homework6_20211020.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6.10 Homework\n",
    "\n",
    "The goal of this homework is to create a tree-based regression model for prediction apartment prices (column `'price'`).\n",
    "\n",
    "In this homework we'll again use the New York City Airbnb Open Data dataset - the same one we used in homework 2 and 3.\n",
    "\n",
    "You can take it from [Kaggle](https://www.kaggle.com/dgomonov/new-york-city-airbnb-open-data?select=AB_NYC_2019.csv)\n",
    "or download from [here](https://raw.githubusercontent.com/alexeygrigorev/datasets/master/AB_NYC_2019.csv)\n",
    "if you don't want to sign up to Kaggle.\n",
    "\n",
    "Let's load the data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = [\n",
    "    'neighbourhood_group', 'room_type', 'latitude', 'longitude',\n",
    "    'minimum_nights', 'number_of_reviews','reviews_per_month',\n",
    "    'calculated_host_listings_count', 'availability_365',\n",
    "    'price'\n",
    "]\n",
    "\n",
    "df = pd.read_csv('AB_NYC_2019.csv', usecols=columns)\n",
    "df.reviews_per_month = df.reviews_per_month.fillna(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Apply the log tranform to `price`\n",
    "* Do train/validation/test split with 60%/20%/20% distribution. \n",
    "* Use the `train_test_split` function and set the `random_state` parameter to 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>neighbourhood_group</th>\n",
       "      <th>latitude</th>\n",
       "      <th>longitude</th>\n",
       "      <th>room_type</th>\n",
       "      <th>price</th>\n",
       "      <th>minimum_nights</th>\n",
       "      <th>number_of_reviews</th>\n",
       "      <th>reviews_per_month</th>\n",
       "      <th>calculated_host_listings_count</th>\n",
       "      <th>availability_365</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Brooklyn</td>\n",
       "      <td>40.64749</td>\n",
       "      <td>-73.97237</td>\n",
       "      <td>Private room</td>\n",
       "      <td>149</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>0.21</td>\n",
       "      <td>6</td>\n",
       "      <td>365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Manhattan</td>\n",
       "      <td>40.75362</td>\n",
       "      <td>-73.98377</td>\n",
       "      <td>Entire home/apt</td>\n",
       "      <td>225</td>\n",
       "      <td>1</td>\n",
       "      <td>45</td>\n",
       "      <td>0.38</td>\n",
       "      <td>2</td>\n",
       "      <td>355</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Manhattan</td>\n",
       "      <td>40.80902</td>\n",
       "      <td>-73.94190</td>\n",
       "      <td>Private room</td>\n",
       "      <td>150</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Brooklyn</td>\n",
       "      <td>40.68514</td>\n",
       "      <td>-73.95976</td>\n",
       "      <td>Entire home/apt</td>\n",
       "      <td>89</td>\n",
       "      <td>1</td>\n",
       "      <td>270</td>\n",
       "      <td>4.64</td>\n",
       "      <td>1</td>\n",
       "      <td>194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Manhattan</td>\n",
       "      <td>40.79851</td>\n",
       "      <td>-73.94399</td>\n",
       "      <td>Entire home/apt</td>\n",
       "      <td>80</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "      <td>0.10</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  neighbourhood_group  latitude  longitude        room_type  price  \\\n",
       "0            Brooklyn  40.64749  -73.97237     Private room    149   \n",
       "1           Manhattan  40.75362  -73.98377  Entire home/apt    225   \n",
       "2           Manhattan  40.80902  -73.94190     Private room    150   \n",
       "3            Brooklyn  40.68514  -73.95976  Entire home/apt     89   \n",
       "4           Manhattan  40.79851  -73.94399  Entire home/apt     80   \n",
       "\n",
       "   minimum_nights  number_of_reviews  reviews_per_month  \\\n",
       "0               1                  9               0.21   \n",
       "1               1                 45               0.38   \n",
       "2               3                  0               0.00   \n",
       "3               1                270               4.64   \n",
       "4              10                  9               0.10   \n",
       "\n",
       "   calculated_host_listings_count  availability_365  \n",
       "0                               6               365  \n",
       "1                               2               355  \n",
       "2                               1               365  \n",
       "3                               1               194  \n",
       "4                               1                 0  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "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>latitude</th>\n",
       "      <th>longitude</th>\n",
       "      <th>price</th>\n",
       "      <th>minimum_nights</th>\n",
       "      <th>number_of_reviews</th>\n",
       "      <th>reviews_per_month</th>\n",
       "      <th>calculated_host_listings_count</th>\n",
       "      <th>availability_365</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>40.728949</td>\n",
       "      <td>-73.952170</td>\n",
       "      <td>152.720687</td>\n",
       "      <td>7.029962</td>\n",
       "      <td>23.274466</td>\n",
       "      <td>1.090910</td>\n",
       "      <td>7.143982</td>\n",
       "      <td>112.781327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.054530</td>\n",
       "      <td>0.046157</td>\n",
       "      <td>240.154170</td>\n",
       "      <td>20.510550</td>\n",
       "      <td>44.550582</td>\n",
       "      <td>1.597283</td>\n",
       "      <td>32.952519</td>\n",
       "      <td>131.622289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>40.499790</td>\n",
       "      <td>-74.244420</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>40.690100</td>\n",
       "      <td>-73.983070</td>\n",
       "      <td>69.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.040000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>40.723070</td>\n",
       "      <td>-73.955680</td>\n",
       "      <td>106.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>0.370000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>45.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>40.763115</td>\n",
       "      <td>-73.936275</td>\n",
       "      <td>175.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>1.580000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>227.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>40.913060</td>\n",
       "      <td>-73.712990</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>1250.000000</td>\n",
       "      <td>629.000000</td>\n",
       "      <td>58.500000</td>\n",
       "      <td>327.000000</td>\n",
       "      <td>365.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           latitude     longitude         price  minimum_nights  \\\n",
       "count  48895.000000  48895.000000  48895.000000    48895.000000   \n",
       "mean      40.728949    -73.952170    152.720687        7.029962   \n",
       "std        0.054530      0.046157    240.154170       20.510550   \n",
       "min       40.499790    -74.244420      0.000000        1.000000   \n",
       "25%       40.690100    -73.983070     69.000000        1.000000   \n",
       "50%       40.723070    -73.955680    106.000000        3.000000   \n",
       "75%       40.763115    -73.936275    175.000000        5.000000   \n",
       "max       40.913060    -73.712990  10000.000000     1250.000000   \n",
       "\n",
       "       number_of_reviews  reviews_per_month  calculated_host_listings_count  \\\n",
       "count       48895.000000       48895.000000                    48895.000000   \n",
       "mean           23.274466           1.090910                        7.143982   \n",
       "std            44.550582           1.597283                       32.952519   \n",
       "min             0.000000           0.000000                        1.000000   \n",
       "25%             1.000000           0.040000                        1.000000   \n",
       "50%             5.000000           0.370000                        1.000000   \n",
       "75%            24.000000           1.580000                        2.000000   \n",
       "max           629.000000          58.500000                      327.000000   \n",
       "\n",
       "       availability_365  \n",
       "count      48895.000000  \n",
       "mean         112.781327  \n",
       "std          131.622289  \n",
       "min            0.000000  \n",
       "25%            0.000000  \n",
       "50%           45.000000  \n",
       "75%          227.000000  \n",
       "max          365.000000  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "df_full_train, df_test = train_test_split(df, test_size=0.2, random_state=1)\n",
    "df_train, df_val = train_test_split(df_full_train, test_size=0.25, random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = df_train.reset_index(drop=True)\n",
    "df_val = df_val.reset_index(drop=True)\n",
    "df_test = df_test.reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train = np.log1p(df_train.price.values)\n",
    "y_val = np.log1p(df_val.price.values)\n",
    "y_test = np.log1p(df_test.price.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "del df_train['price']\n",
    "del df_val['price']\n",
    "del df_test['price']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>latitude</th>\n",
       "      <th>longitude</th>\n",
       "      <th>price</th>\n",
       "      <th>minimum_nights</th>\n",
       "      <th>number_of_reviews</th>\n",
       "      <th>reviews_per_month</th>\n",
       "      <th>calculated_host_listings_count</th>\n",
       "      <th>availability_365</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "      <td>48895.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>40.728949</td>\n",
       "      <td>-73.952170</td>\n",
       "      <td>152.720687</td>\n",
       "      <td>7.029962</td>\n",
       "      <td>23.274466</td>\n",
       "      <td>1.090910</td>\n",
       "      <td>7.143982</td>\n",
       "      <td>112.781327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.054530</td>\n",
       "      <td>0.046157</td>\n",
       "      <td>240.154170</td>\n",
       "      <td>20.510550</td>\n",
       "      <td>44.550582</td>\n",
       "      <td>1.597283</td>\n",
       "      <td>32.952519</td>\n",
       "      <td>131.622289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>40.499790</td>\n",
       "      <td>-74.244420</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>40.690100</td>\n",
       "      <td>-73.983070</td>\n",
       "      <td>69.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.040000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>40.723070</td>\n",
       "      <td>-73.955680</td>\n",
       "      <td>106.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>0.370000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>45.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>40.763115</td>\n",
       "      <td>-73.936275</td>\n",
       "      <td>175.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>24.000000</td>\n",
       "      <td>1.580000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>227.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>40.913060</td>\n",
       "      <td>-73.712990</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>1250.000000</td>\n",
       "      <td>629.000000</td>\n",
       "      <td>58.500000</td>\n",
       "      <td>327.000000</td>\n",
       "      <td>365.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           latitude     longitude         price  minimum_nights  \\\n",
       "count  48895.000000  48895.000000  48895.000000    48895.000000   \n",
       "mean      40.728949    -73.952170    152.720687        7.029962   \n",
       "std        0.054530      0.046157    240.154170       20.510550   \n",
       "min       40.499790    -74.244420      0.000000        1.000000   \n",
       "25%       40.690100    -73.983070     69.000000        1.000000   \n",
       "50%       40.723070    -73.955680    106.000000        3.000000   \n",
       "75%       40.763115    -73.936275    175.000000        5.000000   \n",
       "max       40.913060    -73.712990  10000.000000     1250.000000   \n",
       "\n",
       "       number_of_reviews  reviews_per_month  calculated_host_listings_count  \\\n",
       "count       48895.000000       48895.000000                    48895.000000   \n",
       "mean           23.274466           1.090910                        7.143982   \n",
       "std            44.550582           1.597283                       32.952519   \n",
       "min             0.000000           0.000000                        1.000000   \n",
       "25%             1.000000           0.040000                        1.000000   \n",
       "50%             5.000000           0.370000                        1.000000   \n",
       "75%            24.000000           1.580000                        2.000000   \n",
       "max           629.000000          58.500000                      327.000000   \n",
       "\n",
       "       availability_365  \n",
       "count      48895.000000  \n",
       "mean         112.781327  \n",
       "std          131.622289  \n",
       "min            0.000000  \n",
       "25%            0.000000  \n",
       "50%           45.000000  \n",
       "75%          227.000000  \n",
       "max          365.000000  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "neighbourhood_group               0\n",
       "latitude                          0\n",
       "longitude                         0\n",
       "room_type                         0\n",
       "price                             0\n",
       "minimum_nights                    0\n",
       "number_of_reviews                 0\n",
       "reviews_per_month                 0\n",
       "calculated_host_listings_count    0\n",
       "availability_365                  0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, use `DictVectorizer` to turn train and validation into matrices:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "#from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.feature_extraction import DictVectorizer\n",
    "from sklearn.metrics import roc_auc_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dicts = df_train.to_dict(orient = 'records')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'neighbourhood_group': 'Queens',\n",
       "  'latitude': 40.76434,\n",
       "  'longitude': -73.92132,\n",
       "  'room_type': 'Entire home/apt',\n",
       "  'minimum_nights': 4,\n",
       "  'number_of_reviews': 30,\n",
       "  'reviews_per_month': 0.32,\n",
       "  'calculated_host_listings_count': 1,\n",
       "  'availability_365': 363},\n",
       " {'neighbourhood_group': 'Brooklyn',\n",
       "  'latitude': 40.73442,\n",
       "  'longitude': -73.95854,\n",
       "  'room_type': 'Private room',\n",
       "  'minimum_nights': 5,\n",
       "  'number_of_reviews': 2,\n",
       "  'reviews_per_month': 0.16,\n",
       "  'calculated_host_listings_count': 1,\n",
       "  'availability_365': 0}]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dicts[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "dv = DictVectorizer(sparse=False)\n",
    "X_train = dv.fit_transform(train_dicts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[363.     ,   1.     ,  40.76434, ...,   1.     ,   0.     ,\n",
       "          0.     ],\n",
       "       [  0.     ,   1.     ,  40.73442, ...,   0.     ,   1.     ,\n",
       "          0.     ],\n",
       "       [113.     ,   5.     ,  40.66359, ...,   1.     ,   0.     ,\n",
       "          0.     ],\n",
       "       ...,\n",
       "       [178.     ,   2.     ,  40.69034, ...,   0.     ,   1.     ,\n",
       "          0.     ],\n",
       "       [  0.     ,   1.     ,  40.76324, ...,   1.     ,   0.     ,\n",
       "          0.     ],\n",
       "       [362.     ,  23.     ,  40.74563, ...,   1.     ,   0.     ,\n",
       "          0.     ]])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['availability_365',\n",
       " 'calculated_host_listings_count',\n",
       " 'latitude',\n",
       " 'longitude',\n",
       " 'minimum_nights',\n",
       " 'neighbourhood_group=Bronx',\n",
       " 'neighbourhood_group=Brooklyn',\n",
       " 'neighbourhood_group=Manhattan',\n",
       " 'neighbourhood_group=Queens',\n",
       " 'neighbourhood_group=Staten Island',\n",
       " 'number_of_reviews',\n",
       " 'reviews_per_month',\n",
       " 'room_type=Entire home/apt',\n",
       " 'room_type=Private room',\n",
       " 'room_type=Shared room']"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dv.get_feature_names()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question 1\n",
    "\n",
    "Let's train a decision tree regressor to predict the price variable. \n",
    "\n",
    "* Train a model with `max_depth=1`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeRegressor(max_depth=1)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt = DecisionTreeRegressor(max_depth=1)\n",
    "dt.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "val_dicts = df_val.to_dict(orient='records')\n",
    "X_val = dv.transform(val_dicts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train:  [5.14878079 4.29239363 5.14878079 ... 4.29239363 5.14878079 5.14878079]\n",
      "val:  [4.29239363 5.14878079 5.14878079 ... 5.14878079 5.14878079 5.14878079]\n"
     ]
    }
   ],
   "source": [
    "y_pred = dt.predict(X_train)\n",
    "#auc = roc_auc_score(y_train, y_pred)\n",
    "print('train: ', y_pred)\n",
    "\n",
    "y_pred = dt.predict(X_val)\n",
    "#auc = roc_auc_score(y_val, y_pred)\n",
    "print('val: ', y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4.29239363, 5.14878079, 5.14878079, ..., 5.14878079, 5.14878079,\n",
       "       5.14878079])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import export_text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "|--- room_type=Entire home/apt <= 0.50\n",
      "|   |--- value: [4.29]\n",
      "|--- room_type=Entire home/apt >  0.50\n",
      "|   |--- value: [5.15]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(export_text(dt, feature_names=dv.get_feature_names()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Which feature is used for splitting the data?\n",
    "\n",
    "* `room_type` (V)\n",
    "* `neighbourhood_group`\n",
    "* `number_of_reviews`\n",
    "* `reviews_per_month`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question 2\n",
    "\n",
    "Train a random forest model with these parameters:\n",
    "\n",
    "* `n_estimators=10`\n",
    "* `random_state=1`\n",
    "* `n_jobs=-1`  (optional - to make training faster)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(n_estimators=10, n_jobs=-1, random_state=1)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf = RandomForestRegressor(n_estimators=10, random_state=1, n_jobs=-1)\n",
    "rf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = rf.predict(X_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "rms = mean_squared_error(y_val, y_pred, squared=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.462"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "round(rms, 3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What's the RMSE of this model on validation?\n",
    "\n",
    "* 0.059\n",
    "* 0.259\n",
    "* 0.459 (V)\n",
    "* 0.659"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question 3\n",
    "\n",
    "Now let's experiment with the `n_estimators` parameter\n",
    "\n",
    "* Try different values of this parameter from 10 to 200 with step 10\n",
    "* Set `random_state` to `1`\n",
    "* Evaluate the model on the validation dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "scores = []\n",
    "\n",
    "for n in range(10, 201, 10):\n",
    "    rf = RandomForestRegressor(n_estimators = n, random_state=1)\n",
    "    rf.fit(X_train, y_train)\n",
    "    \n",
    "    y_pred = rf.predict(X_val)\n",
    "    rms = mean_squared_error(y_val, y_pred, squared=False)\n",
    "    scores.append((n, rms))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_scores = pd.DataFrame(scores, columns = ['n_estimators', 'rms'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>n_estimators</th>\n",
       "      <th>rms</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>0.461563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>20</td>\n",
       "      <td>0.448184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>30</td>\n",
       "      <td>0.445535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>40</td>\n",
       "      <td>0.443637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>50</td>\n",
       "      <td>0.442325</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>60</td>\n",
       "      <td>0.441639</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>70</td>\n",
       "      <td>0.441218</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>80</td>\n",
       "      <td>0.441106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>90</td>\n",
       "      <td>0.440533</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>100</td>\n",
       "      <td>0.439958</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>110</td>\n",
       "      <td>0.439459</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>120</td>\n",
       "      <td>0.439223</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>130</td>\n",
       "      <td>0.439290</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>140</td>\n",
       "      <td>0.438997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>150</td>\n",
       "      <td>0.438890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>160</td>\n",
       "      <td>0.438690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>170</td>\n",
       "      <td>0.438617</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>180</td>\n",
       "      <td>0.438756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>190</td>\n",
       "      <td>0.438715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>200</td>\n",
       "      <td>0.438742</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    n_estimators       rms\n",
       "0             10  0.461563\n",
       "1             20  0.448184\n",
       "2             30  0.445535\n",
       "3             40  0.443637\n",
       "4             50  0.442325\n",
       "5             60  0.441639\n",
       "6             70  0.441218\n",
       "7             80  0.441106\n",
       "8             90  0.440533\n",
       "9            100  0.439958\n",
       "10           110  0.439459\n",
       "11           120  0.439223\n",
       "12           130  0.439290\n",
       "13           140  0.438997\n",
       "14           150  0.438890\n",
       "15           160  0.438690\n",
       "16           170  0.438617\n",
       "17           180  0.438756\n",
       "18           190  0.438715\n",
       "19           200  0.438742"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f9373f03670>]"
      ]
     },
     "execution_count": 33,
     "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.plot(df_scores.n_estimators, df_scores.rms)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After which value of `n_estimators` does RMSE stop improving?\n",
    "\n",
    "- 10\n",
    "- 50\n",
    "- 70\n",
    "- 120 (V)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question 4\n",
    "\n",
    "Let's select the best `max_depth`:\n",
    "\n",
    "* Try different values of `max_depth`: `[10, 15, 20, 25]`\n",
    "* For each of these values, try different values of `n_estimators` from 10 till 200 (with step 10)\n",
    "* Fix the random seed: `random_state=1`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "scores = []\n",
    "\n",
    "for d in [10, 15, 20, 25]:\n",
    "    for n in range(10, 201, 10):\n",
    "        rf = RandomForestRegressor(n_estimators = n, \n",
    "                                    max_depth = d, \n",
    "                                    random_state=1)\n",
    "        rf.fit(X_train, y_train)\n",
    "    \n",
    "        y_pred = rf.predict(X_val)\n",
    "        rms = mean_squared_error(y_val, y_pred, squared=False)\n",
    "        scores.append((d, n, rms))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = ['max_depth', 'n_estimators', 'rms']\n",
    "\n",
    "df_scores = pd.DataFrame(scores, columns=columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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>max_depth</th>\n",
       "      <th>n_estimators</th>\n",
       "      <th>rms</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>0.445553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10</td>\n",
       "      <td>20</td>\n",
       "      <td>0.442009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10</td>\n",
       "      <td>30</td>\n",
       "      <td>0.441381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>40</td>\n",
       "      <td>0.441472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10</td>\n",
       "      <td>50</td>\n",
       "      <td>0.441067</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   max_depth  n_estimators       rms\n",
       "0         10            10  0.445553\n",
       "1         10            20  0.442009\n",
       "2         10            30  0.441381\n",
       "3         10            40  0.441472\n",
       "4         10            50  0.441067"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_scores.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f9374199f40>]"
      ]
     },
     "execution_count": 38,
     "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.plot(df_scores.n_estimators, df_scores.rms)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f93736e87f0>"
      ]
     },
     "execution_count": 40,
     "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": [
    "for d in [10, 15, 20, 25]:\n",
    "    df_subset = df_scores[df_scores.max_depth == d]\n",
    "    plt.plot(df_subset.n_estimators, df_subset.rms, label = 'max_depth=%d' % d)\n",
    "    \n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What's the best `max_depth`:\n",
    "\n",
    "* 10\n",
    "* 15\n",
    "* 20\n",
    "* 25\n",
    "\n",
    "Bonus question (not graded):\n",
    "\n",
    "Will the answer be different if we change the seed for the model? (Yes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question 5\n",
    "\n",
    "We can extract feature importance information from tree-based models. \n",
    "\n",
    "At each step of the decision tree learning algorith, it finds the best split. \n",
    "When doint it, we can calculate \"gain\" - the reduction in impurity before and after the split. \n",
    "This gain is quite useful in understanding what are the imporatant features \n",
    "for tree-based models.\n",
    "\n",
    "In Scikit-Learn, tree-based models contain this information in the `feature_importances_` field. \n",
    "\n",
    "For this homework question, we'll find the most important feature:\n",
    "\n",
    "* Train the model with these parametes:\n",
    "    * `n_estimators=10`,\n",
    "    * `max_depth=20`,\n",
    "    * `random_state=1`,\n",
    "    * `n_jobs=-1` (optional)\n",
    "* Get the feature importance information from this model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What's the most important feature? \n",
    "\n",
    "* `neighbourhood_group=Manhattan`\n",
    "* `room_type=Entire home/apt`\t\n",
    "* `longitude`\n",
    "* `latitude`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question 6"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let's train an XGBoost model! For this question, we'll tune the `eta` parameter\n",
    "\n",
    "* Install XGBoost\n",
    "* Create DMatrix for train and validation\n",
    "* Create a watchlist\n",
    "* Train a model with these parameters for 100 rounds:\n",
    "\n",
    "```\n",
    "xgb_params = {\n",
    "    'eta': 0.3, \n",
    "    'max_depth': 6,\n",
    "    'min_child_weight': 1,\n",
    "    \n",
    "    'objective': 'reg:squarederror',\n",
    "    'nthread': 8,\n",
    "    \n",
    "    'seed': 1,\n",
    "    'verbosity': 1,\n",
    "}\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now change `eta` first to `0.1` and then to `0.01`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Which eta leads to the best RMSE score on the validation dataset?\n",
    "\n",
    "* 0.3\n",
    "* 0.1\n",
    "* 0.01"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Submit the results\n",
    "\n",
    "\n",
    "Submit your results here: https://forms.gle/wQgFkYE6CtdDed4w8\n",
    "\n",
    "It's possible that your answers won't match exactly. If it's the case, select the closest one.\n",
    "\n",
    "\n",
    "## Deadline\n",
    "\n",
    "\n",
    "The deadline for submitting is 20 October 2021, 17:00 CET (Wednesday). After that, the form will be closed.\n",
    "\n"
   ]
  }
 ],
 "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}