exe4_logistic_regression.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "76ef1cf5",
   "metadata": {
    "id": "76ef1cf5"
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import scipy.optimize, scipy.special\n",
    "import csv\n",
    "import numpy as np\n",
    "import copy\n",
    "from scipy.special import expit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ef02bcad",
   "metadata": {
    "id": "ef02bcad",
    "outputId": "fab7fce7-adb8-4691-ea4e-2d0bb4bbaf02",
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#X = []\n",
    "y = []\n",
    "X1=[]\n",
    "f = open('ex2data1.txt', 'r', encoding='utf-8')\n",
    "data_reader = csv.reader(f)\n",
    "for line in data_reader:\n",
    "    #print(line)\n",
    "    X1.append([float(line[0]), float(line[1])])\n",
    "    y.append(float(line[2]))\n",
    "\n",
    "f.close()\n",
    "\n",
    "m = len(X1)\n",
    "X=np.array(X1)\n",
    "\n",
    "fg1 = plt.figure()\n",
    "for i in range(m):\n",
    "    if y[i] == 1.0:\n",
    "        plt.plot(X[i][0], X[i][1], 'bo')\n",
    "    else:\n",
    "        plt.plot(X[i][0], X[i][1], 'ro')\n",
    "\n",
    "#H=theta[2]*x1 + theta[1]*x2 +theta[0]\n",
    "ones = np.reshape(np.ones(m), (m, 1)) \n",
    "X = np.reshape(X, (m, 2))\n",
    "X = np.concatenate((ones, X), axis=1) #[1, x1 ,x2] shape=(100,3)\n",
    "y = np.array(y)\n",
    "theta = np.zeros([3, 1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "c202e999",
   "metadata": {
    "id": "c202e999"
   },
   "outputs": [],
   "source": [
    "def sigmoid(x):\n",
    "    a = np.exp(-x)\n",
    "    return 1.0 / (1.0 + a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a031c397",
   "metadata": {
    "id": "a031c397"
   },
   "outputs": [],
   "source": [
    "Js = []\n",
    "epsilon = 0.0000001 # zero division defender\n",
    "\n",
    "########## Gradient Descent ##########################\n",
    "\n",
    "def computeCost(theta, X, y):\n",
    "    m = len(X)\n",
    "    H = sigmoid(X.dot(theta))\n",
    "    J = ((np.log(H + epsilon).T.dot(-y) - np.log(1.0 - H + epsilon).T.dot(1-y)) / m).flatten()\n",
    "    print(J)\n",
    "    Js.append(J[0])\n",
    "    return J"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "634fac09",
   "metadata": {
    "id": "634fac09"
   },
   "outputs": [],
   "source": [
    "iterations = 400\n",
    "\n",
    "def findMinTheta(theta, X, y):\n",
    "    result = scipy.optimize.fmin(computeCost, x0=theta, args=(X, y), maxiter=iterations, full_output=True)\n",
    "    return result[0], result[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c9d10b31",
   "metadata": {
    "id": "c9d10b31",
    "outputId": "6f750815-3c65-4f31-e009-9a8d21342d36",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.69314698]\n",
      "[0.69312199]\n",
      "[0.69018127]\n",
      "[0.6903682]\n",
      "[0.68931536]\n",
      "[0.68744734]\n",
      "[0.68564825]\n",
      "[0.68209698]\n",
      "[0.68280297]\n",
      "[0.67834541]\n",
      "[0.6729512]\n",
      "[0.67155928]\n",
      "[0.66462395]\n",
      "[0.66413987]\n",
      "[0.65648268]\n",
      "[0.65049993]\n",
      "[0.6405817]\n",
      "[0.63909197]\n",
      "[0.63256564]\n",
      "[0.63217106]\n",
      "[0.63870517]\n",
      "[0.6365237]\n",
      "[0.64478848]\n",
      "[0.63350876]\n",
      "[0.64649186]\n",
      "[0.6324008]\n",
      "[0.63304791]\n",
      "[0.63217469]\n",
      "[0.63461661]\n",
      "[0.63196236]\n",
      "[0.63257742]\n",
      "[0.63207753]\n",
      "[0.63228869]\n",
      "[0.63202489]\n",
      "[0.63187016]\n",
      "[0.63193625]\n",
      "[0.63180848]\n",
      "[0.63187296]\n",
      "[0.63170225]\n",
      "[0.63172258]\n",
      "[0.63161325]\n",
      "[0.63164399]\n",
      "[0.63180886]\n",
      "[0.63168961]\n",
      "[0.63142763]\n",
      "[0.63125719]\n",
      "[0.63138265]\n",
      "[0.63151567]\n",
      "[0.63106659]\n",
      "[0.63083513]\n",
      "[0.6310257]\n",
      "[0.63065963]\n",
      "[0.6304322]\n",
      "[0.63018059]\n",
      "[0.62979105]\n",
      "[0.62974111]\n",
      "[0.62951362]\n",
      "[0.62939448]\n",
      "[0.62977114]\n",
      "[0.62927715]\n",
      "[0.62995951]\n",
      "[0.62945661]\n",
      "[0.6296159]\n",
      "[0.62927973]\n",
      "[0.62959765]\n",
      "[0.62925844]\n",
      "[0.6299096]\n",
      "[0.62922096]\n",
      "[0.62927221]\n",
      "[0.62925583]\n",
      "[0.62924853]\n",
      "[0.62943703]\n",
      "[0.62921212]\n",
      "[0.62928038]\n",
      "[0.62922342]\n",
      "[0.62922157]\n",
      "[0.62920604]\n",
      "[0.62922162]\n",
      "[0.62923122]\n",
      "[0.6292078]\n",
      "[0.62926207]\n",
      "[0.62920576]\n",
      "[0.62919628]\n",
      "[0.62919041]\n",
      "[0.62919768]\n",
      "[0.62920993]\n",
      "[0.62919878]\n",
      "[0.62918476]\n",
      "[0.62917732]\n",
      "[0.62917669]\n",
      "[0.62916965]\n",
      "[0.62915459]\n",
      "[0.62913402]\n",
      "[0.62913451]\n",
      "[0.62911444]\n",
      "[0.62909134]\n",
      "[0.62906112]\n",
      "[0.62900964]\n",
      "[0.62902274]\n",
      "[0.6289476]\n",
      "[0.62886882]\n",
      "[0.62883061]\n",
      "[0.62870972]\n",
      "[0.62869001]\n",
      "[0.62853818]\n",
      "[0.62838672]\n",
      "[0.62809607]\n",
      "[0.62802091]\n",
      "[0.62764025]\n",
      "[0.62747936]\n",
      "[0.62694797]\n",
      "[0.6264975]\n",
      "[0.62549948]\n",
      "[0.62537277]\n",
      "[0.62424149]\n",
      "[0.62368407]\n",
      "[0.62221477]\n",
      "[0.62087855]\n",
      "[0.61824474]\n",
      "[0.61865003]\n",
      "[0.6164344]\n",
      "[0.61477273]\n",
      "[0.61284713]\n",
      "[0.61088094]\n",
      "[0.61000689]\n",
      "[0.60847679]\n",
      "[0.61504093]\n",
      "[0.61158437]\n",
      "[0.60957012]\n",
      "[0.60645127]\n",
      "[0.60454545]\n",
      "[0.60504381]\n",
      "[0.60242732]\n",
      "[0.60144212]\n",
      "[0.5970224]\n",
      "[0.59312798]\n",
      "[0.60310949]\n",
      "[0.59373869]\n",
      "[0.59003497]\n",
      "[0.58986093]\n",
      "[0.58016931]\n",
      "[0.57485423]\n",
      "[0.59423057]\n",
      "[0.58577453]\n",
      "[0.56954539]\n",
      "[0.5594415]\n",
      "[0.56163927]\n",
      "[0.54976896]\n",
      "[0.54312706]\n",
      "[0.52706731]\n",
      "[0.50720353]\n",
      "[0.5240248]\n",
      "[0.51052367]\n",
      "[0.48681022]\n",
      "[0.47077161]\n",
      "[0.46193227]\n",
      "[0.43946351]\n",
      "[0.42793773]\n",
      "[0.40212028]\n",
      "[0.39146481]\n",
      "[0.37741404]\n",
      "[0.32910734]\n",
      "[0.2860379]\n",
      "[0.33356913]\n",
      "[0.30835045]\n",
      "[0.25668871]\n",
      "[0.2407715]\n",
      "[0.22890664]\n",
      "[0.22417799]\n",
      "[0.20776723]\n",
      "[0.25450966]\n",
      "[0.2230383]\n",
      "[0.25208244]\n",
      "[0.21223961]\n",
      "[0.22349222]\n",
      "[0.21083252]\n",
      "[0.22217215]\n",
      "[0.20587032]\n",
      "[0.21990795]\n",
      "[0.20544081]\n",
      "[0.2259841]\n",
      "[0.20475102]\n",
      "[0.2148927]\n",
      "[0.20403079]\n",
      "[0.20735867]\n",
      "[0.20392581]\n",
      "[0.20551259]\n",
      "[0.20402779]\n",
      "[0.20639077]\n",
      "[0.2037353]\n",
      "[0.20579527]\n",
      "[0.20354952]\n",
      "[0.20375201]\n",
      "[0.20398409]\n",
      "[0.20359489]\n",
      "[0.20393453]\n",
      "[0.20354185]\n",
      "[0.20423421]\n",
      "[0.2035265]\n",
      "[0.20357164]\n",
      "[0.20351645]\n",
      "[0.20357015]\n",
      "[0.20350985]\n",
      "[0.20357081]\n",
      "[0.20350649]\n",
      "[0.20361187]\n",
      "[0.20350033]\n",
      "[0.20350931]\n",
      "[0.20351274]\n",
      "[0.20350145]\n",
      "[0.20351161]\n",
      "[0.20350103]\n",
      "[0.20350324]\n",
      "[0.20349884]\n",
      "[0.20350452]\n",
      "[0.20349838]\n",
      "[0.20350432]\n",
      "[0.20349824]\n",
      "[0.20350506]\n",
      "[0.20349803]\n",
      "[0.20349852]\n",
      "[0.20349788]\n",
      "[0.20349819]\n",
      "[0.20349862]\n",
      "[0.20349776]\n",
      "[0.203498]\n",
      "[0.20349988]\n",
      "[0.2034976]\n",
      "[0.20349795]\n",
      "[0.20349771]\n",
      "[0.2034981]\n",
      "[0.20349767]\n",
      "[0.20349757]\n",
      "[0.20349764]\n",
      "[0.2034976]\n",
      "[0.20349775]\n",
      "[0.20349758]\n",
      "[0.20349767]\n",
      "[0.20349756]\n",
      "[0.20349762]\n",
      "[0.20349757]\n",
      "[0.20349759]\n",
      "[0.20349756]\n",
      "[0.20349759]\n",
      "[0.20349756]\n",
      "[0.20349757]\n",
      "[0.20349756]\n",
      "[0.20349757]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "[0.20349756]\n",
      "Optimization terminated successfully.\n",
      "         Current function value: 0.203498\n",
      "         Iterations: 157\n",
      "         Function evaluations: 287\n",
      "[-25.16130062   0.20623142   0.20147143]\n",
      "Not admitted\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "theta_result, cost = findMinTheta(theta, X, y)\n",
    "print(theta_result)\n",
    "\n",
    "\n",
    "plot_x = np.array([30, 100])\n",
    "plot_y = -1/theta_result[2] * (theta_result[1] * plot_x + theta_result[0])\n",
    "\n",
    "#theta_result[2] * plot_y + theta_result[1] * plot_x + theta_result[0]= 0\n",
    "\n",
    "plt.plot(plot_x, plot_y, 'y-')\n",
    "\n",
    "pos1 = [i for i in range(len(y)) if y[i]==1]\n",
    "pos0 = [i for i in range(len(y)) if y[i]==0]\n",
    "\n",
    "X1=np.array(X1)\n",
    "a,b=np.split(X1, 2, axis=1)\n",
    "a=a.flatten()\n",
    "b=b.flatten()\n",
    "\n",
    "plt.scatter(a[pos1], b[pos1], label='Admitted')\n",
    "plt.scatter(a[pos0], b[pos0], label='Not Admitted')\n",
    "######### Evaluation #############\n",
    "\n",
    "def eval(x, theta):\n",
    "    x = np.array(x)\n",
    "    H = sigmoid(x.dot(theta))\n",
    "\n",
    "    if H > 0.5:\n",
    "        print(\"admitted\")\n",
    "    else:\n",
    "        print(\"Not admitted\")\n",
    "\n",
    "    plt.plot(x[1], x[2], 'g*',markersize=15)\n",
    "\n",
    "test_1 = [1, 70, 40]\n",
    "\n",
    "eval(test_1, theta_result)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}