Creating ROC curve.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "4eccab22",
   "metadata": {},
   "source": [
    "## Code for ROC curve\n",
    "Receiver Operating Characteristics (ROC) curve gives us the estimation for model performance based on comparing true positive and false positive rate for various tresholds of our binary target variable."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e1a9faba",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4599f94a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# we will create some random but imbalanced data\n",
    "np.random.seed(123)\n",
    "y_real = np.random.choice([0,1], 300, p=[0.9, 0.1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "6af588ac",
   "metadata": {},
   "outputs": [],
   "source": [
    "# we will create fake prediction that will be very bad, using uniform distribution\n",
    "y_pred = np.random.uniform(0,1,300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "f189f741",
   "metadata": {},
   "outputs": [],
   "source": [
    "# we are creating function that transforms the predicted values to the binary outcome according to treshold\n",
    "def binarize(y_hat, threshold):\n",
    "    return (y_hat > threshold).astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "518674ce",
   "metadata": {},
   "outputs": [],
   "source": [
    "def calculate_fpr_tpr(y_real, y_pred_cat, treshold):\n",
    "    \n",
    "    positive_indexes=np.where(y_real==1)\n",
    "    negative_indexes = np.where(y_real==0)\n",
    "    \n",
    "    tp = np.sum(y_real[positive_indexes]==y_pred_cat[positive_indexes])\n",
    "    tn = np.sum(y_real[negative_indexes]==y_pred_cat[negative_indexes])\n",
    "    \n",
    "    fp = np.sum(y_real[negative_indexes]!=y_pred_cat[negative_indexes])\n",
    "    fn = np.sum(y_real[positive_indexes]!=y_pred_cat[positive_indexes])\n",
    "    \n",
    "    tpr = tp/(fn+tp)\n",
    "    fpr = fp/(fp+tn)\n",
    "    \n",
    "    \n",
    "    return tpr, fpr\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c969412c",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plotting_ROC(y_real, y_pred):\n",
    "    tresholds = np.arange(0, 1, 0.01)\n",
    "    tp_rates = []\n",
    "    fp_rates = []\n",
    "    for treshold in tresholds:\n",
    "        y_pred_cat = binarize(y_pred, treshold)\n",
    "        \n",
    "        tpr, fpr = calculate_fpr_tpr(y_real, y_pred_cat, treshold)\n",
    "        \n",
    "        tp_rates.append(tpr)\n",
    "        fp_rates.append(fpr)\n",
    "    \n",
    "    plt.figure(figsize=(4,4))\n",
    "    plt.xlabel('False positive rate')\n",
    "    plt.ylabel('True positive rate')\n",
    "    plt.plot(fp_rates, tp_rates, 'k--', color = 'red')\n",
    "    plt.plot([0,1], [0,1], 'k--')\n",
    "    \n",
    "    \n",
    "    return tp_rates, fp_rates\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "b2f9144c",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "true_positive_rates, false_positive_rates = plotting_ROC(y_real, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "0ff87266",
   "metadata": {},
   "outputs": [],
   "source": [
    "def auc(false_positive_rates, true_positive_rates):\n",
    "    y = np.asanyarray(true_positive_rates)\n",
    "    x = np.asanyarray(false_positive_rates)\n",
    "    widths = np.abs(np.diff(x))\n",
    "    heights = (y[1:] + y[:-1]) / 2.0\n",
    "    area = (widths * heights).sum()\n",
    "    return area"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "14e70db1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Area under curve:  0.48\n"
     ]
    }
   ],
   "source": [
    "print('Area under curve: ',round(auc(false_positive_rates, true_positive_rates), 2))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33572da4",
   "metadata": {},
   "source": [
    "### Good model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "2bed2ec3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Now we will make better model which will have better auc to test our function with famous iris dataset\n",
    "from sklearn.datasets import load_iris\n",
    "from sklearn.model_selection import train_test_split\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.linear_model import LogisticRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "0a2c82d6",
   "metadata": {},
   "outputs": [],
   "source": [
    "iris = load_iris(as_frame=True)\n",
    "iris = pd.DataFrame(data= np.c_[iris['data'], iris['target']],\n",
    "                     columns= iris['feature_names'] + ['target'])\n",
    "iris= iris[(iris['target']==0) | (iris['target']==1)] # we will work only with binary target to test our code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "65804f3f",
   "metadata": {},
   "outputs": [],
   "source": [
    "y = iris['target'].to_numpy()\n",
    "X = iris[['sepal length (cm)', 'sepal width (cm)', 'petal length (cm)',\n",
    "       'petal width (cm)',]].to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "2f2a1b51",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "8cef7cec",
   "metadata": {},
   "outputs": [],
   "source": [
    "logit_model = LogisticRegression(random_state=0).fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "84854716",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = logit_model.predict_proba(X_test)\n",
    "y_pred = np.asarray([i[1] for i in y_pred])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "5109e451",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "true_positive_rates_good, false_positive_rates_good = plotting_ROC(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "956b5a86",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Area under curve:  1.0\n"
     ]
    }
   ],
   "source": [
    "print('Area under curve: ',round(auc(false_positive_rates_good, true_positive_rates_good), 2))"
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}