiSSpE4NEktSJQSJJ6sQgkSR1YpBIkjoxSCRJnRgkkqRODBJJUifD/Kld7QbmLbtlyo69/uIzpuzYkibOGYkkqRODRJLUiUEiSerEIJEkdWKQSJI6MUgkSZ0YJJKkTgwSSVInBokkqRODRJLUiUEiSerEIJEkdWKQSJI6GVqQJPlkkk1Jvt1XOzjJqiTfbc8H9b12YZJ1SR5Mcnpf/aQka9prlydJq++V5IZWvzvJvGGNRZK0fcOckVwDLBxTWwasrqr5wOq2TpLjgEXA8a3NFUlmtTZXAkuB+e0xus8lwJNVdSxwKfCxoY1EkrRdQwuSqvoK8MSY8pnA8ra8HDirr359VT1bVQ8B64CTkxwB7F9Vd1VVAZ8a02Z0XzcCp43OViRJk2eyr5EcXlWPArTnw1p9DvBI33YbWm1OWx5b36ZNVW0FngIOGXTQJEuTjCQZ2bx58y4aiiQJXjgX2wfNJGqc+nhtnl+suqqqFlTVgtmzZ/+KXZQkDTLZQfJYO11Fe97U6huAo/q2mwtsbPW5A+rbtEmyB3AAzz+VJkkasskOkpXA4ra8GLi5r76ofRLrGHoX1e9pp7+2JDmlXf84d0yb0X2dDdzerqNIkibRHsPacZLrgNcChybZAHwQuBhYkWQJ8DBwDkBVrU2yArgf2AqcX1XPtV2dR+8TYHsDt7YHwNXAtUnW0ZuJLBrWWCRJ2ze0IKmqt23npdO2s/1FwEUD6iPACQPqz9CCSJI0dV4oF9slSdOUQSJJ6sQgkSR1YpBIkjoxSCRJnRgkkqRODBJJUicGiSSpE4NEktSJQSJJ6sQgkSR1YpBIkjoxSCRJnQzt7r9SV/OW3TIlx11/8RlTclxpunJGIknqxCCRJHVikEiSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0SS1IlBIknqxCCRJHXiLVKkMbw1i7RzDBLpBWKqAgwMMXXjqS1JUicGiSSpE4NEktSJ10gk+QGDSbQ7XgtzRiJJ6mTaz0iSLAQuA2YBn6iqi6e4S5KmgamcGexupnWQJJkF/A3w28AG4GtJVlbV/VPbM0kT4R/z3cN0P7V1MrCuqr5XVf8KXA+cOcV9kqQZZVrPSIA5wCN96xuAV47dKMlSYGlb/XGSByehby9EhwI/nOpOTLGZ/h44/hk8/nys0/iP3t4L0z1IMqBWzytUXQVcNfzuvLAlGamqBVPdj6k0098Dx+/4hzH+6X5qawNwVN/6XGDjFPVFkmak6R4kXwPmJzkmya8Bi4CVU9wnSZpRpvWpraramuSPgdvoffz3k1W1doq79UI240/v4Xvg+Ge2oYw/Vc+7pCBJ0oRN91NbkqQpZpBIkjoxSHZTST6ZZFOSb/fVDk6yKsl32/NBU9nHYUpyVJIvJ3kgydokF7T6jHgPkrw4yT1JvtnG/+FWnxHjH5VkVpJvJPmHtj7Txr8+yZok9yUZabVd/h4YJLuva4CFY2rLgNVVNR9Y3dZ3V1uB91TVS4FTgPOTHMfMeQ+eBV5fVS8HTgQWJjmFmTP+URcAD/Stz7TxA7yuqk7s+/7ILn8PDJLdVFV9BXhiTPlMYHlbXg6cNZl9mkxV9WhVfb0tb6H3x2QOM+Q9qJ4ft9U926OYIeMHSDIXOAP4RF95xox/HLv8PTBIZpbDq+pR6P2hBQ6b4v5MiiTzgFcAdzOD3oN2Wuc+YBOwqqpm1PiBvwLeB/yirzaTxg+9/zx8Kcm97VZRMIT3YFp/j0TakST7An8HvLuqnk4G3VVn91RVzwEnJjkQuCnJCVPcpUmT5M3Apqq6N8lrp7g7U+nUqtqY5DBgVZLvDOMgzkhmlseSHAHQnjdNcX+GKsme9ELkM1X19608o94DgKr6EXAHvWtmM2X8pwJvTbKe3l3BX5/k08yc8QNQVRvb8ybgJnp3TN/l74FBMrOsBBa35cXAzVPYl6FKb+pxNfBAVV3S99KMeA+SzG4zEZLsDbwB+A4zZPxVdWFVza2qefRunXR7Vb2DGTJ+gCT7JNlvdBl4I/BthvAe+M323VSS64DX0rtt9mPAB4HPAyuA3wAeBs6pqrEX5HcLSV4FfBVYwy/Pkb+f3nWS3f49SPIyehdSZ9H7D+OKqvpIkkOYAePv105tvbeq3jyTxp/k39CbhUDvMsZnq+qiYbwHBokkqRNPbUmSOjFIJEmdGCSSpE4MEklSJwaJJKkTg0Tahdpdhx9KcnBbP6itHz1g28uS/CDJDv8dJvmjJOcOo89SV378V9rFkrwPOLaqlib5W2B9Vf35mG1eBKwHNgLLquqOSe+otIs4I5F2vUuBU5K8G3gV8JcDtnkdvW8ZXwm8bbSY5PIkH2jLpyf5SpIXJflQkve2+ruS3J/kW0muH/ZgpB3xpo3SLlZVP0/y34F/BN5YVf86YLO3AdfRuz3FR5PsWVU/p/fbEF9L8lXgcuBNVfWLMTebXAYcU1XPjt4GRZpKzkik4fgd4FHgeXfcTfJrwJuAz1fV0/Ru2/JGgKr6KfBOYBXw11X1zwP2/S3gM0neQe8HvKQp5YxE2sWSnAj8Nr1fZrwzyY3AF9rLH6d3XeQAYE2babwE+ClwS9vmPwCPA0du5xBnAK8G3gr8jyTHV5WBoiljkEi7ULvr8JX0fv/k4SR/AVxcVSf2bXMd8AdVdV1b3wd4KMlLgNnAe+j9ENcXk3y+/SDVaNsXAUdV1ZeT3Am8HdgX+NGkDFAawFNb0q71TuDhqlrV1q8A/n2S1wC0sDidX84+qKqfAHcCb6F36/v3tt+RWAJ8IsmL+/Y/C/h0kjXAN4BL2++NSFPGj/9KkjpxRiJJ6sQgkSR1YpBIkjoxSCRJnRgkkqRODBJJUicGiSSpk/8PnBJ+15Vo6rkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "#Distribution of data of all continuous variables\n",
    "for cont in df_cont.columns:\n",
    "    plt.hist(df[cont])\n",
    "    plt.xlabel('X-Axis')\n",
    "    plt.ylabel('Y-Axis')\n",
    "    plt.title(cont)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cat=df.copy()\n",
    "df_cat.drop(['fare_amount','passenger_count','distance_km'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "#Distribution of data of all continuous variables\n",
    "for cat in df_cat.columns:\n",
    "    plt.hist(df[cat])\n",
    "    plt.xlabel('X-Axis')\n",
    "    plt.ylabel('Y-Axis')\n",
    "    plt.title(cat)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4. Create categorical data matrix and continuous data matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['pickup_datetime', 'fare_amount', 'fare_class', 'pickup_longitude',\n",
       "       'pickup_latitude', 'dropoff_longitude', 'dropoff_latitude',\n",
       "       'passenger_count', 'distance_km', 'date_timeEST', 'hours', 'weekday',\n",
       "       'AM_PM'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Creating an index of all categorical columns in index\n",
    "cats=['hours','weekday','AM_PM']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "#To use [category dtype].cat.codes, convert all non continuous objects to 'category' codes\n",
    "for cat in cats:\n",
    "    df[cat]=df[cat].astype('category')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pickup_datetime      datetime64[ns, UTC]\n",
       "fare_amount                      float64\n",
       "fare_class                         int64\n",
       "pickup_longitude                 float64\n",
       "pickup_latitude                  float64\n",
       "dropoff_longitude                float64\n",
       "dropoff_latitude                 float64\n",
       "passenger_count                    int64\n",
       "distance_km                      float64\n",
       "date_timeEST         datetime64[ns, UTC]\n",
       "hours                           category\n",
       "weekday                         category\n",
       "AM_PM                           category\n",
       "dtype: object"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Creating a matrix of all categorical columns of all int index values for each category \n",
    "#e.g. weekdays will have 7 codes (0-6), hours will have 24 codes (0-23)\n",
    "cat_cols=np.stack([df[cat].cat.codes for cat in cats],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "#creating index of continuous columns\n",
    "conts=['pickup_longitude',\n",
    "       'pickup_latitude', 'dropoff_longitude', 'dropoff_latitude',\n",
    "       'passenger_count', 'distance_km']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "cont_cols=np.stack([df[col].values for col in conts],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "label=df['fare_amount'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "category shape (120000, 3)\n",
      "continuous shape (120000, 6)\n"
     ]
    }
   ],
   "source": [
    "print('category shape',cat_cols.shape)\n",
    "print('continuous shape',cont_cols.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 5.Create tensors and embedding sizes for categorical data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "category shape torch.Size([120000, 3])\n",
      "continuous shape torch.Size([120000, 6])\n"
     ]
    }
   ],
   "source": [
    "cat_tensor=torch.tensor(cat_cols,dtype=torch.long)\n",
    "cont_tensor=torch.tensor(cont_cols,dtype=torch.float)\n",
    "label_tensor=torch.tensor(label,dtype=torch.float)\n",
    "print('category shape',cat_tensor.shape)\n",
    "print('continuous shape',cont_tensor.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### 5a. Creating the embedding object\n",
    "&nbsp;&nbsp; 1.  Creating embedding sizes : \n",
    "We need to create an embedding matrix of dimensions [original no.of cats, new numbers of cats] This is because we use only one-hotkey encoding, then the dimensions of each vector will by original no. of unique values of the categorical column which will be very large. Hence, we want to squeeze this to a more reasonable number of dimensions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "cat_len=[df[col].nunique() for col in cats]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[24, 7, 2]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cat_len"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "embed_size=[(cat,min(50,int((cat+1)//2))) for cat in cat_len]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(24, 12), (7, 4), (2, 1)]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "embed_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "cat_embeddings = [nn.Embedding(initial,target) for initial,target in embed_size]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### 5b. Trying to create an embedding matrix for the sample to see how it goes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "cat_sample=torch.tensor(cat_cols[:4,:],dtype=torch.long)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[ 4,  1,  0],\n",
      "        [11,  2,  0],\n",
      "        [ 7,  2,  0],\n",
      "        [17,  3,  1]])\n"
     ]
    }
   ],
   "source": [
    "print(cat_sample)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Embedding(24, 12), Embedding(7, 4), Embedding(2, 1)]"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cat_embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Embedding(24, 12)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Attempting only for the hours data - to convert a coded 24-dimension feature into an embedding\n",
    "embed_hour=cat_embeddings[0]\n",
    "embed_hour"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-37-581f525cfd62>:2: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
      "  test_hour=torch.tensor(cat_sample[:,0],dtype=torch.long)\n"
     ]
    }
   ],
   "source": [
    "#test sample of hours data\n",
    "test_hour=torch.tensor(cat_sample[:,0],dtype=torch.long)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([ 4, 11,  7, 17])"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_hour"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[-0.8622, -0.4589, -0.3882,  2.1875,  0.7319, -0.6826,  0.9976,  0.3093,\n",
      "         -0.5769,  0.3855,  2.0841, -0.6578],\n",
      "        [ 0.1719, -1.4984,  1.5151,  0.1399,  1.6173,  0.2143, -2.1506,  0.9505,\n",
      "         -1.2036,  1.2818, -0.9280, -0.1563],\n",
      "        [-1.4625,  0.1978,  0.4445, -1.5443,  0.1917,  0.5728,  1.5959,  0.5367,\n",
      "          1.0606, -0.8834,  0.3144,  0.3381],\n",
      "        [-0.5294,  1.0591, -0.5086, -0.3494, -0.4488, -0.6346, -0.9714, -0.6094,\n",
      "          1.7215,  0.1053,  1.5604, -0.4385]], grad_fn=<EmbeddingBackward>)\n"
     ]
    }
   ],
   "source": [
    "print(embed_hour(test_hour))\n",
    "#We see that a 24 dimension variable now has only 12, which is a victory!!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Now attempting the same for all features in cat_sample\n",
    "embeddings_test=[]\n",
    "for i, embed in enumerate(cat_embeddings):\n",
    "    embeddings_test.append(embed(cat_sample[:,i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "hours\n",
      "torch.Size([4, 12])\n",
      "tensor([[-0.8622, -0.4589, -0.3882,  2.1875,  0.7319, -0.6826,  0.9976,  0.3093,\n",
      "         -0.5769,  0.3855,  2.0841, -0.6578],\n",
      "        [ 0.1719, -1.4984,  1.5151,  0.1399,  1.6173,  0.2143, -2.1506,  0.9505,\n",
      "         -1.2036,  1.2818, -0.9280, -0.1563],\n",
      "        [-1.4625,  0.1978,  0.4445, -1.5443,  0.1917,  0.5728,  1.5959,  0.5367,\n",
      "          1.0606, -0.8834,  0.3144,  0.3381],\n",
      "        [-0.5294,  1.0591, -0.5086, -0.3494, -0.4488, -0.6346, -0.9714, -0.6094,\n",
      "          1.7215,  0.1053,  1.5604, -0.4385]], grad_fn=<EmbeddingBackward>)\n",
      "weekday\n",
      "torch.Size([4, 4])\n",
      "tensor([[-0.1605, -0.9461,  0.6792,  0.6212],\n",
      "        [-0.7289,  0.9246, -0.9244,  0.9909],\n",
      "        [-0.7289,  0.9246, -0.9244,  0.9909],\n",
      "        [ 0.3013,  0.8733, -1.1197, -1.1564]], grad_fn=<EmbeddingBackward>)\n",
      "AM_PM\n",
      "torch.Size([4, 1])\n",
      "tensor([[ 0.9581],\n",
      "        [ 0.9581],\n",
      "        [ 0.9581],\n",
      "        [-1.6076]], grad_fn=<EmbeddingBackward>)\n"
     ]
    }
   ],
   "source": [
    "for i,e in enumerate(embeddings_test):\n",
    "    print(cats[i])\n",
    "    print(e.shape)\n",
    "    print(e)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###### SUCCESS ON SAMPLE!!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 6.Create model, optimizer and criterion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([120000, 6])"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cont_tensor.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Model(nn.Module):\n",
    "    def __init__(self,cat_cols,cont_cols,layer_count,output_features,embed_size_list,p=0.5): \n",
    "        '''\n",
    "        args:\n",
    "        cat_cols is the tensor of all categorical values (pre-embedding)\n",
    "        n_cont=number of continuous variables (for batch normalization)\n",
    "        cont_cols is the tensor of all continuous values\n",
    "        input_features - number of parameters of input\n",
    "        layer_count - a tuple of number of nodes of each hidden layer\n",
    "        output_features = number of outputs expected\n",
    "        embed_size_list is list of embedding sizes for the categorical values\n",
    "        p = basically, the % of nodes to be nullified during dropout layer\n",
    "       \n",
    "       Approach: In the constructor, create all the layers (Linear, ReLU, Batch and Dropout) for each hidden layer as per layer_count)\n",
    "       and add them to sequential(). Fwd() will have all the data manipulation and final embedding\n",
    "       \n",
    "        '''\n",
    "        super().__init__()\n",
    "        #self.input_features=input_features\n",
    "        self.output_features=output_features\n",
    "        self.layer_count=layer_count\n",
    "        self.embed_size_list=embed_size_list\n",
    "        #Create embeddings from categorical columns as seen in the test case\n",
    "        self.embeddings=[nn.Embedding(base_dim, target_dim) for base_dim, target_dim in embed_size_list]\n",
    "#         print(self.embeddings)\n",
    "        n_cont=cont_cols.shape[1]\n",
    "        self.batch_norm=nn.BatchNorm1d(n_cont)\n",
    "        self.dropout=nn.Dropout(p=0.5)\n",
    "        #To create the layers we need to start with input sizes\n",
    "        n_cont=cont_cols.shape[1]\n",
    "        #We need to find the total number of columns in the data. \n",
    "        #cat_cols currently only has 4 columns which post embedding will go to 23 \n",
    "        n_in=sum(nf for ni,nf in self.embed_size_list)+n_cont\n",
    "#         print('n_in',n_in)\n",
    "        self.layers=[]\n",
    "        for l in self.layer_count:\n",
    "            self.layers.append(nn.Linear(n_in,l))\n",
    "            self.layers.append(nn.ReLU(inplace=True))\n",
    "            self.layers.append(nn.Dropout(p))\n",
    "            self.layers.append(nn.BatchNorm1d(l))            \n",
    "            n_in=l\n",
    "        self.layers.append(nn.Linear(self.layer_count[-1],self.output_features))\n",
    "#         print(self.layers)\n",
    "        self.final_layers=nn.Sequential(*self.layers)\n",
    "#         print(self.final_layers)\n",
    "        \n",
    "    def forward(self,cat_cols, cont_cols):\n",
    "        '''\n",
    "        1. Create the embedding for cat and create one final input value for the forward path (only one time this gets created)\n",
    "        2. Create the dropout for this input layer\n",
    "        3. Create a batch norm for this layer\\\n",
    "        4. Pass this into the all the layers moving forward (starting with the first linear layer)\n",
    "        '''\n",
    "        #Creating the embedding for categorical columns\n",
    "        embeds=[]\n",
    "        for i,e in enumerate(self.embeddings):\n",
    "#             print('e' ,e)\n",
    "            embeds.append(e(cat_cols[:,i]))\n",
    "        cat_final=torch.cat(embeds,axis=1)\n",
    "#         print(cat_final.shape) #17 columns cos 12 for hours, 2 for AM/PM and 4 for days post embedding\n",
    "#         print(cont_cols.shape) # 6 columns one for each of the 6 features\n",
    "        \n",
    "        #Batch normalize the continuous variables first\n",
    "        cont_cols=self.batch_norm(cont_cols)\n",
    "        cat_final=self.dropout(cat_final)\n",
    "        self.X=torch.cat((cat_final, cont_cols),axis=1)\n",
    "#         print('X shape', self.X.shape)\n",
    "#         print('X dtype',self.X.dtype)\n",
    "#         print('X class',type(self.X))\n",
    "        #Dropout for the complete data set\n",
    "        #self.X=self.dropout(self.X)\n",
    "        self.X=self.final_layers(self.X)\n",
    "        return self.X\n",
    "            \n",
    "\n",
    "        #Creating final data set with cat and cols\n",
    "          \n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.manual_seed(33)\n",
    "model = Model(cat_tensor,cont_tensor,layer_count=[200,100],output_features=1,embed_size_list=embed_size,p=0.4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[-0.5855],\n",
      "        [-0.8058],\n",
      "        [ 0.1915],\n",
      "        ...,\n",
      "        [-0.4630],\n",
      "        [ 0.1225],\n",
      "        [-0.0557]], grad_fn=<AddmmBackward>)\n"
     ]
    }
   ],
   "source": [
    "#Testing if forward() works\n",
    "z = model.forward(cat_tensor,cont_tensor)\n",
    "print(z)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "criterion=nn.MSELoss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "optimizer=torch.optim.Adam(model.parameters(),lr=0.001)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 7.Create train, test split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Kaggle data set says that the data is already shuffled. If the data is not shuffled, then we need to do a train test split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(120000,)"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "label.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test size 24000\n",
      "train size 96000\n",
      "X_train_cont torch.Size([24000, 6])\n",
      "X_train_cat torch.Size([24000, 3])\n",
      "X_test_cont torch.Size([96000, 6])\n",
      "X_test_cat torch.Size([96000, 3])\n",
      "label_train torch.Size([24000, 1])\n",
      "label_test torch.Size([96000, 1])\n"
     ]
    }
   ],
   "source": [
    "test_size=int(0.2*df.shape[0])\n",
    "print('test size',test_size)\n",
    "print('train size',df.shape[0]-test_size)\n",
    "cat_cols\n",
    "X_train_cont=cont_tensor[:test_size,:]\n",
    "X_train_cat=cat_tensor[:test_size,:]\n",
    "X_test_cont=cont_tensor[test_size:,:]\n",
    "X_test_cat=cat_tensor[test_size:]\n",
    "label_train=label_tensor[:test_size].reshape(-1,1)\n",
    "label_test=label_tensor[test_size:].reshape(-1,1)\n",
    "print('X_train_cont',X_train_cont.shape)\n",
    "print('X_train_cat',X_train_cat.shape)\n",
    "print('X_test_cont',X_test_cont.shape)\n",
    "print('X_test_cat',X_test_cat.shape)\n",
    "print('label_train',label_train.shape)\n",
    "print('label_test',label_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 8. Building the train model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Loss at epoch 10 is 11.973340034484863\n",
      " Loss at epoch 20 is 11.446440696716309\n",
      " Loss at epoch 30 is 11.045608520507812\n",
      " Loss at epoch 40 is 10.684078216552734\n",
      " Loss at epoch 50 is 10.370415687561035\n",
      " Loss at epoch 60 is 10.076183319091797\n",
      " Loss at epoch 70 is 9.8255615234375\n",
      " Loss at epoch 80 is 9.548290252685547\n",
      " Loss at epoch 90 is 9.270340919494629\n",
      " Loss at epoch 100 is 8.994292259216309\n",
      " Loss at epoch 110 is 8.701070785522461\n",
      " Loss at epoch 120 is 8.375794410705566\n",
      " Loss at epoch 130 is 8.000714302062988\n",
      " Loss at epoch 140 is 7.601511001586914\n",
      " Loss at epoch 150 is 7.159296035766602\n",
      " Loss at epoch 160 is 6.678114891052246\n",
      " Loss at epoch 170 is 6.224819660186768\n",
      " Loss at epoch 180 is 5.713708400726318\n",
      " Loss at epoch 190 is 5.243253231048584\n",
      " Loss at epoch 200 is 4.778747081756592\n",
      " Loss at epoch 210 is 4.378846645355225\n",
      " Loss at epoch 220 is 4.063196182250977\n",
      " Loss at epoch 230 is 3.8786251544952393\n",
      " Loss at epoch 240 is 3.6867034435272217\n",
      " Loss at epoch 250 is 3.6485307216644287\n",
      " Loss at epoch 260 is 3.5889039039611816\n",
      " Loss at epoch 270 is 3.581583023071289\n",
      " Loss at epoch 280 is 3.567286252975464\n",
      " Loss at epoch 290 is 3.5726301670074463\n",
      " Loss at epoch 300 is 3.540520429611206\n"
     ]
    }
   ],
   "source": [
    "epochs=300\n",
    "losses3=[]\n",
    "for e in range(epochs):\n",
    "    e+=1\n",
    "    y_pred=model.forward(X_train_cat,X_train_cont)\n",
    "    loss=criterion(y_pred,label_train)**0.5\n",
    "    losses3.append(loss)\n",
    "    if e%10==0:\n",
    "        print(f' Loss at epoch {e} is {loss}')\n",
    "    \n",
    "    optimizer.zero_grad()\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(range(epochs),losses3,'b')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Losses')\n",
    "plt.title('Loss for lr=0.1 (red) and lr=0.001 (green)')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 9. Running on test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "with torch.no_grad():\n",
    "    y_preds=model.forward(X_test_cat,X_test_cont)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor(3.5872)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loss_test=criterion(y_preds, label_test)\n",
    "loss_test**0.5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 10. Analysing the error spread"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "error=(y_preds-label_test).numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                  0\n",
      "count  96000.000000\n",
      "mean      -0.174317\n",
      "std        3.582977\n",
      "min      -51.335407\n",
      "25%       -1.369182\n",
      "50%        0.111890\n",
      "75%        1.467132\n",
      "max       63.726105\n",
      "Mean of error is  -0.1743\n",
      "Std dev of error is   3.5830\n",
      "Range of error is 115.0615\n",
      "Max of error is  63.7261 at sample 48310\n",
      "Max of error is -51.3354 at sample 47838\n"
     ]
    },
    {
     "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": [
    "print(pd.DataFrame(error).describe())\n",
    "print('Mean of error is {:8.4f}'.format(error.mean() ))\n",
    "print('Std dev of error is {:8.4f}'.format(error.std()))\n",
    "print('Range of error is {:8.4f}'.format(error.max()-error.min()))\n",
    "print('Max of error is {:8.4f} at sample {}'.format(error.max(), (error.argmax())))\n",
    "print('Max of error is {:8.4f} at sample {}'.format(error.min(), (error.argmin())))\n",
    "plt.hist(error)\n",
    "plt.title('Distribution of the error variable for test case')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "one_std_dev_p=error.mean()+error.std()\n",
    "one_std_dev_n=error.mean()-error.std()\n",
    "ranges=((error<=one_std_dev_p) & (error>=one_std_dev_n)).sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "% of error within one standard deviation on either side is:   86.910%\n"
     ]
    }
   ],
   "source": [
    "print('% of error within one standard deviation on either side is: {:8.3f}%'.format(ranges/len(error)*100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 11. Saving the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "if len(losses3)==epochs:\n",
    "    torch.save(model.state_dict(),'uber_model_weights.pt')\n",
    "    torch.save(model,'uber_model.pkl')\n",
    "else:\n",
    "    print(\"You haven't trained this model! Only trained models should be saved!\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 12. Create program to directly run data transformation and through model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Refer Solution Part 2.ipynb </b>"
   ]
  },
  {
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}