module4_logistic_regression.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "# CPSC 4970 AI + ML: Module 4 -- Logistic regression (classifier)\n",
    "\n",
    "New stuff:\n",
    "- [LogisticRegression](https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression)\n",
    "- [plot confusion matrix](https://scikit-learn.org/stable/modules/generated/sklearn.metrics.plot_confusion_matrix.html#sklearn.metrics.plot_confusion_matrix)\n",
    "- [metrics](https://scikit-learn.org/stable/modules/classes.html#sklearn-metrics-metrics)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import OneHotEncoder, LabelEncoder\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.metrics import ConfusionMatrixDisplay\n",
    "from IPython.display import display\n",
    "import pandas as pd\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use(\"dark_background\")\n",
    "\n",
    "\n",
    "# Data from https://www.openml.org/d/31\n",
    "# Data originally from https://archive.ics.uci.edu/ml/datasets/statlog+(german+credit+data)\n",
    "# but the openml version is cleaned up a bit.\n",
    "df = pd.read_csv('data/dataset_31_credit-g.csv')"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "outputs": [],
   "source": [
    "ct = ColumnTransformer(\n",
    "    [('one-hot', OneHotEncoder(), ['checking_status', 'credit_history', 'purpose',\n",
    "                                  'savings_status', 'employment', 'personal_status',\n",
    "                                  'other_parties', 'property_magnitude', 'other_payment_plans',\n",
    "                                  'housing', 'job', 'own_telephone', 'foreign_worker'])],\n",
    "    remainder='passthrough'\n",
    ")\n",
    "\n",
    "X = df.iloc[:, :-1]\n",
    "y = df['class']\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y)\n",
    "\n",
    "X_train_transformed = ct.fit_transform(X_train)\n",
    "X_test_transformed = ct.transform(X_test)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training accuracy:  0.8026666666666666\n",
      "Training precision:  0.8268551236749117\n",
      "Confusion matrix:  [[134  98]\n",
      " [ 50 468]]\n"
     ]
    },
    {
     "data": {
      "text/plain": "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x141440940>"
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 2 Axes>",
      "image/png": "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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 2 Axes>",
      "image/png": "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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score, precision_score, confusion_matrix, plot_confusion_matrix\n",
    "\n",
    "lr = LogisticRegression(max_iter=1000).fit(X_train_transformed, y_train)\n",
    "print(\"Training accuracy: \", accuracy_score(y_train, lr.predict(X_train_transformed)))\n",
    "print(\"Training precision: \", precision_score(y_train, lr.predict(X_train_transformed), pos_label=lr.classes_[1]))\n",
    "print(\"Confusion matrix: \", confusion_matrix(y_train, lr.predict(X_train_transformed)))\n",
    "ConfusionMatrixDisplay.from_estimator(lr, X_train_transformed, y_train)\n",
    "ConfusionMatrixDisplay.from_estimator(lr, X_test_transformed, y_test)"
   ],
   "metadata": {
    "collapsed": false
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}