notebooks/49_tune_multiple_models.ipynb

{
 "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%2F49_tune_multiple_models.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/49_tune_multiple_models.ipynb)\n",
    "\n",
    "# 🤖⚡ scikit-learn tip #49 ([video](https://www.youtube.com/watch?v=v2QpvCJ1ar8&list=PL5-da3qGB5ID7YYAqireYEew2mWVvgmj6&index=49))\n",
    "\n",
    "You can tune 2+ models using the same grid search! Here's how:\n",
    "\n",
    "1. Create multiple parameter dictionaries\n",
    "2. Specify the model within each dictionary\n",
    "3. Put the dictionaries in a list\n",
    "\n",
    "See example 👇"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "df = pd.read_csv('http://bit.ly/kaggletrain')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols = ['Sex', 'Name', 'Age']\n",
    "X = df[cols]\n",
    "y = df['Survived']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.model_selection import GridSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# this will be the first Pipeline step\n",
    "ct = ColumnTransformer(\n",
    "    [('ohe', OneHotEncoder(), ['Sex']),\n",
    "     ('vectorizer', CountVectorizer(), 'Name'),\n",
    "     ('imputer', SimpleImputer(), ['Age'])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# each of these models will take a turn as the second Pipeline step\n",
    "clf1 = LogisticRegression(solver='liblinear', random_state=1)\n",
    "clf2 = RandomForestClassifier(random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create the Pipeline\n",
    "pipe = Pipeline([('preprocessor', ct), ('classifier', clf1)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create the parameter dictionary for clf1\n",
    "params1 = {}\n",
    "params1['preprocessor__vectorizer__ngram_range'] = [(1, 1), (1, 2)]\n",
    "params1['classifier__penalty'] = ['l1', 'l2']\n",
    "params1['classifier__C'] = [0.1, 1, 10]\n",
    "params1['classifier'] = [clf1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create the parameter dictionary for clf2\n",
    "params2 = {}\n",
    "params2['preprocessor__vectorizer__ngram_range'] = [(1, 1), (1, 2)]\n",
    "params2['classifier__n_estimators'] = [100, 200]\n",
    "params2['classifier__min_samples_leaf'] = [1, 2]\n",
    "params2['classifier'] = [clf2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create a list of parameter dictionaries\n",
    "params = [params1, params2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'classifier': LogisticRegression(C=10, penalty='l1', random_state=1, solver='liblinear'),\n",
       " 'classifier__C': 10,\n",
       " 'classifier__penalty': 'l1',\n",
       " 'preprocessor__vectorizer__ngram_range': (1, 2)}"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# this will search every parameter combination within each dictionary\n",
    "grid = GridSearchCV(pipe, params)\n",
    "grid.fit(X, y)\n",
    "grid.best_params_"
   ]
  },
  {
   "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
}