{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[![Open in Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/justmarkham/scikit-learn-tips/master?filepath=notebooks%2F43_ordinal_encoding_for_trees.ipynb)\n",
    "\n",
    "[![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/justmarkham/scikit-learn-tips/blob/master/notebooks/43_ordinal_encoding_for_trees.ipynb)\n",
    "\n",
    "# 🤖⚡ scikit-learn tip #43 ([video](https://www.youtube.com/watch?v=n_x40CdPZss&list=PL5-da3qGB5ID7YYAqireYEew2mWVvgmj6&index=43))\n",
    "\n",
    "With a tree-based model, try OrdinalEncoder instead of OneHotEncoder even for nominal (unordered) features.\n",
    "\n",
    "Accuracy will often be similar, but OrdinalEncoder will be much faster!\n",
    "\n",
    "See example 👇"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "df = pd.read_csv('https://www.openml.org/data/get_csv/1595261/adult-census.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import OneHotEncoder, OrdinalEncoder\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.model_selection import cross_val_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "categorical_cols = ['workclass', 'education', 'marital-status',\n",
    "                    'occupation', 'relationship', 'race', 'sex']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df[categorical_cols]\n",
    "y = df['class']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(48842, 60)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# OneHotEncoder creates 60 columns\n",
    "ohe = OneHotEncoder()\n",
    "ohe.fit_transform(X).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(48842, 7)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# OrdinalEncoder creates 7 columns\n",
    "oe = OrdinalEncoder()\n",
    "oe.fit_transform(X).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Random Forests is a tree-based model\n",
    "rf = RandomForestClassifier(random_state=1, n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1.95 s, sys: 189 ms, total: 2.14 s\n",
      "Wall time: 23.2 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.8262561170407418"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Pipeline containing OneHotEncoder\n",
    "ohe_pipe = make_pipeline(ohe, rf)\n",
    "%time cross_val_score(ohe_pipe, X, y).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1.67 s, sys: 133 ms, total: 1.81 s\n",
      "Wall time: 3.83 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.8256623624061437"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Pipeline containing OrdinalEncoder\n",
    "oe_pipe = make_pipeline(oe, rf)\n",
    "%time cross_val_score(oe_pipe, X, y).mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Want more tips? [View all tips on GitHub](https://github.com/justmarkham/scikit-learn-tips) or [Sign up to receive 2 tips by email every week](https://scikit-learn.tips) 💌\n",
    "\n",
    "© 2020 [Data School](https://www.dataschool.io). All rights reserved."
   ]
  }
 ],
 "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": "3.9.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}