03-06-Nonlinear Relationships.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Nonlinear Relationships"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set_style('whitegrid')\n",
    "%matplotlib inline\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.16.5\n",
      "0.25.1\n",
      "3.7.4 (default, Aug 13 2019, 15:17:50) \n",
      "[Clang 4.0.1 (tags/RELEASE_401/final)]\n",
      "0.21.3\n"
     ]
    }
   ],
   "source": [
    "print(np.__version__)\n",
    "print(pd.__version__)\n",
    "import sys\n",
    "print(sys.version)\n",
    "import sklearn\n",
    "print(sklearn.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_boston\n",
    "boston_data = load_boston()\n",
    "df = pd.DataFrame(boston_data.data, columns=boston_data.feature_names)"
   ]
  },
  {
   "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>CRIM</th>\n",
       "      <th>ZN</th>\n",
       "      <th>INDUS</th>\n",
       "      <th>CHAS</th>\n",
       "      <th>NOX</th>\n",
       "      <th>RM</th>\n",
       "      <th>AGE</th>\n",
       "      <th>DIS</th>\n",
       "      <th>RAD</th>\n",
       "      <th>TAX</th>\n",
       "      <th>PTRATIO</th>\n",
       "      <th>B</th>\n",
       "      <th>LSTAT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.00632</td>\n",
       "      <td>18.0</td>\n",
       "      <td>2.31</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.538</td>\n",
       "      <td>6.575</td>\n",
       "      <td>65.2</td>\n",
       "      <td>4.0900</td>\n",
       "      <td>1.0</td>\n",
       "      <td>296.0</td>\n",
       "      <td>15.3</td>\n",
       "      <td>396.90</td>\n",
       "      <td>4.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.02731</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.469</td>\n",
       "      <td>6.421</td>\n",
       "      <td>78.9</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2.0</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.8</td>\n",
       "      <td>396.90</td>\n",
       "      <td>9.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.02729</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.469</td>\n",
       "      <td>7.185</td>\n",
       "      <td>61.1</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2.0</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.8</td>\n",
       "      <td>392.83</td>\n",
       "      <td>4.03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.03237</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.18</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.458</td>\n",
       "      <td>6.998</td>\n",
       "      <td>45.8</td>\n",
       "      <td>6.0622</td>\n",
       "      <td>3.0</td>\n",
       "      <td>222.0</td>\n",
       "      <td>18.7</td>\n",
       "      <td>394.63</td>\n",
       "      <td>2.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0.06905</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.18</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.458</td>\n",
       "      <td>7.147</td>\n",
       "      <td>54.2</td>\n",
       "      <td>6.0622</td>\n",
       "      <td>3.0</td>\n",
       "      <td>222.0</td>\n",
       "      <td>18.7</td>\n",
       "      <td>396.90</td>\n",
       "      <td>5.33</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      CRIM    ZN  INDUS  CHAS    NOX     RM   AGE     DIS  RAD    TAX  \\\n",
       "0  0.00632  18.0   2.31   0.0  0.538  6.575  65.2  4.0900  1.0  296.0   \n",
       "1  0.02731   0.0   7.07   0.0  0.469  6.421  78.9  4.9671  2.0  242.0   \n",
       "2  0.02729   0.0   7.07   0.0  0.469  7.185  61.1  4.9671  2.0  242.0   \n",
       "3  0.03237   0.0   2.18   0.0  0.458  6.998  45.8  6.0622  3.0  222.0   \n",
       "4  0.06905   0.0   2.18   0.0  0.458  7.147  54.2  6.0622  3.0  222.0   \n",
       "\n",
       "   PTRATIO       B  LSTAT  \n",
       "0     15.3  396.90   4.98  \n",
       "1     17.8  396.90   9.14  \n",
       "2     17.8  392.83   4.03  \n",
       "3     18.7  394.63   2.94  \n",
       "4     18.7  396.90   5.33  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = boston_data.target"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "df = pd.read_csv('housing.data', delim_whitespace=True, header=None)\n",
    "df.columns = ['CRIM', 'ZN' , 'INDUS', 'CHAS', 'NOX', 'RM', 'AGE', \n",
    "              'DIS', 'RAD', 'TAX', 'PTRATIO', 'B', 'LSTAT', 'MEDV']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Brief Introduction to Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df[['LSTAT']].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree = DecisionTreeRegressor(max_depth=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeRegressor(criterion='mse', max_depth=5, max_features=None,\n",
       "                      max_leaf_nodes=None, min_impurity_decrease=0.0,\n",
       "                      min_impurity_split=None, min_samples_leaf=1,\n",
       "                      min_samples_split=2, min_weight_fraction_leaf=0.0,\n",
       "                      presort=False, random_state=None, splitter='best')"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "sort_idx = X.flatten().argsort()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,8))\n",
    "plt.scatter(X[sort_idx], y[sort_idx])\n",
    "plt.plot(X[sort_idx], tree.predict(X[sort_idx]), color='k')\n",
    "\n",
    "plt.xlabel('LSTAT')\n",
    "plt.ylabel('MEDV');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using `max_depth` of 5 led to overfitting. Let's try 2 instead."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "tree = DecisionTreeRegressor(max_depth=2)\n",
    "tree.fit(X, y)\n",
    "sort_idx = X.flatten().argsort()\n",
    "plt.figure(figsize=(10,8))\n",
    "plt.scatter(X[sort_idx], y[sort_idx])\n",
    "plt.plot(X[sort_idx], tree.predict(X[sort_idx]), color='k')\n",
    "\n",
    "plt.xlabel('LSTAT')\n",
    "plt.ylabel('MEDV');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Brief Introduction to Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_squared_error, r2_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.values\n",
    "#y = df['MEDV'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, \n",
    "                                                    test_size=0.3, \n",
    "                                                    random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "forest = RandomForestRegressor(n_estimators=500, criterion='mse', \n",
    "                               random_state=42, n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=None,\n",
       "                      max_features='auto', max_leaf_nodes=None,\n",
       "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                      min_samples_leaf=1, min_samples_split=2,\n",
       "                      min_weight_fraction_leaf=0.0, n_estimators=500, n_jobs=-1,\n",
       "                      oob_score=False, random_state=42, verbose=0,\n",
       "                      warm_start=False)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forest.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train_pred = forest.predict(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test_pred = forest.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE train: 1.8700, test: 9.3921\n"
     ]
    }
   ],
   "source": [
    "print(\"MSE train: {0:.4f}, test: {1:.4f}\".\\\n",
    "      format(mean_squared_error(y_train, y_train_pred), \n",
    "             mean_squared_error(y_test, y_test_pred)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R^2 train: 0.9787, test: 0.8740\n"
     ]
    }
   ],
   "source": [
    "print(\"R^2 train: {0:.4f}, test: {1:.4f}\".\\\n",
    "      format(r2_score(y_train, y_train_pred),\n",
    "             r2_score(y_test, y_test_pred)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Brief Introduction to AdaBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import AdaBoostRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "ada = AdaBoostRegressor(DecisionTreeRegressor(max_depth=4), \n",
    "                        n_estimators=500, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AdaBoostRegressor(base_estimator=DecisionTreeRegressor(criterion='mse',\n",
       "                                                       max_depth=4,\n",
       "                                                       max_features=None,\n",
       "                                                       max_leaf_nodes=None,\n",
       "                                                       min_impurity_decrease=0.0,\n",
       "                                                       min_impurity_split=None,\n",
       "                                                       min_samples_leaf=1,\n",
       "                                                       min_samples_split=2,\n",
       "                                                       min_weight_fraction_leaf=0.0,\n",
       "                                                       presort=False,\n",
       "                                                       random_state=None,\n",
       "                                                       splitter='best'),\n",
       "                  learning_rate=1.0, loss='linear', n_estimators=500,\n",
       "                  random_state=42)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ada.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train_pred = ada.predict(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test_pred = ada.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE train: 4.6605, test: 13.4949\n"
     ]
    }
   ],
   "source": [
    "print(\"MSE train: {0:.4f}, test: {1:.4f}\".\\\n",
    "      format(mean_squared_error(y_train, y_train_pred), \n",
    "             mean_squared_error(y_test, y_test_pred)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R^2 train: 0.9470, test: 0.8189\n"
     ]
    }
   ],
   "source": [
    "print(\"R^2 train: {0:.4f}, test: {1:.4f}\".\\\n",
    "      format(r2_score(y_train, y_train_pred),\n",
    "             r2_score(y_test, y_test_pred)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Revisiting Feature Importance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "13 features.\n",
    "\n",
    "Are they all equally important?\n",
    "\n",
    "Which features are more important?\n",
    "\n",
    "Can scikit-learn help us with this?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## According to AdaBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.03104826, 0.00243815, 0.01083651, 0.00085548, 0.0371141 ,\n",
       "       0.23589575, 0.00958776, 0.13743614, 0.01893706, 0.02188852,\n",
       "       0.04185043, 0.02193283, 0.43017901])"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ada.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['CRIM', 'ZN', 'INDUS', 'CHAS', 'NOX', 'RM', 'AGE', 'DIS', 'RAD', 'TAX',\n",
       "       'PTRATIO', 'B', 'LSTAT'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "result = pd.DataFrame(ada.feature_importances_, df.columns)\n",
    "result.columns = ['feature']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "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>feature</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>LSTAT</td>\n",
       "      <td>0.430179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>RM</td>\n",
       "      <td>0.235896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>DIS</td>\n",
       "      <td>0.137436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>PTRATIO</td>\n",
       "      <td>0.041850</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>NOX</td>\n",
       "      <td>0.037114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>CRIM</td>\n",
       "      <td>0.031048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>B</td>\n",
       "      <td>0.021933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>TAX</td>\n",
       "      <td>0.021889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>RAD</td>\n",
       "      <td>0.018937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>INDUS</td>\n",
       "      <td>0.010837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>AGE</td>\n",
       "      <td>0.009588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>ZN</td>\n",
       "      <td>0.002438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>CHAS</td>\n",
       "      <td>0.000855</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          feature\n",
       "LSTAT    0.430179\n",
       "RM       0.235896\n",
       "DIS      0.137436\n",
       "PTRATIO  0.041850\n",
       "NOX      0.037114\n",
       "CRIM     0.031048\n",
       "B        0.021933\n",
       "TAX      0.021889\n",
       "RAD      0.018937\n",
       "INDUS    0.010837\n",
       "AGE      0.009588\n",
       "ZN       0.002438\n",
       "CHAS     0.000855"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result.sort_values(by='feature', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "result.sort_values(by='feature', ascending=False).plot(kind='bar');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## According to Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.03213849, 0.00166784, 0.00677542, 0.00131141, 0.01508521,\n",
       "       0.4038535 , 0.01414013, 0.06280815, 0.00467201, 0.0120068 ,\n",
       "       0.01833147, 0.0127986 , 0.41441098])"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forest.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "result = pd.DataFrame(forest.feature_importances_, df.columns)\n",
    "result.columns = ['feature']\n",
    "result.sort_values(by='feature', ascending=False).plot(kind='bar');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Exercise"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Time for you to try your hands on building machine learning model. All the codes have been provided. All you have to do is study the codes and make some light modifications to tacklet the problem that I am presenting to you now.\n",
    "\n",
    "In previous and this lesson, I performed modelling of our data using Random Forest and AdaBoost. I would like you to make use of Decision Tree Regressor and perform the following:\n",
    "\n",
    "1. Split your data to train and test set, reserving 30% of your data for testing\n",
    "2. Instantiate, fit and predict \n",
    "3. Calculate MSE and R-square\n",
    "4. Extract feature importance and visualise it using bar chart"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's try this out\n",
    "\n",
    ".\n",
    "\n",
    ".\n",
    "\n",
    ".\n",
    "\n",
    "Your turn."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree = DecisionTreeRegressor(max_depth=3)\n",
    "tree.fit(X_train, y_train)\n",
    "\n",
    "y_train_pred = tree.predict(X_train)\n",
    "y_test_pred = tree.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE train: 15.3680, test: 16.9282\n"
     ]
    }
   ],
   "source": [
    "print(\"MSE train: {0:.4f}, test: {1:.4f}\".\\\n",
    "      format(mean_squared_error(y_train, y_train_pred), \n",
    "             mean_squared_error(y_test, y_test_pred)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R^2 train: 0.8252, test: 0.7728\n"
     ]
    }
   ],
   "source": [
    "print(\"R^2 train: {0:.4f}, test: {1:.4f}\".\\\n",
    "      format(r2_score(y_train, y_train_pred),\n",
    "             r2_score(y_test, y_test_pred)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "result = pd.DataFrame(tree.feature_importances_, df.columns)\n",
    "result.columns = ['feature']\n",
    "result.sort_values(by='feature', ascending=False).plot(kind='bar');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}