DS_Lecture10.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# From Trees to Forests\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.tree import DecisionTreeClassifier, DecisionTreeRegressor\n",
    "from sklearn.ensemble import BaggingClassifier, BaggingRegressor\n",
    "\n",
    "from sklearn import tree\n",
    "#from graphviz import Source\n",
    "#import graphviz\n",
    "from IPython.display import Image  \n",
    "#from sklearn.tree import export_graphviz\n",
    "#import pydotplus\n",
    "from six import StringIO  \n",
    "from sklearn.inspection import permutation_importance"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Classification Tree - Titanic\n",
    "\n",
    "Let us build a classification tree on the Titanic data using `Survived` as the outcome and `Sex, Pclass, Age, PassengerId` as explanatory variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 891 entries, 0 to 890\n",
      "Data columns (total 12 columns):\n",
      " #   Column       Non-Null Count  Dtype  \n",
      "---  ------       --------------  -----  \n",
      " 0   PassengerId  891 non-null    int64  \n",
      " 1   Survived     891 non-null    int64  \n",
      " 2   Pclass       891 non-null    int64  \n",
      " 3   Name         891 non-null    object \n",
      " 4   Sex          891 non-null    object \n",
      " 5   Age          714 non-null    float64\n",
      " 6   SibSp        891 non-null    int64  \n",
      " 7   Parch        891 non-null    int64  \n",
      " 8   Ticket       891 non-null    object \n",
      " 9   Fare         891 non-null    float64\n",
      " 10  Cabin        204 non-null    object \n",
      " 11  Embarked     889 non-null    object \n",
      "dtypes: float64(2), int64(5), object(5)\n",
      "memory usage: 83.7+ KB\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv('data/TitanicTrain.csv') # Load the data\n",
    "df.info()"
   ]
  },
  {
   "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>Pclass</th>\n",
       "      <th>Age</th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Sex_female</th>\n",
       "      <th>Sex_male</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>26.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>35.0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Pclass   Age  PassengerId  Survived  Sex_female  Sex_male\n",
       "0       3  22.0            1         0           0         1\n",
       "1       1  38.0            2         1           1         0\n",
       "2       3  26.0            3         1           1         0\n",
       "3       1  35.0            4         1           1         0\n",
       "4       3  35.0            5         0           0         1"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "one_hot_data = pd.get_dummies(df[[\"Sex\", \"Pclass\", \"Age\", \"PassengerId\",\"Survived\"]])\n",
    "one_hot_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "train=one_hot_data[[\"Sex_male\", \"Pclass\", \"Age\", \"PassengerId\",\"Survived\"]].dropna()\n",
    "X_train = train[[\"Sex_male\", \"Pclass\", \"Age\", \"PassengerId\"]]\n",
    "Y_train = train[\"Survived\"]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Shallow Tree of Depth 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn import tree\n",
    "plt.rcParams[\"figure.figsize\"]=6,5\n",
    "\n",
    "clf = tree.DecisionTreeClassifier(max_depth=2)\n",
    "clf = clf.fit(X_train, Y_train)\n",
    "\n",
    "\n",
    "tree.plot_tree(clf);"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Variable Importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#MDI\n",
    "MDI_importances = clf.feature_importances_\n",
    "indices = np.argsort(MDI_importances)\n",
    "features = X_train.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "#permutation importance:\n",
    "MDA_importances = permutation_importance(clf, X_train, Y_train, n_repeats=5,random_state=42, n_jobs=2)\n",
    "sorted_idx = MDA_importances.importances_mean.argsort()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams[\"figure.figsize\"]=10,4\n",
    "fig, (ax1, ax2) = plt.subplots(1, 2)\n",
    "\n",
    "ax1.set_title('MDI Importances')\n",
    "ax1.barh(range(len(indices)), MDI_importances[indices], color='b', align='center')\n",
    "ax1.set_yticks( np.arange(4))\n",
    "ax1.set_yticklabels(features[indices])\n",
    "ax1.set(xlabel='Relative Importance')\n",
    "\n",
    "\n",
    "ax2.boxplot(MDA_importances.importances[sorted_idx].T,\n",
    "           vert=False, labels=features[sorted_idx])\n",
    "tpmp=ax2.set_title(\"Permutation Importances\")\n",
    "fig.tight_layout()\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Increasing Tree Depth \n",
    "\n",
    "Notice the finer and finer splits on Age and PassengerId"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "clf = tree.DecisionTreeClassifier(max_depth=5)\n",
    "clf = clf.fit(X_train, Y_train)\n",
    "\n",
    "tmp=tree.plot_tree(clf);"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Variable Importance\n",
    "\n",
    "Notice that passenger ID and age are gaining MDI importance as the tree depth is increased!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "#permutation importance:\n",
    "MDA_importances = permutation_importance(clf, X_train, Y_train, n_repeats=5,random_state=42, n_jobs=2)\n",
    "sorted_idx = MDA_importances.importances_mean.argsort()\n",
    "\n",
    "#MDI\n",
    "MDI_importances = clf.feature_importances_\n",
    "indices = np.argsort(MDI_importances)\n",
    "features = X_train.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams[\"figure.figsize\"]=10,4\n",
    "fig, (ax1, ax2) = plt.subplots(1, 2)\n",
    "\n",
    "ax1.set_title('MDI Importances')\n",
    "ax1.barh(range(len(indices)), MDI_importances[indices], color='b', align='center')\n",
    "ax1.set_yticks( np.arange(4))\n",
    "ax1.set_yticklabels(features[indices])\n",
    "ax1.set(xlabel='Relative Importance')\n",
    "\n",
    "\n",
    "ax2.boxplot(MDA_importances.importances[sorted_idx].T,\n",
    "           vert=False, labels=features[sorted_idx])\n",
    "tpmp=ax2.set_title(\"Permutation Importances\")\n",
    "fig.tight_layout()\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ensembles of Estimators\n",
    "\n",
    "This notion—that multiple overfitting estimators can be combined to reduce the effect of this overfitting—is what underlies an ensemble method called *bagging*.\n",
    "Bagging makes use of an ensemble (a grab bag, perhaps) of parallel estimators, each of which over-fits the data, and averages the results to find a better classification.\n",
    "An ensemble of randomized decision trees is known as a *random forest*.\n",
    "\n",
    "This type of bagging classification can be done manually using Scikit-Learn's ``BaggingClassifier`` meta-estimator, as shown here:"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Loading and Splitting the Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "#`load_boston` has been removed from scikit-learn since version 1.2.\n",
    "#from sklearn.datasets import load_boston\n",
    "import random\n",
    "from sklearn.metrics import mean_squared_error,confusion_matrix, classification_report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>black</th>\n",
       "      <th>lstat</th>\n",
       "      <th>medv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.00632</td>\n",
       "      <td>18.0</td>\n",
       "      <td>2.31</td>\n",
       "      <td>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</td>\n",
       "      <td>296</td>\n",
       "      <td>15.3</td>\n",
       "      <td>396.90</td>\n",
       "      <td>4.98</td>\n",
       "      <td>24.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.02731</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.07</td>\n",
       "      <td>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</td>\n",
       "      <td>242</td>\n",
       "      <td>17.8</td>\n",
       "      <td>396.90</td>\n",
       "      <td>9.14</td>\n",
       "      <td>21.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.02729</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.07</td>\n",
       "      <td>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</td>\n",
       "      <td>242</td>\n",
       "      <td>17.8</td>\n",
       "      <td>392.83</td>\n",
       "      <td>4.03</td>\n",
       "      <td>34.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.03237</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.18</td>\n",
       "      <td>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</td>\n",
       "      <td>222</td>\n",
       "      <td>18.7</td>\n",
       "      <td>394.63</td>\n",
       "      <td>2.94</td>\n",
       "      <td>33.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.06905</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.18</td>\n",
       "      <td>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</td>\n",
       "      <td>222</td>\n",
       "      <td>18.7</td>\n",
       "      <td>396.90</td>\n",
       "      <td>5.33</td>\n",
       "      <td>36.2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      crim    zn  indus  chas    nox     rm   age     dis  rad  tax  ptratio  \\\n",
       "0  0.00632  18.0   2.31     0  0.538  6.575  65.2  4.0900    1  296     15.3   \n",
       "1  0.02731   0.0   7.07     0  0.469  6.421  78.9  4.9671    2  242     17.8   \n",
       "2  0.02729   0.0   7.07     0  0.469  7.185  61.1  4.9671    2  242     17.8   \n",
       "3  0.03237   0.0   2.18     0  0.458  6.998  45.8  6.0622    3  222     18.7   \n",
       "4  0.06905   0.0   2.18     0  0.458  7.147  54.2  6.0622    3  222     18.7   \n",
       "\n",
       "    black  lstat  medv  \n",
       "0  396.90   4.98  24.0  \n",
       "1  396.90   9.14  21.6  \n",
       "2  392.83   4.03  34.7  \n",
       "3  394.63   2.94  33.4  \n",
       "4  396.90   5.33  36.2  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#`load_boston` has been removed from scikit-learn since version 1.2.\n",
    "#boston = load_boston()\n",
    "url=\"https://raw.githubusercontent.com/markusloecher/DataScience-HWR/main/data/Boston.csv\"\n",
    "boston_df=pd.read_csv(url)\n",
    "boston_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(354, 13)\n",
      "(152, 13)\n"
     ]
    }
   ],
   "source": [
    "y = boston_df['medv']\n",
    "X = boston_df.drop('medv',axis=1)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=123)\n",
    "print(X_train.shape)\n",
    "print(X_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.224170737677708"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree_B1 = DecisionTreeRegressor(random_state=0)\n",
    "tree_B1.fit(X_train, y_train)\n",
    "\n",
    "singleTreePreds = tree_B1.predict(X_test)\n",
    "np.sqrt(mean_squared_error(y_test, singleTreePreds))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Bagging"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "baggedBoston = BaggingRegressor(base_estimator=tree_B1,n_estimators=200,oob_score=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/loecherm/.pyenv/versions/aug_hs_env/lib/python3.9/site-packages/sklearn/ensemble/_base.py:166: FutureWarning: `base_estimator` was renamed to `estimator` in version 1.2 and will be removed in 1.4.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "3.699118861118715"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The folloing line of code does all of these steps:\n",
    "# It fits 200 trees on 200 bootstrapped vrsions of your training data !!\n",
    "baggedBoston.fit(X_train, y_train) \n",
    "\n",
    "# The predict function now computes 200 predistions for each data point and averages them !!\n",
    "baggedPreds = baggedBoston.predict(X_test)\n",
    "\n",
    "np.sqrt(mean_squared_error(y_test, baggedPreds))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8727022246437939"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "baggedBoston.oob_score_ # this is ythe equivalent of your test set loss !"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.4412431529172727"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "\n",
    "forest = RandomForestRegressor(200, max_features=0.5, oob_score=True)\n",
    "forest.fit(X_train, y_train)\n",
    "\n",
    "forestPreds = forest.predict(X_test)\n",
    "np.sqrt(mean_squared_error(y_test, forestPreds))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8758200098217869"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forest.oob_score_"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tasks\n",
    "\n",
    "1. Find the best value for `mtry`\n",
    "2. Plot the two kinds of variable importances"
   ]
  }
 ],
 "metadata": {
  "hide_input": false,
  "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": "3.9.0"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}