HW/HW6/HW6-decision_trees.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-JrenrSGmAeg"
   },
   "source": [
    "**HOMEWORK**  \n",
    "\n",
    "The goal of this homework is to create a regression model for predicting housing prices (column 'median_house_value').  \n",
    "\n",
    "In this homework we'll again use the California Housing Prices dataset - the same one we used in homework 2 and 3.\n",
    "\n",
    "You can take it from [Kaggle](https://www.kaggle.com/datasets/camnugent/california-housing-prices) or download using wget link mentioned below:\n",
    "\n",
    "```\n",
    "wget https://raw.githubusercontent.com/alexeygrigorev/datasets/master/housing.csv\n",
    "\n",
    "```\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "nLosiCNJUjDM"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Mona\\anaconda3\\envs\\mlzoomcamp\\lib\\site-packages\\xgboost\\compat.py:36: FutureWarning: pandas.Int64Index is deprecated and will be removed from pandas in a future version. Use pandas.Index with the appropriate dtype instead.\n",
      "  from pandas import MultiIndex, Int64Index\n",
      "Intel(R) Extension for Scikit-learn* enabled (https://github.com/intel/scikit-learn-intelex)\n"
     ]
    }
   ],
   "source": [
    "#@ IMPORTING LIBRARIES AND DEPENDENCIES:\n",
    "import wget\n",
    "import re\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import xgboost as xgb\n",
    "import matplotlib.pyplot as plt\n",
    "from tqdm.auto import tqdm\n",
    "# from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.tree import export_text\n",
    "from sklearn.feature_extraction import DictVectorizer\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "from sklearnex import patch_sklearn, config_context\n",
    "patch_sklearn()\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "4Psk224nnyIF"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100% [..........................................................................] 1423529 / 1423529"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'housing.csv'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#@ DOWNLOADING THE DATASET: UNCOMMENT BELOW:\n",
    "wget.download(\"https://raw.githubusercontent.com/alexeygrigorev/datasets/master/housing.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "492Hn6-3n08G"
   },
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>longitude</th>\n",
       "      <td>-122.23</td>\n",
       "      <td>-122.22</td>\n",
       "      <td>-122.24</td>\n",
       "      <td>-122.25</td>\n",
       "      <td>-122.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>latitude</th>\n",
       "      <td>37.88</td>\n",
       "      <td>37.86</td>\n",
       "      <td>37.85</td>\n",
       "      <td>37.85</td>\n",
       "      <td>37.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>housing_median_age</th>\n",
       "      <td>41.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>52.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_rooms</th>\n",
       "      <td>880.0</td>\n",
       "      <td>7099.0</td>\n",
       "      <td>1467.0</td>\n",
       "      <td>1274.0</td>\n",
       "      <td>1627.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_bedrooms</th>\n",
       "      <td>129.0</td>\n",
       "      <td>1106.0</td>\n",
       "      <td>190.0</td>\n",
       "      <td>235.0</td>\n",
       "      <td>280.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>population</th>\n",
       "      <td>322.0</td>\n",
       "      <td>2401.0</td>\n",
       "      <td>496.0</td>\n",
       "      <td>558.0</td>\n",
       "      <td>565.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>households</th>\n",
       "      <td>126.0</td>\n",
       "      <td>1138.0</td>\n",
       "      <td>177.0</td>\n",
       "      <td>219.0</td>\n",
       "      <td>259.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>median_income</th>\n",
       "      <td>8.3252</td>\n",
       "      <td>8.3014</td>\n",
       "      <td>7.2574</td>\n",
       "      <td>5.6431</td>\n",
       "      <td>3.8462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>median_house_value</th>\n",
       "      <td>452600.0</td>\n",
       "      <td>358500.0</td>\n",
       "      <td>352100.0</td>\n",
       "      <td>341300.0</td>\n",
       "      <td>342200.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ocean_proximity</th>\n",
       "      <td>NEAR BAY</td>\n",
       "      <td>NEAR BAY</td>\n",
       "      <td>NEAR BAY</td>\n",
       "      <td>NEAR BAY</td>\n",
       "      <td>NEAR BAY</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           0         1         2         3         4\n",
       "longitude            -122.23   -122.22   -122.24   -122.25   -122.25\n",
       "latitude               37.88     37.86     37.85     37.85     37.85\n",
       "housing_median_age      41.0      21.0      52.0      52.0      52.0\n",
       "total_rooms            880.0    7099.0    1467.0    1274.0    1627.0\n",
       "total_bedrooms         129.0    1106.0     190.0     235.0     280.0\n",
       "population             322.0    2401.0     496.0     558.0     565.0\n",
       "households             126.0    1138.0     177.0     219.0     259.0\n",
       "median_income         8.3252    8.3014    7.2574    5.6431    3.8462\n",
       "median_house_value  452600.0  358500.0  352100.0  341300.0  342200.0\n",
       "ocean_proximity     NEAR BAY  NEAR BAY  NEAR BAY  NEAR BAY  NEAR BAY"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#@ READING DATASET:\n",
    "PATH = \"./housing.csv\"\n",
    "select_cols = [\"longitude\", \"latitude\", \"housing_median_age\", \"total_rooms\", \"total_bedrooms\", \"population\", \"households\", \n",
    "               \"median_income\", \"median_house_value\", \"ocean_proximity\"]\n",
    "df = pd.read_csv(PATH, usecols=select_cols)\n",
    "df.total_bedrooms = df.total_bedrooms.fillna(0)\n",
    "df.head().T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "longitude             float64\n",
       "latitude              float64\n",
       "housing_median_age    float64\n",
       "total_rooms           float64\n",
       "total_bedrooms        float64\n",
       "population            float64\n",
       "households            float64\n",
       "median_income         float64\n",
       "median_house_value    float64\n",
       "ocean_proximity        object\n",
       "dtype: object"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>longitude</th>\n",
       "      <td>-122.23</td>\n",
       "      <td>-122.22</td>\n",
       "      <td>-122.24</td>\n",
       "      <td>-122.25</td>\n",
       "      <td>-122.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>latitude</th>\n",
       "      <td>37.88</td>\n",
       "      <td>37.86</td>\n",
       "      <td>37.85</td>\n",
       "      <td>37.85</td>\n",
       "      <td>37.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>housing_median_age</th>\n",
       "      <td>41.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>52.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_rooms</th>\n",
       "      <td>880.0</td>\n",
       "      <td>7099.0</td>\n",
       "      <td>1467.0</td>\n",
       "      <td>1274.0</td>\n",
       "      <td>1627.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_bedrooms</th>\n",
       "      <td>129.0</td>\n",
       "      <td>1106.0</td>\n",
       "      <td>190.0</td>\n",
       "      <td>235.0</td>\n",
       "      <td>280.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>population</th>\n",
       "      <td>322.0</td>\n",
       "      <td>2401.0</td>\n",
       "      <td>496.0</td>\n",
       "      <td>558.0</td>\n",
       "      <td>565.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>households</th>\n",
       "      <td>126.0</td>\n",
       "      <td>1138.0</td>\n",
       "      <td>177.0</td>\n",
       "      <td>219.0</td>\n",
       "      <td>259.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>median_income</th>\n",
       "      <td>8.3252</td>\n",
       "      <td>8.3014</td>\n",
       "      <td>7.2574</td>\n",
       "      <td>5.6431</td>\n",
       "      <td>3.8462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>median_house_value</th>\n",
       "      <td>452600.0</td>\n",
       "      <td>358500.0</td>\n",
       "      <td>352100.0</td>\n",
       "      <td>341300.0</td>\n",
       "      <td>342200.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ocean_proximity</th>\n",
       "      <td>near_bay</td>\n",
       "      <td>near_bay</td>\n",
       "      <td>near_bay</td>\n",
       "      <td>near_bay</td>\n",
       "      <td>near_bay</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           0         1         2         3         4\n",
       "longitude            -122.23   -122.22   -122.24   -122.25   -122.25\n",
       "latitude               37.88     37.86     37.85     37.85     37.85\n",
       "housing_median_age      41.0      21.0      52.0      52.0      52.0\n",
       "total_rooms            880.0    7099.0    1467.0    1274.0    1627.0\n",
       "total_bedrooms         129.0    1106.0     190.0     235.0     280.0\n",
       "population             322.0    2401.0     496.0     558.0     565.0\n",
       "households             126.0    1138.0     177.0     219.0     259.0\n",
       "median_income         8.3252    8.3014    7.2574    5.6431    3.8462\n",
       "median_house_value  452600.0  358500.0  352100.0  341300.0  342200.0\n",
       "ocean_proximity     near_bay  near_bay  near_bay  near_bay  near_bay"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# for column names:\n",
    "\n",
    "df.columns = df.columns.str.lower().str.replace(' ', '_')\n",
    "\n",
    "# for data in the dataframe:\n",
    "\n",
    "strings = list(df.dtypes[df.dtypes == 'object'].index)\n",
    "for col in strings:\n",
    "    df[col] = df[col].str.lower().str.replace(' ', '_')\n",
    "\n",
    "df.head(5).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "longitude             0\n",
       "latitude              0\n",
       "housing_median_age    0\n",
       "total_rooms           0\n",
       "total_bedrooms        0\n",
       "population            0\n",
       "households            0\n",
       "median_income         0\n",
       "median_house_value    0\n",
       "ocean_proximity       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# checking if any missing values exist in any of the missing columns:\n",
    "df.isna().sum(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# column_names = df.columns[df.isnull().any()].to_list()\n",
    "\n",
    "# for column in column_names:\n",
    "#     df[column].fillna(0, inplace=True) \n",
    "    \n",
    "# df.isna().sum(axis=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "OqXppJ_Ro0H0"
   },
   "source": [
    "- Apply the log transform to `median_house_value`. \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": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        13.022766\n",
       "1        12.789687\n",
       "2        12.771673\n",
       "3        12.740520\n",
       "4        12.743154\n",
       "           ...    \n",
       "20635    11.265758\n",
       "20636    11.252872\n",
       "20637    11.432810\n",
       "20638    11.346883\n",
       "20639    11.400887\n",
       "Name: median_house_value, Length: 20640, dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features = df.drop(columns='median_house_value')\n",
    "features.head()\n",
    "target = df[\"median_house_value\"]\n",
    "target.head()\n",
    "log_target = np.log1p(target)\n",
    "log_target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "id": "W1MmdcRDpHS1"
   },
   "outputs": [],
   "source": [
    "#@ SPLITTING THE DATASET FOR TRAINING AND TEST:\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "\n",
    "x_train_full, x_test, y_train_full, y_test = train_test_split(features, log_target, test_size=0.2, random_state=1)\n",
    "x_train, x_val, y_train, y_val = train_test_split(x_train_full, y_train_full, test_size=0.25, random_state=1)\n",
    "\n",
    "splits = [x_train_full, x_train, x_val, x_test]\n",
    "for split in splits:\n",
    "        split.reset_index(inplace = True)\n",
    "\n",
    "splits = [y_train_full,y_train, y_val, y_test]\n",
    "for split in splits:\n",
    "         split.reset_index(drop = True, inplace = True)    \n",
    "       "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5VKq8FX7qIVM"
   },
   "source": [
    "- We will use `DictVectorizer` to turn train and validation into matrices."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "id": "WnfFrWe9qtuc"
   },
   "outputs": [],
   "source": [
    "#@ IMPLEMENTATION OF DICTVECTORIZER:\n",
    "\n",
    "dv = DictVectorizer(sparse=False)\n",
    "\n",
    "train_dict = x_train.to_dict(orient='records')\n",
    "x_train_OH = dv.fit_transform(train_dict)\n",
    "\n",
    "val_dict = x_val.to_dict(orient='records')\n",
    "x_val_OH = dv.transform(val_dict)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "qvxmhaD5hmAK"
   },
   "source": [
    "**Question 1**\n",
    "\n",
    "Let's train a decision tree regressor to predict the `median_house_value` variable.\n",
    "\n",
    "Train a model with `max_depth=1`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "id": "cqBcSbpIq7kJ"
   },
   "outputs": [],
   "source": [
    "#@ TRAINING THE REGRESSION MODEL:\n",
    "\n",
    "dt = DecisionTreeRegressor(max_depth=1)\n",
    "dt.fit(x_train_OH, y_train)\n",
    "y_pred = dt.predict(x_val_OH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "sLAnqRwuikqQ"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "|--- ocean_proximity=inland <= 0.50\n",
      "|   |--- value: [12.31]\n",
      "|--- ocean_proximity=inland >  0.50\n",
      "|   |--- value: [11.61]\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Mona\\anaconda3\\envs\\mlzoomcamp\\lib\\site-packages\\sklearn\\utils\\deprecation.py:87: FutureWarning: Function get_feature_names is deprecated; get_feature_names is deprecated in 1.0 and will be removed in 1.2. Please use get_feature_names_out instead.\n",
      "  warnings.warn(msg, category=FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "#@ INSPECTION:\n",
    "\n",
    "print(export_text(dt, feature_names=dv.get_feature_names()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "JNBm-qeRp6aZ"
   },
   "source": [
    "- Which feature is used for splitting the data?\n",
    "\n",
    "- Answer: **ocean_proximity=inland**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Xhnbdy-CqQoB"
   },
   "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": 14,
   "metadata": {
    "id": "2w-YSGYMisFt"
   },
   "outputs": [],
   "source": [
    "#@ TRAINING RANDOM FOREST MODEL:\n",
    "\n",
    "n_estimators=10  \n",
    "random_state=1  \n",
    "n_jobs=-1 \n",
    "\n",
    "rf = RandomForestRegressor(n_estimators=n_estimators,\n",
    "                            random_state=random_state,\n",
    "                           n_jobs = n_jobs)\n",
    "rf.fit(x_train_OH, y_train)\n",
    "y_pred = rf.predict(x_val_OH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "id": "jCxUAO2Rq1Ps"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.237"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#@ CALCULATING MEAN SQUARED ERROR:\n",
    "# squaredbool, default=True\n",
    "#    If True returns MSE value, if False returns RMSE value.\n",
    "\n",
    "round(mean_squared_error(y_val, y_pred, squared=False),3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jfcDllHxrTqu"
   },
   "source": [
    "- What's the RMSE of this model on validation?\n",
    "\n",
    "- Answer: **0.237**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Y7K69bFSrg_V"
   },
   "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": 17,
   "metadata": {
    "id": "-oUFCQwOrQNN"
   },
   "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>RMSE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>0.236755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>20</td>\n",
       "      <td>0.229005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>30</td>\n",
       "      <td>0.226680</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>40</td>\n",
       "      <td>0.226321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>50</td>\n",
       "      <td>0.226023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>60</td>\n",
       "      <td>0.225379</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>70</td>\n",
       "      <td>0.224517</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>80</td>\n",
       "      <td>0.224515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>90</td>\n",
       "      <td>0.224199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>100</td>\n",
       "      <td>0.224151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>110</td>\n",
       "      <td>0.223843</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>120</td>\n",
       "      <td>0.223798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>130</td>\n",
       "      <td>0.223836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>140</td>\n",
       "      <td>0.223845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>150</td>\n",
       "      <td>0.223700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>160</td>\n",
       "      <td>0.223788</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>170</td>\n",
       "      <td>0.223752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>180</td>\n",
       "      <td>0.224003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>190</td>\n",
       "      <td>0.224002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>200</td>\n",
       "      <td>0.223990</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    n_estimators      RMSE\n",
       "0             10  0.236755\n",
       "1             20  0.229005\n",
       "2             30  0.226680\n",
       "3             40  0.226321\n",
       "4             50  0.226023\n",
       "5             60  0.225379\n",
       "6             70  0.224517\n",
       "7             80  0.224515\n",
       "8             90  0.224199\n",
       "9            100  0.224151\n",
       "10           110  0.223843\n",
       "11           120  0.223798\n",
       "12           130  0.223836\n",
       "13           140  0.223845\n",
       "14           150  0.223700\n",
       "15           160  0.223788\n",
       "16           170  0.223752\n",
       "17           180  0.224003\n",
       "18           190  0.224002\n",
       "19           200  0.223990"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#@ TRAINING THE RANDOM FOREST MODEL:\n",
    "\n",
    "scores = []\n",
    "range_n_estimators = np.arange(10, 210, 10)\n",
    "\n",
    "for n_est in range_n_estimators:\n",
    "    \n",
    "        rf = RandomForestRegressor(n_estimators=n_est,\n",
    "                                    random_state=1,\n",
    "                                    n_jobs = -1)\n",
    "        rf.fit(x_train_OH, y_train)\n",
    "\n",
    "        y_pred = rf.predict(x_val_OH)\n",
    "        rmse = mean_squared_error(y_val, y_pred, squared=False)\n",
    "        scores.append((n_est,rmse))\n",
    "        \n",
    "df_scores = pd.DataFrame(scores, columns=['n_estimators', 'RMSE'])\n",
    "df_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "id": "3J-wFLh4s8Ky"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'RMSE')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#@ INSPECTING THE RMSE SCORES:\n",
    "\n",
    "plt.plot(df_scores.n_estimators, df_scores.RMSE)\n",
    "plt.xlabel(\"n_estimators\")\n",
    "plt.ylabel(\"RMSE\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "NAXZcW3AuC-d"
   },
   "source": [
    "- After which value of `n_estimators` does RMSE stop improving?\n",
    "\n",
    "- Answer: **150**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "CYhRv3kEvWjK"
   },
   "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": 19,
   "metadata": {
    "id": "fO9TackKvKcS"
   },
   "outputs": [],
   "source": [
    "#@ TRAINING THE MODEL WITH DEPTH:\n",
    "\n",
    "scores = []\n",
    "max_depth = [10, 15, 20, 25]\n",
    "range_n_estimators = np.arange(10, 210, 10)\n",
    "\n",
    "for d in max_depth:\n",
    "    for n_est in range_n_estimators:\n",
    "\n",
    "            rf = RandomForestRegressor(n_estimators=n_est,\n",
    "                                        max_depth= d,\n",
    "                                        random_state=1,\n",
    "                                        n_jobs = -1)\n",
    "            rf.fit(x_train_OH, y_train)\n",
    "\n",
    "            y_pred = rf.predict(x_val_OH)\n",
    "            rmse = mean_squared_error(y_val, y_pred, squared=False)\n",
    "\n",
    "            scores.append((d, n_est,rmse))\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>RMSE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>0.253136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10</td>\n",
       "      <td>20</td>\n",
       "      <td>0.250160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10</td>\n",
       "      <td>30</td>\n",
       "      <td>0.250099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>40</td>\n",
       "      <td>0.249999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10</td>\n",
       "      <td>50</td>\n",
       "      <td>0.250147</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>25</td>\n",
       "      <td>160</td>\n",
       "      <td>0.223528</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>25</td>\n",
       "      <td>170</td>\n",
       "      <td>0.223510</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>25</td>\n",
       "      <td>180</td>\n",
       "      <td>0.223809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>25</td>\n",
       "      <td>190</td>\n",
       "      <td>0.223822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>25</td>\n",
       "      <td>200</td>\n",
       "      <td>0.223896</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>80 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    max_depth  n_estimators      RMSE\n",
       "0          10            10  0.253136\n",
       "1          10            20  0.250160\n",
       "2          10            30  0.250099\n",
       "3          10            40  0.249999\n",
       "4          10            50  0.250147\n",
       "..        ...           ...       ...\n",
       "75         25           160  0.223528\n",
       "76         25           170  0.223510\n",
       "77         25           180  0.223809\n",
       "78         25           190  0.223822\n",
       "79         25           200  0.223896\n",
       "\n",
       "[80 rows x 3 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "columns = ['max_depth', 'n_estimators', 'RMSE']\n",
    "df_scores = pd.DataFrame(scores, columns=columns)\n",
    "df_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x22af8a80d30>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for d in max_depth:\n",
    "    df_subset = df_scores[df_scores.max_depth == d]\n",
    "    \n",
    "    plt.plot(df_subset.n_estimators, df_subset.RMSE,\n",
    "             label='max_depth=%d' % d)\n",
    "    \n",
    "    \n",
    "plt.xlabel(\"n_estimators\")\n",
    "plt.ylabel(\"RMSE\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "eCRCaKlSwp7D"
   },
   "source": [
    "- What's the best `max_depth`:\n",
    "\n",
    "- Answer: **25**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BDLbx3N5xAoY"
   },
   "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. When doint it, we can calculate \"gain\" - the reduction in impurity before and after the split. This gain is quite useful in understanding what are the imporatant features 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",
    "\n",
    "Get the feature importance information from this model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "id": "UICqST2Qwhbj"
   },
   "outputs": [],
   "source": [
    "#@ TRAINING THE RANDOM FOREST MODEL:\n",
    "\n",
    "n_estimators=10\n",
    "max_depth=20\n",
    "random_state=1  \n",
    "n_jobs=-1 \n",
    "\n",
    "rf = RandomForestRegressor(n_estimators=n_estimators,\n",
    "                           max_depth=max_depth,\n",
    "                           random_state=random_state,\n",
    "                           n_jobs = n_jobs)\n",
    "rf.fit(x_train_OH, y_train)\n",
    "\n",
    "importances = rf.feature_importances_\n",
    "std = np.std([tree.feature_importances_ for tree in rf.estimators_], axis=0)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Mona\\anaconda3\\envs\\mlzoomcamp\\lib\\site-packages\\sklearn\\utils\\deprecation.py:87: FutureWarning: Function get_feature_names is deprecated; get_feature_names is deprecated in 1.0 and will be removed in 1.2. Please use get_feature_names_out instead.\n",
      "  warnings.warn(msg, category=FutureWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "forest_importances = pd.Series(importances, index= dv.get_feature_names())\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "forest_importances.plot.bar(yerr=std, ax=ax)\n",
    "ax.set_title(\"Feature importances using MDI\")\n",
    "ax.set_ylabel(\"Mean decrease in impurity\")\n",
    "fig.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kSCqlx9jye1s"
   },
   "source": [
    "- What's the most important feature?\n",
    "\n",
    "- Answer: **median_income**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "K5KEdiTMzPaI"
   },
   "source": [
    "**Question 6**\n",
    "\n",
    "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",
    "```\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "id": "TUP0uoL5yZTc"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['households',\n",
       " 'housing_median_age',\n",
       " 'index',\n",
       " 'latitude',\n",
       " 'longitude',\n",
       " 'median_income',\n",
       " 'ocean_proximity=<1h_ocean',\n",
       " 'ocean_proximity=inland',\n",
       " 'ocean_proximity=island',\n",
       " 'ocean_proximity=near_bay',\n",
       " 'ocean_proximity=near_ocean',\n",
       " 'population',\n",
       " 'total_bedrooms',\n",
       " 'total_rooms']"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#@ CREATING THE DMARTIX:\n",
    "features = dv.feature_names_\n",
    "features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ValueError: feature_names must be string, and may not contain [, ] or <"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['households',\n",
       " 'housing_median_age',\n",
       " 'index',\n",
       " 'latitude',\n",
       " 'longitude',\n",
       " 'median_income',\n",
       " 'ocean_proximity_1h_ocean',\n",
       " 'ocean_proximity=inland',\n",
       " 'ocean_proximity=island',\n",
       " 'ocean_proximity=near_bay',\n",
       " 'ocean_proximity=near_ocean',\n",
       " 'population',\n",
       " 'total_bedrooms',\n",
       " 'total_rooms']"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regex = re.compile(r\"=<|\\[|\\]\", re.IGNORECASE)\n",
    "features = [regex.sub(\"_\", col) if any(x in str(col) for x in set(('=','<','[', ']'))) else col for col in features]\n",
    "features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "dtrain = xgb.DMatrix(x_train_OH, label=y_train, feature_names=features)\n",
    "dval = xgb.DMatrix(x_val_OH, label=y_val, feature_names=features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "id": "lRRzLznP-Z-Q"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\ttrain-rmse:8.12733\teval-rmse:8.13245\n",
      "[1]\ttrain-rmse:5.69586\teval-rmse:5.70070\n",
      "[2]\ttrain-rmse:3.99534\teval-rmse:3.99961\n",
      "[3]\ttrain-rmse:2.80673\teval-rmse:2.81054\n",
      "[4]\ttrain-rmse:1.97720\teval-rmse:1.98060\n",
      "[5]\ttrain-rmse:1.40055\teval-rmse:1.40602\n",
      "[6]\ttrain-rmse:0.99978\teval-rmse:1.00616\n",
      "[7]\ttrain-rmse:0.72601\teval-rmse:0.73398\n",
      "[8]\ttrain-rmse:0.54258\teval-rmse:0.55300\n",
      "[9]\ttrain-rmse:0.41736\teval-rmse:0.42933\n",
      "[10]\ttrain-rmse:0.33893\teval-rmse:0.35412\n",
      "[11]\ttrain-rmse:0.29138\teval-rmse:0.30958\n",
      "[12]\ttrain-rmse:0.26111\teval-rmse:0.28264\n",
      "[13]\ttrain-rmse:0.24329\teval-rmse:0.26688\n",
      "[14]\ttrain-rmse:0.23230\teval-rmse:0.25818\n",
      "[15]\ttrain-rmse:0.22253\teval-rmse:0.25095\n",
      "[16]\ttrain-rmse:0.21892\teval-rmse:0.24870\n",
      "[17]\ttrain-rmse:0.21494\teval-rmse:0.24610\n",
      "[18]\ttrain-rmse:0.21136\teval-rmse:0.24455\n",
      "[19]\ttrain-rmse:0.20852\teval-rmse:0.24297\n",
      "[20]\ttrain-rmse:0.20725\teval-rmse:0.24240\n",
      "[21]\ttrain-rmse:0.20616\teval-rmse:0.24173\n",
      "[22]\ttrain-rmse:0.20181\teval-rmse:0.23875\n",
      "[23]\ttrain-rmse:0.20098\teval-rmse:0.23887\n",
      "[24]\ttrain-rmse:0.19913\teval-rmse:0.23777\n",
      "[25]\ttrain-rmse:0.19828\teval-rmse:0.23732\n",
      "[26]\ttrain-rmse:0.19547\teval-rmse:0.23543\n",
      "[27]\ttrain-rmse:0.19337\teval-rmse:0.23525\n",
      "[28]\ttrain-rmse:0.19215\teval-rmse:0.23495\n",
      "[29]\ttrain-rmse:0.19137\teval-rmse:0.23467\n",
      "[30]\ttrain-rmse:0.18889\teval-rmse:0.23401\n",
      "[31]\ttrain-rmse:0.18722\teval-rmse:0.23342\n",
      "[32]\ttrain-rmse:0.18665\teval-rmse:0.23336\n",
      "[33]\ttrain-rmse:0.18511\teval-rmse:0.23287\n",
      "[34]\ttrain-rmse:0.18246\teval-rmse:0.23168\n",
      "[35]\ttrain-rmse:0.18131\teval-rmse:0.23107\n",
      "[36]\ttrain-rmse:0.18000\teval-rmse:0.23086\n",
      "[37]\ttrain-rmse:0.17886\teval-rmse:0.23065\n",
      "[38]\ttrain-rmse:0.17822\teval-rmse:0.23049\n",
      "[39]\ttrain-rmse:0.17604\teval-rmse:0.22937\n",
      "[40]\ttrain-rmse:0.17554\teval-rmse:0.22904\n",
      "[41]\ttrain-rmse:0.17364\teval-rmse:0.22863\n",
      "[42]\ttrain-rmse:0.17284\teval-rmse:0.22842\n",
      "[43]\ttrain-rmse:0.17199\teval-rmse:0.22805\n",
      "[44]\ttrain-rmse:0.17164\teval-rmse:0.22794\n",
      "[45]\ttrain-rmse:0.17040\teval-rmse:0.22794\n",
      "[46]\ttrain-rmse:0.17009\teval-rmse:0.22789\n",
      "[47]\ttrain-rmse:0.16836\teval-rmse:0.22696\n",
      "[48]\ttrain-rmse:0.16722\teval-rmse:0.22661\n",
      "[49]\ttrain-rmse:0.16625\teval-rmse:0.22623\n",
      "[50]\ttrain-rmse:0.16567\teval-rmse:0.22609\n",
      "[51]\ttrain-rmse:0.16467\teval-rmse:0.22545\n",
      "[52]\ttrain-rmse:0.16422\teval-rmse:0.22544\n",
      "[53]\ttrain-rmse:0.16337\teval-rmse:0.22523\n",
      "[54]\ttrain-rmse:0.16192\teval-rmse:0.22492\n",
      "[55]\ttrain-rmse:0.16089\teval-rmse:0.22483\n",
      "[56]\ttrain-rmse:0.16038\teval-rmse:0.22461\n",
      "[57]\ttrain-rmse:0.15917\teval-rmse:0.22406\n",
      "[58]\ttrain-rmse:0.15804\teval-rmse:0.22429\n",
      "[59]\ttrain-rmse:0.15660\teval-rmse:0.22377\n",
      "[60]\ttrain-rmse:0.15611\teval-rmse:0.22364\n",
      "[61]\ttrain-rmse:0.15568\teval-rmse:0.22349\n",
      "[62]\ttrain-rmse:0.15535\teval-rmse:0.22355\n",
      "[63]\ttrain-rmse:0.15432\teval-rmse:0.22370\n",
      "[64]\ttrain-rmse:0.15343\teval-rmse:0.22350\n",
      "[65]\ttrain-rmse:0.15250\teval-rmse:0.22325\n",
      "[66]\ttrain-rmse:0.15158\teval-rmse:0.22300\n",
      "[67]\ttrain-rmse:0.15137\teval-rmse:0.22297\n",
      "[68]\ttrain-rmse:0.15026\teval-rmse:0.22281\n",
      "[69]\ttrain-rmse:0.14945\teval-rmse:0.22275\n",
      "[70]\ttrain-rmse:0.14881\teval-rmse:0.22254\n",
      "[71]\ttrain-rmse:0.14761\teval-rmse:0.22221\n",
      "[72]\ttrain-rmse:0.14704\teval-rmse:0.22200\n",
      "[73]\ttrain-rmse:0.14672\teval-rmse:0.22201\n",
      "[74]\ttrain-rmse:0.14615\teval-rmse:0.22183\n",
      "[75]\ttrain-rmse:0.14556\teval-rmse:0.22176\n",
      "[76]\ttrain-rmse:0.14512\teval-rmse:0.22168\n",
      "[77]\ttrain-rmse:0.14428\teval-rmse:0.22136\n",
      "[78]\ttrain-rmse:0.14357\teval-rmse:0.22128\n",
      "[79]\ttrain-rmse:0.14244\teval-rmse:0.22098\n",
      "[80]\ttrain-rmse:0.14162\teval-rmse:0.22105\n",
      "[81]\ttrain-rmse:0.14095\teval-rmse:0.22096\n",
      "[82]\ttrain-rmse:0.14047\teval-rmse:0.22089\n",
      "[83]\ttrain-rmse:0.13931\teval-rmse:0.22035\n",
      "[84]\ttrain-rmse:0.13850\teval-rmse:0.22062\n",
      "[85]\ttrain-rmse:0.13765\teval-rmse:0.22039\n",
      "[86]\ttrain-rmse:0.13731\teval-rmse:0.22026\n",
      "[87]\ttrain-rmse:0.13670\teval-rmse:0.22008\n",
      "[88]\ttrain-rmse:0.13608\teval-rmse:0.22007\n",
      "[89]\ttrain-rmse:0.13528\teval-rmse:0.21997\n",
      "[90]\ttrain-rmse:0.13511\teval-rmse:0.22002\n",
      "[91]\ttrain-rmse:0.13445\teval-rmse:0.21998\n",
      "[92]\ttrain-rmse:0.13395\teval-rmse:0.21987\n",
      "[93]\ttrain-rmse:0.13325\teval-rmse:0.21980\n",
      "[94]\ttrain-rmse:0.13285\teval-rmse:0.21975\n",
      "[95]\ttrain-rmse:0.13267\teval-rmse:0.21979\n",
      "[96]\ttrain-rmse:0.13184\teval-rmse:0.21961\n",
      "[97]\ttrain-rmse:0.13104\teval-rmse:0.21955\n",
      "[98]\ttrain-rmse:0.13049\teval-rmse:0.21959\n",
      "[99]\ttrain-rmse:0.13001\teval-rmse:0.21954\n"
     ]
    }
   ],
   "source": [
    "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",
    "\n",
    "\n",
    "# xgb_param['eval_metric'] = 'rmse'\n",
    "watchlist = [(dtrain, 'train'), (dval, 'eval')]\n",
    "\n",
    "num_round = 100\n",
    "bst = xgb.train(xgb_params, dtrain, num_round, watchlist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2195412561585616"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = bst.predict(dval)\n",
    "rmse = mean_squared_error(y_val,y_pred, squared= False)\n",
    "rmse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "3RmcQ1BQ64X-"
   },
   "source": [
    "- Now, change eta first to 0.1 and then to 0.01."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training with eta 0.100:\n",
      "\n",
      "\n",
      "[0]\ttrain-rmse:10.44371\teval-rmse:10.44938\n",
      "[1]\ttrain-rmse:9.40118\teval-rmse:9.40684\n",
      "[2]\ttrain-rmse:8.46290\teval-rmse:8.46831\n",
      "[3]\ttrain-rmse:7.61859\teval-rmse:7.62391\n",
      "[4]\ttrain-rmse:6.85878\teval-rmse:6.86368\n",
      "[5]\ttrain-rmse:6.17513\teval-rmse:6.17987\n",
      "[6]\ttrain-rmse:5.55992\teval-rmse:5.56416\n",
      "[7]\ttrain-rmse:5.00639\teval-rmse:5.01018\n",
      "[8]\ttrain-rmse:4.50838\teval-rmse:4.51201\n",
      "[9]\ttrain-rmse:4.06030\teval-rmse:4.06384\n",
      "[10]\ttrain-rmse:3.65719\teval-rmse:3.66051\n",
      "[11]\ttrain-rmse:3.29460\teval-rmse:3.29767\n",
      "[12]\ttrain-rmse:2.96847\teval-rmse:2.97146\n",
      "[13]\ttrain-rmse:2.67518\teval-rmse:2.67816\n",
      "[14]\ttrain-rmse:2.41155\teval-rmse:2.41423\n",
      "[15]\ttrain-rmse:2.17452\teval-rmse:2.17720\n",
      "[16]\ttrain-rmse:1.96127\teval-rmse:1.96403\n",
      "[17]\ttrain-rmse:1.76988\teval-rmse:1.77244\n",
      "[18]\ttrain-rmse:1.59791\teval-rmse:1.60080\n",
      "[19]\ttrain-rmse:1.44355\teval-rmse:1.44665\n",
      "[20]\ttrain-rmse:1.30443\teval-rmse:1.30770\n",
      "[21]\ttrain-rmse:1.18002\teval-rmse:1.18332\n",
      "[22]\ttrain-rmse:1.06853\teval-rmse:1.07243\n",
      "[23]\ttrain-rmse:0.96805\teval-rmse:0.97228\n",
      "[24]\ttrain-rmse:0.87809\teval-rmse:0.88264\n",
      "[25]\ttrain-rmse:0.79814\teval-rmse:0.80317\n",
      "[26]\ttrain-rmse:0.72627\teval-rmse:0.73184\n",
      "[27]\ttrain-rmse:0.66261\teval-rmse:0.66886\n",
      "[28]\ttrain-rmse:0.60592\teval-rmse:0.61294\n",
      "[29]\ttrain-rmse:0.55581\teval-rmse:0.56368\n",
      "[30]\ttrain-rmse:0.51050\teval-rmse:0.51933\n",
      "[31]\ttrain-rmse:0.47075\teval-rmse:0.48067\n",
      "[32]\ttrain-rmse:0.43624\teval-rmse:0.44725\n",
      "[33]\ttrain-rmse:0.40555\teval-rmse:0.41778\n",
      "[34]\ttrain-rmse:0.37907\teval-rmse:0.39231\n",
      "[35]\ttrain-rmse:0.35548\teval-rmse:0.37010\n",
      "[36]\ttrain-rmse:0.33547\teval-rmse:0.35109\n",
      "[37]\ttrain-rmse:0.31766\teval-rmse:0.33469\n",
      "[38]\ttrain-rmse:0.30172\teval-rmse:0.32001\n",
      "[39]\ttrain-rmse:0.28822\teval-rmse:0.30778\n",
      "[40]\ttrain-rmse:0.27602\teval-rmse:0.29663\n",
      "[41]\ttrain-rmse:0.26633\teval-rmse:0.28781\n",
      "[42]\ttrain-rmse:0.25824\teval-rmse:0.28062\n",
      "[43]\ttrain-rmse:0.25120\teval-rmse:0.27429\n",
      "[44]\ttrain-rmse:0.24483\teval-rmse:0.26910\n",
      "[45]\ttrain-rmse:0.23902\teval-rmse:0.26400\n",
      "[46]\ttrain-rmse:0.23453\teval-rmse:0.26033\n",
      "[47]\ttrain-rmse:0.23023\teval-rmse:0.25682\n",
      "[48]\ttrain-rmse:0.22628\teval-rmse:0.25357\n",
      "[49]\ttrain-rmse:0.22381\teval-rmse:0.25153\n",
      "[50]\ttrain-rmse:0.22083\teval-rmse:0.24904\n",
      "[51]\ttrain-rmse:0.21850\teval-rmse:0.24729\n",
      "[52]\ttrain-rmse:0.21632\teval-rmse:0.24591\n",
      "[53]\ttrain-rmse:0.21406\teval-rmse:0.24438\n",
      "[54]\ttrain-rmse:0.21241\teval-rmse:0.24328\n",
      "[55]\ttrain-rmse:0.21100\teval-rmse:0.24241\n",
      "[56]\ttrain-rmse:0.20962\teval-rmse:0.24187\n",
      "[57]\ttrain-rmse:0.20826\teval-rmse:0.24102\n",
      "[58]\ttrain-rmse:0.20685\teval-rmse:0.24021\n",
      "[59]\ttrain-rmse:0.20554\teval-rmse:0.23965\n",
      "[60]\ttrain-rmse:0.20427\teval-rmse:0.23890\n",
      "[61]\ttrain-rmse:0.20327\teval-rmse:0.23815\n",
      "[62]\ttrain-rmse:0.20217\teval-rmse:0.23754\n",
      "[63]\ttrain-rmse:0.20124\teval-rmse:0.23714\n",
      "[64]\ttrain-rmse:0.20073\teval-rmse:0.23677\n",
      "[65]\ttrain-rmse:0.20024\teval-rmse:0.23636\n",
      "[66]\ttrain-rmse:0.19928\teval-rmse:0.23572\n",
      "[67]\ttrain-rmse:0.19842\teval-rmse:0.23552\n",
      "[68]\ttrain-rmse:0.19727\teval-rmse:0.23482\n",
      "[69]\ttrain-rmse:0.19624\teval-rmse:0.23444\n",
      "[70]\ttrain-rmse:0.19590\teval-rmse:0.23425\n",
      "[71]\ttrain-rmse:0.19529\teval-rmse:0.23408\n",
      "[72]\ttrain-rmse:0.19471\teval-rmse:0.23396\n",
      "[73]\ttrain-rmse:0.19370\teval-rmse:0.23327\n",
      "[74]\ttrain-rmse:0.19282\teval-rmse:0.23300\n",
      "[75]\ttrain-rmse:0.19206\teval-rmse:0.23273\n",
      "[76]\ttrain-rmse:0.19159\teval-rmse:0.23255\n",
      "[77]\ttrain-rmse:0.19057\teval-rmse:0.23206\n",
      "[78]\ttrain-rmse:0.18969\teval-rmse:0.23206\n",
      "[79]\ttrain-rmse:0.18887\teval-rmse:0.23150\n",
      "[80]\ttrain-rmse:0.18833\teval-rmse:0.23130\n",
      "[81]\ttrain-rmse:0.18796\teval-rmse:0.23100\n",
      "[82]\ttrain-rmse:0.18708\teval-rmse:0.23050\n",
      "[83]\ttrain-rmse:0.18629\teval-rmse:0.23026\n",
      "[84]\ttrain-rmse:0.18580\teval-rmse:0.23014\n",
      "[85]\ttrain-rmse:0.18500\teval-rmse:0.22957\n",
      "[86]\ttrain-rmse:0.18450\teval-rmse:0.22953\n",
      "[87]\ttrain-rmse:0.18342\teval-rmse:0.22895\n",
      "[88]\ttrain-rmse:0.18300\teval-rmse:0.22891\n",
      "[89]\ttrain-rmse:0.18272\teval-rmse:0.22880\n",
      "[90]\ttrain-rmse:0.18231\teval-rmse:0.22871\n",
      "[91]\ttrain-rmse:0.18190\teval-rmse:0.22863\n",
      "[92]\ttrain-rmse:0.18100\teval-rmse:0.22816\n",
      "[93]\ttrain-rmse:0.18029\teval-rmse:0.22798\n",
      "[94]\ttrain-rmse:0.17992\teval-rmse:0.22792\n",
      "[95]\ttrain-rmse:0.17945\teval-rmse:0.22776\n",
      "[96]\ttrain-rmse:0.17888\teval-rmse:0.22746\n",
      "[97]\ttrain-rmse:0.17838\teval-rmse:0.22733\n",
      "[98]\ttrain-rmse:0.17785\teval-rmse:0.22710\n",
      "[99]\ttrain-rmse:0.17730\teval-rmse:0.22691\n",
      "\n",
      "\n",
      "\n",
      "Training with eta 0.010:\n",
      "\n",
      "\n",
      "[0]\ttrain-rmse:11.48634\teval-rmse:11.49225\n",
      "[1]\ttrain-rmse:11.37165\teval-rmse:11.37753\n",
      "[2]\ttrain-rmse:11.25810\teval-rmse:11.26397\n",
      "[3]\ttrain-rmse:11.14569\teval-rmse:11.15153\n",
      "[4]\ttrain-rmse:11.03441\teval-rmse:11.04022\n",
      "[5]\ttrain-rmse:10.92424\teval-rmse:10.93003\n",
      "[6]\ttrain-rmse:10.81517\teval-rmse:10.82094\n",
      "[7]\ttrain-rmse:10.70718\teval-rmse:10.71290\n",
      "[8]\ttrain-rmse:10.60029\teval-rmse:10.60598\n",
      "[9]\ttrain-rmse:10.49445\teval-rmse:10.50010\n",
      "[10]\ttrain-rmse:10.38969\teval-rmse:10.39532\n",
      "[11]\ttrain-rmse:10.28597\teval-rmse:10.29160\n",
      "[12]\ttrain-rmse:10.18328\teval-rmse:10.18889\n",
      "[13]\ttrain-rmse:10.08163\teval-rmse:10.08723\n",
      "[14]\ttrain-rmse:9.98099\teval-rmse:9.98657\n",
      "[15]\ttrain-rmse:9.88136\teval-rmse:9.88695\n",
      "[16]\ttrain-rmse:9.78273\teval-rmse:9.78832\n",
      "[17]\ttrain-rmse:9.68508\teval-rmse:9.69063\n",
      "[18]\ttrain-rmse:9.58841\teval-rmse:9.59393\n",
      "[19]\ttrain-rmse:9.49271\teval-rmse:9.49823\n",
      "[20]\ttrain-rmse:9.39796\teval-rmse:9.40346\n",
      "[21]\ttrain-rmse:9.30416\teval-rmse:9.30966\n",
      "[22]\ttrain-rmse:9.21131\teval-rmse:9.21680\n",
      "[23]\ttrain-rmse:9.11938\teval-rmse:9.12486\n",
      "[24]\ttrain-rmse:9.02836\teval-rmse:9.03382\n",
      "[25]\ttrain-rmse:8.93827\teval-rmse:8.94370\n",
      "[26]\ttrain-rmse:8.84906\teval-rmse:8.85448\n",
      "[27]\ttrain-rmse:8.76076\teval-rmse:8.76616\n",
      "[28]\ttrain-rmse:8.67333\teval-rmse:8.67871\n",
      "[29]\ttrain-rmse:8.58678\teval-rmse:8.59214\n",
      "[30]\ttrain-rmse:8.50110\teval-rmse:8.50645\n",
      "[31]\ttrain-rmse:8.41628\teval-rmse:8.42160\n",
      "[32]\ttrain-rmse:8.33230\teval-rmse:8.33760\n",
      "[33]\ttrain-rmse:8.24917\teval-rmse:8.25445\n",
      "[34]\ttrain-rmse:8.16687\teval-rmse:8.17212\n",
      "[35]\ttrain-rmse:8.08539\teval-rmse:8.09064\n",
      "[36]\ttrain-rmse:8.00473\teval-rmse:8.00996\n",
      "[37]\ttrain-rmse:7.92488\teval-rmse:7.93008\n",
      "[38]\ttrain-rmse:7.84582\teval-rmse:7.85101\n",
      "[39]\ttrain-rmse:7.76756\teval-rmse:7.77273\n",
      "[40]\ttrain-rmse:7.69008\teval-rmse:7.69524\n",
      "[41]\ttrain-rmse:7.61338\teval-rmse:7.61852\n",
      "[42]\ttrain-rmse:7.53744\teval-rmse:7.54256\n",
      "[43]\ttrain-rmse:7.46227\teval-rmse:7.46738\n",
      "[44]\ttrain-rmse:7.38785\teval-rmse:7.39293\n",
      "[45]\ttrain-rmse:7.31418\teval-rmse:7.31924\n",
      "[46]\ttrain-rmse:7.24123\teval-rmse:7.24628\n",
      "[47]\ttrain-rmse:7.16903\teval-rmse:7.17406\n",
      "[48]\ttrain-rmse:7.09755\teval-rmse:7.10257\n",
      "[49]\ttrain-rmse:7.02678\teval-rmse:7.03178\n",
      "[50]\ttrain-rmse:6.95672\teval-rmse:6.96171\n",
      "[51]\ttrain-rmse:6.88736\teval-rmse:6.89235\n",
      "[52]\ttrain-rmse:6.81870\teval-rmse:6.82366\n",
      "[53]\ttrain-rmse:6.75073\teval-rmse:6.75567\n",
      "[54]\ttrain-rmse:6.68343\teval-rmse:6.68837\n",
      "[55]\ttrain-rmse:6.61681\teval-rmse:6.62173\n",
      "[56]\ttrain-rmse:6.55086\teval-rmse:6.55575\n",
      "[57]\ttrain-rmse:6.48557\teval-rmse:6.49045\n",
      "[58]\ttrain-rmse:6.42093\teval-rmse:6.42577\n",
      "[59]\ttrain-rmse:6.35694\teval-rmse:6.36176\n",
      "[60]\ttrain-rmse:6.29359\teval-rmse:6.29837\n",
      "[61]\ttrain-rmse:6.23088\teval-rmse:6.23563\n",
      "[62]\ttrain-rmse:6.16880\teval-rmse:6.17353\n",
      "[63]\ttrain-rmse:6.10733\teval-rmse:6.11205\n",
      "[64]\ttrain-rmse:6.04649\teval-rmse:6.05116\n",
      "[65]\ttrain-rmse:5.98625\teval-rmse:5.99091\n",
      "[66]\ttrain-rmse:5.92661\teval-rmse:5.93123\n",
      "[67]\ttrain-rmse:5.86758\teval-rmse:5.87214\n",
      "[68]\ttrain-rmse:5.80913\teval-rmse:5.81365\n",
      "[69]\ttrain-rmse:5.75127\teval-rmse:5.75574\n",
      "[70]\ttrain-rmse:5.69399\teval-rmse:5.69842\n",
      "[71]\ttrain-rmse:5.63728\teval-rmse:5.64167\n",
      "[72]\ttrain-rmse:5.58115\teval-rmse:5.58551\n",
      "[73]\ttrain-rmse:5.52557\teval-rmse:5.52990\n",
      "[74]\ttrain-rmse:5.47055\teval-rmse:5.47483\n",
      "[75]\ttrain-rmse:5.41608\teval-rmse:5.42032\n",
      "[76]\ttrain-rmse:5.36217\teval-rmse:5.36638\n",
      "[77]\ttrain-rmse:5.30879\teval-rmse:5.31296\n",
      "[78]\ttrain-rmse:5.25594\teval-rmse:5.26007\n",
      "[79]\ttrain-rmse:5.20363\teval-rmse:5.20771\n",
      "[80]\ttrain-rmse:5.15184\teval-rmse:5.15589\n",
      "[81]\ttrain-rmse:5.10057\teval-rmse:5.10458\n",
      "[82]\ttrain-rmse:5.04981\teval-rmse:5.05379\n",
      "[83]\ttrain-rmse:4.99956\teval-rmse:5.00352\n",
      "[84]\ttrain-rmse:4.94982\teval-rmse:4.95374\n",
      "[85]\ttrain-rmse:4.90058\teval-rmse:4.90445\n",
      "[86]\ttrain-rmse:4.85182\teval-rmse:4.85565\n",
      "[87]\ttrain-rmse:4.80356\teval-rmse:4.80736\n",
      "[88]\ttrain-rmse:4.75578\teval-rmse:4.75955\n",
      "[89]\ttrain-rmse:4.70848\teval-rmse:4.71223\n",
      "[90]\ttrain-rmse:4.66166\teval-rmse:4.66539\n",
      "[91]\ttrain-rmse:4.61530\teval-rmse:4.61899\n",
      "[92]\ttrain-rmse:4.56940\teval-rmse:4.57307\n",
      "[93]\ttrain-rmse:4.52397\teval-rmse:4.52762\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[94]\ttrain-rmse:4.47900\teval-rmse:4.48263\n",
      "[95]\ttrain-rmse:4.43448\teval-rmse:4.43807\n",
      "[96]\ttrain-rmse:4.39040\teval-rmse:4.39395\n",
      "[97]\ttrain-rmse:4.34677\teval-rmse:4.35028\n",
      "[98]\ttrain-rmse:4.30357\teval-rmse:4.30704\n",
      "[99]\ttrain-rmse:4.26081\teval-rmse:4.26426\n",
      "\n",
      "\n",
      "\n",
      "Training complete\n",
      "\n",
      "\n",
      "For eta 0.100, RMSE is 0.22691 \n",
      "For eta 0.010, RMSE is 4.26426 \n"
     ]
    }
   ],
   "source": [
    "eta_list =  [0.1,0.01]\n",
    "RMSE = []\n",
    "\n",
    "for eta in eta_list:\n",
    "    xgb_params['eta'] = eta\n",
    "    num_round = 100\n",
    "    print (\"Training with eta %0.3f:\\n\\n\" %(eta))\n",
    "    bst = xgb.train(xgb_params, dtrain, num_round, watchlist)\n",
    "    print (\"\\n\\n\")\n",
    "    y_pred = bst.predict(dval)\n",
    "    rmse = mean_squared_error(y_val,y_pred, squared= False)\n",
    "    RMSE.append(rmse)\n",
    "\n",
    "\n",
    "print(\"Training complete\\n\\n\")\n",
    "for i,eta in enumerate(eta_list):\n",
    "    print (\"For eta %0.3f, RMSE is %0.5f \" %(eta, RMSE[i]))    \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Yl6Xn6Zb76iH"
   },
   "source": [
    "- Which eta leads to the best RMSE score on the validation dataset?\n",
    "-Answer:\n",
    "     \n",
    "     - RMSE of 0.01 is worst at 4.26426, \n",
    "     - for 0.1 it is 0.22691, \n",
    "     - whereas 0.3 gives best RMSE of 0.21954"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}