JARVIS.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# The code used for the JARVIS AI hackathon at IIT Madras. This was held in 2019 in the month of January. After that, this notebook was updated for removing clumsiness and for easy execution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pd.read_csv('train.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "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>id</th>\n",
       "      <th>region_id</th>\n",
       "      <th>category_id</th>\n",
       "      <th>district_id</th>\n",
       "      <th>city_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>created_at_first</th>\n",
       "      <th>title</th>\n",
       "      <th>description</th>\n",
       "      <th>listing_num_images</th>\n",
       "      <th>listing_price</th>\n",
       "      <th>l1</th>\n",
       "      <th>l2</th>\n",
       "      <th>demand</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1000001</td>\n",
       "      <td>5040</td>\n",
       "      <td>910</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61232</td>\n",
       "      <td>1047742</td>\n",
       "      <td>2018-06-05 10:58:33</td>\n",
       "      <td>Heavy duty steel prison bars, can be used as g...</td>\n",
       "      <td>Heavy duty steel prison bars, can be used as g...</td>\n",
       "      <td>1</td>\n",
       "      <td>1500.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Tools &amp; DIY</td>\n",
       "      <td>0.203867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1000002</td>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>7000598</td>\n",
       "      <td>1113762</td>\n",
       "      <td>2018-06-05 11:01:46</td>\n",
       "      <td>6 seater dinning room suite available only had...</td>\n",
       "      <td>Hi I have moved to a smaller place and there i...</td>\n",
       "      <td>2</td>\n",
       "      <td>3000.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Furniture &amp; Decor</td>\n",
       "      <td>0.323121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1000006</td>\n",
       "      <td>5048</td>\n",
       "      <td>378</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61491</td>\n",
       "      <td>1176953</td>\n",
       "      <td>2018-06-05 11:10:16</td>\n",
       "      <td>Toyota Hilux 2011 model very low milege FSH wi...</td>\n",
       "      <td>2011 Hilux for sale  very low milege  canopy i...</td>\n",
       "      <td>3</td>\n",
       "      <td>142000.0</td>\n",
       "      <td>Vehicles</td>\n",
       "      <td>Cars &amp; Bakkies</td>\n",
       "      <td>0.407734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1000007</td>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61323</td>\n",
       "      <td>1031691</td>\n",
       "      <td>2018-06-05 11:11:44</td>\n",
       "      <td>Black wooden coffee table</td>\n",
       "      <td>Excellent condition. Price negotiable</td>\n",
       "      <td>1</td>\n",
       "      <td>760.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Furniture &amp; Decor</td>\n",
       "      <td>0.407734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1000008</td>\n",
       "      <td>5043</td>\n",
       "      <td>803</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61370</td>\n",
       "      <td>1176918</td>\n",
       "      <td>2018-06-05 11:11:55</td>\n",
       "      <td>Repairs and fixing of all laptops and computer...</td>\n",
       "      <td>we are expert in all laptops repairs and fixin...</td>\n",
       "      <td>1</td>\n",
       "      <td>450.0</td>\n",
       "      <td>Electronics &amp; Computers</td>\n",
       "      <td>Computers &amp; Laptops</td>\n",
       "      <td>0.203867</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  region_id  category_id  district_id  city_id  user_id  \\\n",
       "0  1000001       5040          910      65535.0    61232  1047742   \n",
       "1  1000002       5042          807      65535.0  7000598  1113762   \n",
       "2  1000006       5048          378      65535.0    61491  1176953   \n",
       "3  1000007       5042          807      65535.0    61323  1031691   \n",
       "4  1000008       5043          803      65535.0    61370  1176918   \n",
       "\n",
       "      created_at_first                                              title  \\\n",
       "0  2018-06-05 10:58:33  Heavy duty steel prison bars, can be used as g...   \n",
       "1  2018-06-05 11:01:46  6 seater dinning room suite available only had...   \n",
       "2  2018-06-05 11:10:16  Toyota Hilux 2011 model very low milege FSH wi...   \n",
       "3  2018-06-05 11:11:44                          Black wooden coffee table   \n",
       "4  2018-06-05 11:11:55  Repairs and fixing of all laptops and computer...   \n",
       "\n",
       "                                         description  listing_num_images  \\\n",
       "0  Heavy duty steel prison bars, can be used as g...                   1   \n",
       "1  Hi I have moved to a smaller place and there i...                   2   \n",
       "2  2011 Hilux for sale  very low milege  canopy i...                   3   \n",
       "3              Excellent condition. Price negotiable                   1   \n",
       "4  we are expert in all laptops repairs and fixin...                   1   \n",
       "\n",
       "   listing_price                       l1                   l2    demand  \n",
       "0         1500.0     Home, Garden & Tools          Tools & DIY  0.203867  \n",
       "1         3000.0     Home, Garden & Tools    Furniture & Decor  0.323121  \n",
       "2       142000.0                 Vehicles       Cars & Bakkies  0.407734  \n",
       "3          760.0     Home, Garden & Tools    Furniture & Decor  0.407734  \n",
       "4          450.0  Electronics & Computers  Computers & Laptops  0.203867  "
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 347014 entries, 0 to 347013\n",
      "Data columns (total 14 columns):\n",
      "id                    347014 non-null int64\n",
      "region_id             347014 non-null int64\n",
      "category_id           347014 non-null int64\n",
      "district_id           346864 non-null float64\n",
      "city_id               347014 non-null int64\n",
      "user_id               347014 non-null int64\n",
      "created_at_first      347014 non-null object\n",
      "title                 347014 non-null object\n",
      "description           267376 non-null object\n",
      "listing_num_images    347014 non-null int64\n",
      "listing_price         346292 non-null float64\n",
      "l1                    347014 non-null object\n",
      "l2                    347014 non-null object\n",
      "demand                347014 non-null float64\n",
      "dtypes: float64(3), int64(6), object(5)\n",
      "memory usage: 37.1+ MB\n"
     ]
    }
   ],
   "source": [
    "train.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb1506db4a8>"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.heatmap(train.isnull(),cmap='viridis')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "train.drop(['description'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "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>id</th>\n",
       "      <th>region_id</th>\n",
       "      <th>category_id</th>\n",
       "      <th>district_id</th>\n",
       "      <th>city_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>created_at_first</th>\n",
       "      <th>title</th>\n",
       "      <th>listing_num_images</th>\n",
       "      <th>listing_price</th>\n",
       "      <th>l1</th>\n",
       "      <th>l2</th>\n",
       "      <th>demand</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1000001</td>\n",
       "      <td>5040</td>\n",
       "      <td>910</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61232</td>\n",
       "      <td>1047742</td>\n",
       "      <td>2018-06-05 10:58:33</td>\n",
       "      <td>Heavy duty steel prison bars, can be used as g...</td>\n",
       "      <td>1</td>\n",
       "      <td>1500.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Tools &amp; DIY</td>\n",
       "      <td>0.203867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1000002</td>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>7000598</td>\n",
       "      <td>1113762</td>\n",
       "      <td>2018-06-05 11:01:46</td>\n",
       "      <td>6 seater dinning room suite available only had...</td>\n",
       "      <td>2</td>\n",
       "      <td>3000.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Furniture &amp; Decor</td>\n",
       "      <td>0.323121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1000006</td>\n",
       "      <td>5048</td>\n",
       "      <td>378</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61491</td>\n",
       "      <td>1176953</td>\n",
       "      <td>2018-06-05 11:10:16</td>\n",
       "      <td>Toyota Hilux 2011 model very low milege FSH wi...</td>\n",
       "      <td>3</td>\n",
       "      <td>142000.0</td>\n",
       "      <td>Vehicles</td>\n",
       "      <td>Cars &amp; Bakkies</td>\n",
       "      <td>0.407734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1000007</td>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61323</td>\n",
       "      <td>1031691</td>\n",
       "      <td>2018-06-05 11:11:44</td>\n",
       "      <td>Black wooden coffee table</td>\n",
       "      <td>1</td>\n",
       "      <td>760.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Furniture &amp; Decor</td>\n",
       "      <td>0.407734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1000008</td>\n",
       "      <td>5043</td>\n",
       "      <td>803</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61370</td>\n",
       "      <td>1176918</td>\n",
       "      <td>2018-06-05 11:11:55</td>\n",
       "      <td>Repairs and fixing of all laptops and computer...</td>\n",
       "      <td>1</td>\n",
       "      <td>450.0</td>\n",
       "      <td>Electronics &amp; Computers</td>\n",
       "      <td>Computers &amp; Laptops</td>\n",
       "      <td>0.203867</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  region_id  category_id  district_id  city_id  user_id  \\\n",
       "0  1000001       5040          910      65535.0    61232  1047742   \n",
       "1  1000002       5042          807      65535.0  7000598  1113762   \n",
       "2  1000006       5048          378      65535.0    61491  1176953   \n",
       "3  1000007       5042          807      65535.0    61323  1031691   \n",
       "4  1000008       5043          803      65535.0    61370  1176918   \n",
       "\n",
       "      created_at_first                                              title  \\\n",
       "0  2018-06-05 10:58:33  Heavy duty steel prison bars, can be used as g...   \n",
       "1  2018-06-05 11:01:46  6 seater dinning room suite available only had...   \n",
       "2  2018-06-05 11:10:16  Toyota Hilux 2011 model very low milege FSH wi...   \n",
       "3  2018-06-05 11:11:44                          Black wooden coffee table   \n",
       "4  2018-06-05 11:11:55  Repairs and fixing of all laptops and computer...   \n",
       "\n",
       "   listing_num_images  listing_price                       l1  \\\n",
       "0                   1         1500.0     Home, Garden & Tools   \n",
       "1                   2         3000.0     Home, Garden & Tools   \n",
       "2                   3       142000.0                 Vehicles   \n",
       "3                   1          760.0     Home, Garden & Tools   \n",
       "4                   1          450.0  Electronics & Computers   \n",
       "\n",
       "                    l2    demand  \n",
       "0          Tools & DIY  0.203867  \n",
       "1    Furniture & Decor  0.323121  \n",
       "2       Cars & Bakkies  0.407734  \n",
       "3    Furniture & Decor  0.407734  \n",
       "4  Computers & Laptops  0.203867  "
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb150620358>"
      ]
     },
     "execution_count": 91,
     "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": [
    "sns.scatterplot(x=train['region_id'],y=train['demand'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb15059dbe0>"
      ]
     },
     "execution_count": 92,
     "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": [
    "sns.barplot(x='region_id',y='demand',data=train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "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>id</th>\n",
       "      <th>region_id</th>\n",
       "      <th>category_id</th>\n",
       "      <th>district_id</th>\n",
       "      <th>city_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>created_at_first</th>\n",
       "      <th>title</th>\n",
       "      <th>listing_num_images</th>\n",
       "      <th>listing_price</th>\n",
       "      <th>l1</th>\n",
       "      <th>l2</th>\n",
       "      <th>demand</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1000001</td>\n",
       "      <td>5040</td>\n",
       "      <td>910</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61232</td>\n",
       "      <td>1047742</td>\n",
       "      <td>2018-06-05 10:58:33</td>\n",
       "      <td>Heavy duty steel prison bars, can be used as g...</td>\n",
       "      <td>1</td>\n",
       "      <td>1500.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Tools &amp; DIY</td>\n",
       "      <td>0.203867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1000002</td>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>7000598</td>\n",
       "      <td>1113762</td>\n",
       "      <td>2018-06-05 11:01:46</td>\n",
       "      <td>6 seater dinning room suite available only had...</td>\n",
       "      <td>2</td>\n",
       "      <td>3000.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Furniture &amp; Decor</td>\n",
       "      <td>0.323121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1000006</td>\n",
       "      <td>5048</td>\n",
       "      <td>378</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61491</td>\n",
       "      <td>1176953</td>\n",
       "      <td>2018-06-05 11:10:16</td>\n",
       "      <td>Toyota Hilux 2011 model very low milege FSH wi...</td>\n",
       "      <td>3</td>\n",
       "      <td>142000.0</td>\n",
       "      <td>Vehicles</td>\n",
       "      <td>Cars &amp; Bakkies</td>\n",
       "      <td>0.407734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1000007</td>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61323</td>\n",
       "      <td>1031691</td>\n",
       "      <td>2018-06-05 11:11:44</td>\n",
       "      <td>Black wooden coffee table</td>\n",
       "      <td>1</td>\n",
       "      <td>760.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Furniture &amp; Decor</td>\n",
       "      <td>0.407734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1000008</td>\n",
       "      <td>5043</td>\n",
       "      <td>803</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61370</td>\n",
       "      <td>1176918</td>\n",
       "      <td>2018-06-05 11:11:55</td>\n",
       "      <td>Repairs and fixing of all laptops and computer...</td>\n",
       "      <td>1</td>\n",
       "      <td>450.0</td>\n",
       "      <td>Electronics &amp; Computers</td>\n",
       "      <td>Computers &amp; Laptops</td>\n",
       "      <td>0.203867</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  region_id  category_id  district_id  city_id  user_id  \\\n",
       "0  1000001       5040          910      65535.0    61232  1047742   \n",
       "1  1000002       5042          807      65535.0  7000598  1113762   \n",
       "2  1000006       5048          378      65535.0    61491  1176953   \n",
       "3  1000007       5042          807      65535.0    61323  1031691   \n",
       "4  1000008       5043          803      65535.0    61370  1176918   \n",
       "\n",
       "      created_at_first                                              title  \\\n",
       "0  2018-06-05 10:58:33  Heavy duty steel prison bars, can be used as g...   \n",
       "1  2018-06-05 11:01:46  6 seater dinning room suite available only had...   \n",
       "2  2018-06-05 11:10:16  Toyota Hilux 2011 model very low milege FSH wi...   \n",
       "3  2018-06-05 11:11:44                          Black wooden coffee table   \n",
       "4  2018-06-05 11:11:55  Repairs and fixing of all laptops and computer...   \n",
       "\n",
       "   listing_num_images  listing_price                       l1  \\\n",
       "0                   1         1500.0     Home, Garden & Tools   \n",
       "1                   2         3000.0     Home, Garden & Tools   \n",
       "2                   3       142000.0                 Vehicles   \n",
       "3                   1          760.0     Home, Garden & Tools   \n",
       "4                   1          450.0  Electronics & Computers   \n",
       "\n",
       "                    l2    demand  \n",
       "0          Tools & DIY  0.203867  \n",
       "1    Furniture & Decor  0.323121  \n",
       "2       Cars & Bakkies  0.407734  \n",
       "3    Furniture & Decor  0.407734  \n",
       "4  Computers & Laptops  0.203867  "
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "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>id</th>\n",
       "      <th>region_id</th>\n",
       "      <th>category_id</th>\n",
       "      <th>district_id</th>\n",
       "      <th>city_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>created_at_first</th>\n",
       "      <th>title</th>\n",
       "      <th>listing_num_images</th>\n",
       "      <th>listing_price</th>\n",
       "      <th>l1</th>\n",
       "      <th>l2</th>\n",
       "      <th>demand</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1000001</td>\n",
       "      <td>5040</td>\n",
       "      <td>910</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61232</td>\n",
       "      <td>1047742</td>\n",
       "      <td>2018-06-05 10:58:33</td>\n",
       "      <td>Heavy duty steel prison bars, can be used as g...</td>\n",
       "      <td>1</td>\n",
       "      <td>1500.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Tools &amp; DIY</td>\n",
       "      <td>0.203867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1000002</td>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>7000598</td>\n",
       "      <td>1113762</td>\n",
       "      <td>2018-06-05 11:01:46</td>\n",
       "      <td>6 seater dinning room suite available only had...</td>\n",
       "      <td>2</td>\n",
       "      <td>3000.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Furniture &amp; Decor</td>\n",
       "      <td>0.323121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1000006</td>\n",
       "      <td>5048</td>\n",
       "      <td>378</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61491</td>\n",
       "      <td>1176953</td>\n",
       "      <td>2018-06-05 11:10:16</td>\n",
       "      <td>Toyota Hilux 2011 model very low milege FSH wi...</td>\n",
       "      <td>3</td>\n",
       "      <td>142000.0</td>\n",
       "      <td>Vehicles</td>\n",
       "      <td>Cars &amp; Bakkies</td>\n",
       "      <td>0.407734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1000007</td>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61323</td>\n",
       "      <td>1031691</td>\n",
       "      <td>2018-06-05 11:11:44</td>\n",
       "      <td>Black wooden coffee table</td>\n",
       "      <td>1</td>\n",
       "      <td>760.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Furniture &amp; Decor</td>\n",
       "      <td>0.407734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1000008</td>\n",
       "      <td>5043</td>\n",
       "      <td>803</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61370</td>\n",
       "      <td>1176918</td>\n",
       "      <td>2018-06-05 11:11:55</td>\n",
       "      <td>Repairs and fixing of all laptops and computer...</td>\n",
       "      <td>1</td>\n",
       "      <td>450.0</td>\n",
       "      <td>Electronics &amp; Computers</td>\n",
       "      <td>Computers &amp; Laptops</td>\n",
       "      <td>0.203867</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  region_id  category_id  district_id  city_id  user_id  \\\n",
       "0  1000001       5040          910      65535.0    61232  1047742   \n",
       "1  1000002       5042          807      65535.0  7000598  1113762   \n",
       "2  1000006       5048          378      65535.0    61491  1176953   \n",
       "3  1000007       5042          807      65535.0    61323  1031691   \n",
       "4  1000008       5043          803      65535.0    61370  1176918   \n",
       "\n",
       "      created_at_first                                              title  \\\n",
       "0  2018-06-05 10:58:33  Heavy duty steel prison bars, can be used as g...   \n",
       "1  2018-06-05 11:01:46  6 seater dinning room suite available only had...   \n",
       "2  2018-06-05 11:10:16  Toyota Hilux 2011 model very low milege FSH wi...   \n",
       "3  2018-06-05 11:11:44                          Black wooden coffee table   \n",
       "4  2018-06-05 11:11:55  Repairs and fixing of all laptops and computer...   \n",
       "\n",
       "   listing_num_images  listing_price                       l1  \\\n",
       "0                   1         1500.0     Home, Garden & Tools   \n",
       "1                   2         3000.0     Home, Garden & Tools   \n",
       "2                   3       142000.0                 Vehicles   \n",
       "3                   1          760.0     Home, Garden & Tools   \n",
       "4                   1          450.0  Electronics & Computers   \n",
       "\n",
       "                    l2    demand  \n",
       "0          Tools & DIY  0.203867  \n",
       "1    Furniture & Decor  0.323121  \n",
       "2       Cars & Bakkies  0.407734  \n",
       "3    Furniture & Decor  0.407734  \n",
       "4  Computers & Laptops  0.203867  "
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = train[['region_id','category_id','city_id','listing_num_images']]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "#X = pd.concat([X,train['demand']],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "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>region_id</th>\n",
       "      <th>category_id</th>\n",
       "      <th>city_id</th>\n",
       "      <th>listing_num_images</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5040</td>\n",
       "      <td>910</td>\n",
       "      <td>61232</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>7000598</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5048</td>\n",
       "      <td>378</td>\n",
       "      <td>61491</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>61323</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5043</td>\n",
       "      <td>803</td>\n",
       "      <td>61370</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   region_id  category_id  city_id  listing_num_images\n",
       "0       5040          910    61232                   1\n",
       "1       5042          807  7000598                   2\n",
       "2       5048          378    61491                   3\n",
       "3       5042          807    61323                   1\n",
       "4       5043          803    61370                   1"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "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>region_id</th>\n",
       "      <th>category_id</th>\n",
       "      <th>city_id</th>\n",
       "      <th>listing_num_images</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5040</td>\n",
       "      <td>910</td>\n",
       "      <td>61232</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>7000598</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5048</td>\n",
       "      <td>378</td>\n",
       "      <td>61491</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>61323</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5043</td>\n",
       "      <td>803</td>\n",
       "      <td>61370</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   region_id  category_id  city_id  listing_num_images\n",
       "0       5040          910    61232                   1\n",
       "1       5042          807  7000598                   2\n",
       "2       5048          378    61491                   3\n",
       "3       5042          807    61323                   1\n",
       "4       5043          803    61370                   1"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 347014 entries, 0 to 347013\n",
      "Data columns (total 4 columns):\n",
      "region_id             347014 non-null int64\n",
      "category_id           347014 non-null int64\n",
      "city_id               347014 non-null int64\n",
      "listing_num_images    347014 non-null int64\n",
      "dtypes: int64(4)\n",
      "memory usage: 10.6 MB\n"
     ]
    }
   ],
   "source": [
    "X.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb15053f390>"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.heatmap(X.isnull(),cmap='viridis')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "c = pd.get_dummies(train['l1'],drop_first=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = pd.concat([X,c],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "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>region_id</th>\n",
       "      <th>category_id</th>\n",
       "      <th>city_id</th>\n",
       "      <th>listing_num_images</th>\n",
       "      <th>Farming &amp; Industrial</th>\n",
       "      <th>Fashion &amp; Beauty</th>\n",
       "      <th>Hobbies &amp; Interests</th>\n",
       "      <th>Home, Garden &amp; Tools</th>\n",
       "      <th>Jobs</th>\n",
       "      <th>Kids &amp; Baby</th>\n",
       "      <th>Office &amp; Business</th>\n",
       "      <th>Pets</th>\n",
       "      <th>Property</th>\n",
       "      <th>Services</th>\n",
       "      <th>Sports &amp; Outdoors</th>\n",
       "      <th>Vehicles</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5040</td>\n",
       "      <td>910</td>\n",
       "      <td>61232</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>7000598</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5048</td>\n",
       "      <td>378</td>\n",
       "      <td>61491</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>61323</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5043</td>\n",
       "      <td>803</td>\n",
       "      <td>61370</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   region_id  category_id  city_id  listing_num_images  Farming & Industrial  \\\n",
       "0       5040          910    61232                   1                     0   \n",
       "1       5042          807  7000598                   2                     0   \n",
       "2       5048          378    61491                   3                     0   \n",
       "3       5042          807    61323                   1                     0   \n",
       "4       5043          803    61370                   1                     0   \n",
       "\n",
       "   Fashion & Beauty  Hobbies & Interests  Home, Garden & Tools  Jobs  \\\n",
       "0                 0                    0                     1     0   \n",
       "1                 0                    0                     1     0   \n",
       "2                 0                    0                     0     0   \n",
       "3                 0                    0                     1     0   \n",
       "4                 0                    0                     0     0   \n",
       "\n",
       "   Kids & Baby  Office & Business  Pets  Property  Services  \\\n",
       "0            0                  0     0         0         0   \n",
       "1            0                  0     0         0         0   \n",
       "2            0                  0     0         0         0   \n",
       "3            0                  0     0         0         0   \n",
       "4            0                  0     0         0         0   \n",
       "\n",
       "   Sports & Outdoors  Vehicles  \n",
       "0                  0         0  \n",
       "1                  0         0  \n",
       "2                  0         1  \n",
       "3                  0         0  \n",
       "4                  0         0  "
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "c = pd.get_dummies(train['l2'],drop_first=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = pd.concat([X,c],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "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>region_id</th>\n",
       "      <th>category_id</th>\n",
       "      <th>city_id</th>\n",
       "      <th>listing_num_images</th>\n",
       "      <th>Farming &amp; Industrial</th>\n",
       "      <th>Fashion &amp; Beauty</th>\n",
       "      <th>Hobbies &amp; Interests</th>\n",
       "      <th>Home, Garden &amp; Tools</th>\n",
       "      <th>Jobs</th>\n",
       "      <th>Kids &amp; Baby</th>\n",
       "      <th>...</th>\n",
       "      <th>Rooms for rent &amp; Shared</th>\n",
       "      <th>Shop &amp; Catering Equipment</th>\n",
       "      <th>TV, Audio &amp; Visual</th>\n",
       "      <th>Toddler's Clothing &amp; Accessories</th>\n",
       "      <th>Tools &amp; DIY</th>\n",
       "      <th>Toys, Games &amp; Remote Control</th>\n",
       "      <th>Transport</th>\n",
       "      <th>Trucks &amp; Commercial Vehicles</th>\n",
       "      <th>Vacation Rentals</th>\n",
       "      <th>iPads &amp; Tablets</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5040</td>\n",
       "      <td>910</td>\n",
       "      <td>61232</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>7000598</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5048</td>\n",
       "      <td>378</td>\n",
       "      <td>61491</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>61323</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5043</td>\n",
       "      <td>803</td>\n",
       "      <td>61370</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 70 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   region_id  category_id  city_id  listing_num_images  Farming & Industrial  \\\n",
       "0       5040          910    61232                   1                     0   \n",
       "1       5042          807  7000598                   2                     0   \n",
       "2       5048          378    61491                   3                     0   \n",
       "3       5042          807    61323                   1                     0   \n",
       "4       5043          803    61370                   1                     0   \n",
       "\n",
       "   Fashion & Beauty  Hobbies & Interests  Home, Garden & Tools  Jobs  \\\n",
       "0                 0                    0                     1     0   \n",
       "1                 0                    0                     1     0   \n",
       "2                 0                    0                     0     0   \n",
       "3                 0                    0                     1     0   \n",
       "4                 0                    0                     0     0   \n",
       "\n",
       "   Kids & Baby  ...  Rooms for rent & Shared  Shop & Catering Equipment  \\\n",
       "0            0  ...                        0                          0   \n",
       "1            0  ...                        0                          0   \n",
       "2            0  ...                        0                          0   \n",
       "3            0  ...                        0                          0   \n",
       "4            0  ...                        0                          0   \n",
       "\n",
       "   TV, Audio & Visual  Toddler's Clothing & Accessories  Tools & DIY  \\\n",
       "0                   0                                 0            1   \n",
       "1                   0                                 0            0   \n",
       "2                   0                                 0            0   \n",
       "3                   0                                 0            0   \n",
       "4                   0                                 0            0   \n",
       "\n",
       "   Toys, Games & Remote Control  Transport  Trucks & Commercial Vehicles  \\\n",
       "0                             0          0                             0   \n",
       "1                             0          0                             0   \n",
       "2                             0          0                             0   \n",
       "3                             0          0                             0   \n",
       "4                             0          0                             0   \n",
       "\n",
       "   Vacation Rentals  iPads & Tablets  \n",
       "0                 0                0  \n",
       "1                 0                0  \n",
       "2                 0                0  \n",
       "3                 0                0  \n",
       "4                 0                0  \n",
       "\n",
       "[5 rows x 70 columns]"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0.203867\n",
       "1    0.323121\n",
       "2    0.407734\n",
       "3    0.407734\n",
       "4    0.203867\n",
       "Name: demand, dtype: float64"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = train['demand']\n",
    "y.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = pd.concat([train['id'],X],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "lm = LinearRegression()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None, normalize=False)"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lm.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "prediction = lm.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7fb150432d68>"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(y_test,prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MAE: 0.17733808349333632\n",
      "MSE: 0.04953557327073543\n",
      "RMSE: 0.2225658852356655\n"
     ]
    }
   ],
   "source": [
    "print('MAE:', metrics.mean_absolute_error(y_test, prediction))\n",
    "print('MSE:', metrics.mean_squared_error(y_test, prediction))\n",
    "print('RMSE:', np.sqrt(metrics.mean_squared_error(y_test, prediction)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None, normalize=False)"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [],
   "source": [
    "test = pd.read_csv('test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "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>id</th>\n",
       "      <th>region_id</th>\n",
       "      <th>category_id</th>\n",
       "      <th>district_id</th>\n",
       "      <th>city_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>created_at_first</th>\n",
       "      <th>title</th>\n",
       "      <th>description</th>\n",
       "      <th>listing_num_images</th>\n",
       "      <th>listing_price</th>\n",
       "      <th>l1</th>\n",
       "      <th>l2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1000020</td>\n",
       "      <td>5042</td>\n",
       "      <td>882</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>299952</td>\n",
       "      <td>1148634</td>\n",
       "      <td>2018-06-05 11:57:42</td>\n",
       "      <td>Bicycle</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "      <td>1900.0</td>\n",
       "      <td>Sports &amp; Outdoors</td>\n",
       "      <td>Bicycles</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1000025</td>\n",
       "      <td>5042</td>\n",
       "      <td>377</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61323</td>\n",
       "      <td>1027629</td>\n",
       "      <td>2018-06-05 12:20:21</td>\n",
       "      <td>Toyota New &amp; Used Tail Lights</td>\n",
       "      <td>We at AMS Specialise in Used and New Replaceme...</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Vehicles</td>\n",
       "      <td>Car Parts &amp; Accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1000047</td>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>7000502</td>\n",
       "      <td>1195460</td>\n",
       "      <td>2018-06-19 13:39:07</td>\n",
       "      <td>A Double bed. Base and Mattress set with white...</td>\n",
       "      <td>A Double bed. Base  and Mattress set with whit...</td>\n",
       "      <td>8</td>\n",
       "      <td>2200.0</td>\n",
       "      <td>Home, Garden &amp; Tools</td>\n",
       "      <td>Furniture &amp; Decor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1000054</td>\n",
       "      <td>5040</td>\n",
       "      <td>5171</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61231</td>\n",
       "      <td>1007869</td>\n",
       "      <td>2018-06-19 14:02:54</td>\n",
       "      <td>Screen printing Business for sale</td>\n",
       "      <td>Ideal for starting your own business at home a...</td>\n",
       "      <td>5</td>\n",
       "      <td>8500.0</td>\n",
       "      <td>Office &amp; Business</td>\n",
       "      <td>Businesses for Sale</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1000064</td>\n",
       "      <td>5048</td>\n",
       "      <td>378</td>\n",
       "      <td>65535.0</td>\n",
       "      <td>61491</td>\n",
       "      <td>1155621</td>\n",
       "      <td>2018-06-19 14:39:06</td>\n",
       "      <td>2003 Volkswagen Jetta 4</td>\n",
       "      <td>Jetta 4 2003 2L. Full House. Leather interior,...</td>\n",
       "      <td>7</td>\n",
       "      <td>50000.0</td>\n",
       "      <td>Vehicles</td>\n",
       "      <td>Cars &amp; Bakkies</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  region_id  category_id  district_id  city_id  user_id  \\\n",
       "0  1000020       5042          882      65535.0   299952  1148634   \n",
       "1  1000025       5042          377      65535.0    61323  1027629   \n",
       "2  1000047       5042          807      65535.0  7000502  1195460   \n",
       "3  1000054       5040         5171      65535.0    61231  1007869   \n",
       "4  1000064       5048          378      65535.0    61491  1155621   \n",
       "\n",
       "      created_at_first                                              title  \\\n",
       "0  2018-06-05 11:57:42                                            Bicycle   \n",
       "1  2018-06-05 12:20:21                      Toyota New & Used Tail Lights   \n",
       "2  2018-06-19 13:39:07  A Double bed. Base and Mattress set with white...   \n",
       "3  2018-06-19 14:02:54                  Screen printing Business for sale   \n",
       "4  2018-06-19 14:39:06                            2003 Volkswagen Jetta 4   \n",
       "\n",
       "                                         description  listing_num_images  \\\n",
       "0                                                NaN                   6   \n",
       "1  We at AMS Specialise in Used and New Replaceme...                   7   \n",
       "2  A Double bed. Base  and Mattress set with whit...                   8   \n",
       "3  Ideal for starting your own business at home a...                   5   \n",
       "4  Jetta 4 2003 2L. Full House. Leather interior,...                   7   \n",
       "\n",
       "   listing_price                    l1                       l2  \n",
       "0         1900.0     Sports & Outdoors                 Bicycles  \n",
       "1            1.0              Vehicles  Car Parts & Accessories  \n",
       "2         2200.0  Home, Garden & Tools        Furniture & Decor  \n",
       "3         8500.0     Office & Business      Businesses for Sale  \n",
       "4        50000.0              Vehicles           Cars & Bakkies  "
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "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>region_id</th>\n",
       "      <th>category_id</th>\n",
       "      <th>city_id</th>\n",
       "      <th>listing_num_images</th>\n",
       "      <th>Farming &amp; Industrial</th>\n",
       "      <th>Fashion &amp; Beauty</th>\n",
       "      <th>Hobbies &amp; Interests</th>\n",
       "      <th>Home, Garden &amp; Tools</th>\n",
       "      <th>Jobs</th>\n",
       "      <th>Kids &amp; Baby</th>\n",
       "      <th>...</th>\n",
       "      <th>Rooms for rent &amp; Shared</th>\n",
       "      <th>Shop &amp; Catering Equipment</th>\n",
       "      <th>TV, Audio &amp; Visual</th>\n",
       "      <th>Toddler's Clothing &amp; Accessories</th>\n",
       "      <th>Tools &amp; DIY</th>\n",
       "      <th>Toys, Games &amp; Remote Control</th>\n",
       "      <th>Transport</th>\n",
       "      <th>Trucks &amp; Commercial Vehicles</th>\n",
       "      <th>Vacation Rentals</th>\n",
       "      <th>iPads &amp; Tablets</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5042</td>\n",
       "      <td>882</td>\n",
       "      <td>299952</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5042</td>\n",
       "      <td>377</td>\n",
       "      <td>61323</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5042</td>\n",
       "      <td>807</td>\n",
       "      <td>7000502</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5040</td>\n",
       "      <td>5171</td>\n",
       "      <td>61231</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5048</td>\n",
       "      <td>378</td>\n",
       "      <td>61491</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 70 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   region_id  category_id  city_id  listing_num_images  Farming & Industrial  \\\n",
       "0       5042          882   299952                   6                     0   \n",
       "1       5042          377    61323                   7                     0   \n",
       "2       5042          807  7000502                   8                     0   \n",
       "3       5040         5171    61231                   5                     0   \n",
       "4       5048          378    61491                   7                     0   \n",
       "\n",
       "   Fashion & Beauty  Hobbies & Interests  Home, Garden & Tools  Jobs  \\\n",
       "0                 0                    0                     0     0   \n",
       "1                 0                    0                     0     0   \n",
       "2                 0                    0                     1     0   \n",
       "3                 0                    0                     0     0   \n",
       "4                 0                    0                     0     0   \n",
       "\n",
       "   Kids & Baby  ...  Rooms for rent & Shared  Shop & Catering Equipment  \\\n",
       "0            0  ...                        0                          0   \n",
       "1            0  ...                        0                          0   \n",
       "2            0  ...                        0                          0   \n",
       "3            0  ...                        0                          0   \n",
       "4            0  ...                        0                          0   \n",
       "\n",
       "   TV, Audio & Visual  Toddler's Clothing & Accessories  Tools & DIY  \\\n",
       "0                   0                                 0            0   \n",
       "1                   0                                 0            0   \n",
       "2                   0                                 0            0   \n",
       "3                   0                                 0            0   \n",
       "4                   0                                 0            0   \n",
       "\n",
       "   Toys, Games & Remote Control  Transport  Trucks & Commercial Vehicles  \\\n",
       "0                             0          0                             0   \n",
       "1                             0          0                             0   \n",
       "2                             0          0                             0   \n",
       "3                             0          0                             0   \n",
       "4                             0          0                             0   \n",
       "\n",
       "   Vacation Rentals  iPads & Tablets  \n",
       "0                 0                0  \n",
       "1                 0                0  \n",
       "2                 0                0  \n",
       "3                 0                0  \n",
       "4                 0                0  \n",
       "\n",
       "[5 rows x 70 columns]"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_testing = test[['region_id','category_id','city_id','listing_num_images']]\n",
    "c = pd.get_dummies(test['l1'],drop_first=True)\n",
    "X_testing = pd.concat([X_testing,c],axis=1)\n",
    "c = pd.get_dummies(test['l2'],drop_first=True)\n",
    "X_testing = pd.concat([X_testing,c],axis=1)\n",
    "X_testing.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_testing = pd.concat([test['id'],X_testing],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb14ed83518>"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.heatmap(X_testing.isnull(),cmap='viridis')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [],
   "source": [
    "prediction = lm.predict(X_testing)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [],
   "source": [
    "sub = pd.DataFrame(prediction,index=X_testing['id'],columns=['demand'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1000020</th>\n",
       "      <td>0.354142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1000025</th>\n",
       "      <td>0.287073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1000047</th>\n",
       "      <td>0.331427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1000054</th>\n",
       "      <td>0.331476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1000064</th>\n",
       "      <td>0.387908</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           demand\n",
       "id               \n",
       "1000020  0.354142\n",
       "1000025  0.287073\n",
       "1000047  0.331427\n",
       "1000054  0.331476\n",
       "1000064  0.387908"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sub.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 86400 entries, 1000020 to 1050345\n",
      "Data columns (total 1 columns):\n",
      "demand    86400 non-null float64\n",
      "dtypes: float64(1)\n",
      "memory usage: 1.3 MB\n"
     ]
    }
   ],
   "source": [
    "sub.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [],
   "source": [
    "sub.reset_index(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "sub.to_csv(\"samplesub.csv\",index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}