LogisticRegression_simple_py3_oop.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2 Logistic Regression OOP Way"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt \n",
    "import pprint"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10, 20, 30, 40, 50, 60, 70]\n",
      "[0, 0, 0, 1, 1, 1, 1]\n"
     ]
    }
   ],
   "source": [
    "x=[10, 20, 30, 40, 50, 60, 70]\n",
    "y=[0, 0, 0, 1, 1, 1, 1]\n",
    "print(x)\n",
    "print(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.0, 10.0], [1.0, 20.0], [1.0, 30.0], [1.0, 40.0], [1.0, 50.0], [1.0, 60.0], [1.0, 70.0]]\n"
     ]
    }
   ],
   "source": [
    "xx=[ [1.0, float(ix) ] for ix in x]\n",
    "print(xx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7, 2)\n",
      "<class 'numpy.ndarray'>\n",
      "[[ 1. 10.]\n",
      " [ 1. 20.]\n",
      " [ 1. 30.]\n",
      " [ 1. 40.]\n",
      " [ 1. 50.]\n",
      " [ 1. 60.]\n",
      " [ 1. 70.]]\n"
     ]
    }
   ],
   "source": [
    "X=np.array(xx)\n",
    "y=np.array(y)\n",
    "print(X.shape)\n",
    "print ( type(X))\n",
    "print(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7, 2)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "7\n"
     ]
    }
   ],
   "source": [
    "print(X.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAFFdJREFUeJzt3X+U5XV93/HnC1YDKwgqk0hZYNBsUIqJwEig1oQYaRaqi6a2geKJplROW4nxaNqDR6FKJTnRnqhJaXQ1RosLiFjDlmCJQUx6rOLOgiKw3eMW+TEHlPE3laNIfPeP++XDdZidmV3mO3fu8nycM+fe7+f7me99v89e5sX3+73f701VIUkSwD6jLkCStHoYCpKkxlCQJDWGgiSpMRQkSY2hIElqDAVJUmMoSJIaQ0GS1BgKEpDk2Um+neT4bvkfJPlmklPmmXt+kqvmjL03yZ90z1+T5I4kDyT5WpKz59nGM5M8mOQZQ2MnJJlN8qRlb1BaonibC2kgyWuBNwInAJ8EvlJVvz/PvCOB7cAzq+r7SfYFZoBXAF8B7gNeUFU7khwKPL2qbptnO9cC/6Oq/qxbfjewpqp+t58OpcW5pyB1quoDwFeBG4FDgbfsYt5dwE3Ay7uhFwMPVtUXuuWfAMcm2b+q7psvEDofAV4F0AXLWcCly9GLtKcMBemnfQA4FvjTqvrRAvMuY/BHHOBfdstU1Q+A3wL+DXBfkr9K8pxdbONq4JgkzwJOBb5XVV9chh6kPebhI6mT5ADgy8ANwGnA86rq27uYOwHcDawHbgVOrqrtc+bsD7wDOLGqXrSL7byfwaGn5wDbq+ody9SOtEfcU5Ae9V5gW1X9a+CvgPftamJVzQKfBf4C+NojgZDk55JsTPIU4EfA/wP+foHX/G/Aa4CNwEeXoQfpcTEUJCDJGcAGBod9YHDC+fj5Pjk05DLgJd3jI/YB3gTcC3wb+FXg3+1qA1X1OQbnIG6qqjv3tH5puXj4SBqxJJ8BLquqD466FslQkEYoyQuATwOHV9UDo65H8vCRNCJJPgL8DfAGA0GrhXsKkqTGPQVJUrNm1AXsrkMOOaQmJydHXYYkjZVt27Z9s6omFps3dqEwOTnJ9PT0qMuQpLGS5K6lzPPwkSSpMRQkSY2hIElqDAVJUmMoSJIaQ0GS1BgKkqTGUJAkNb2FQpIPJbk/ya27WJ8kf5JkZ5JbkhzfVy1SrzZvhslJ2GefwePmzaOuaM/Zy+q0kr1UVS8/wK8AxwO37mL96cCngAAnATcuZbsnnHBCSavGRz9atXZtFTz6s3btYHzc2MvqtEy9ANO1hL+xvd4lNckkcE1VHTvPuvcDn62qy7vlHcApVXXfQtucmpoqb3OhVWNyEu6a5+4BRx4Jd9650tU8PvayOi1TL0m2VdXUYvNGeU7hMOCeoeWZbuwxkpybZDrJ9Ozs7IoUJy3J3Xfv3vhqZi+r0wr3MspQyDxj8+62VNWmqpqqqqmJiUVv8ietnCOO2L3x1cxeVqcV7mWUoTADHD60vI7Bl51L4+Pii2Ht2p8eW7t2MD5u7GV1WuFeRhkKW4Df7j6FdBLwvcXOJ0irztlnw6ZNg+O7yeBx06bB+Lixl9VphXvp7URzksuBU4BDgG8A/xF4EkBVvS9JgP8CbAAeBH6nqhY9g+yJZknafUs90dzbl+xU1VmLrC/gdX29viRp93lFsySpMRQkSY2hIElqDAVJUmMoSJIaQ0GS1BgKkqTGUJAkNYaCJKkxFCRJjaEgSWoMBUlSYyhIkhpDQZLUGAqSpMZQkCQ1hoIkqTEUJEmNoSBJagwFSVJjKEiSGkNBktQYCpKkxlCQJDWGgiSpMRQkSY2hIElqDAVJUmMoSJKaXkMhyYYkO5LsTHL+POuPSHJDkpuT3JLk9D7rkSQtrLdQSLIvcAlwGnAMcFaSY+ZMeytwZVUdB5wJ/Ne+6pEkLa7PPYUTgZ1VdUdVPQRcAZwxZ04BT+2eHwTc22M9kqRF9BkKhwH3DC3PdGPD3ga8KskMcC3wu/NtKMm5SaaTTM/OzvZRqySJfkMh84zVnOWzgA9X1TrgdODSJI+pqao2VdVUVU1NTEz0UKokCfoNhRng8KHldTz28NA5wJUAVfV5YD/gkB5rkiQtoM9Q2AqsT3JUkiczOJG8Zc6cu4FfB0jyXAah4PEhSRqR3kKhqh4GzgOuA7Yz+JTRbUkuSrKxm/Ym4LVJvgxcDrymquYeYpIkrZA1fW68qq5lcAJ5eOzCoee3Ay/sswZJ0tJ5RbMkqTEUJEmNoSBJagwFSVJjKEiSGkNBktQYCpKkxlCQJDWGgiSpMRQkSY2hIElqDAVJUmMoSJIaQ0GS1BgKkqTGUJAkNYaCJKkxFCRJjaEgSWoMBUlSYyhIkhpDQZLUGAqSpMZQkCQ1hoIkqTEUJEmNoSBJagwFSVJjKEiSGkNBktT0GgpJNiTZkWRnkvN3MedfJLk9yW1JLuuzHknSwtb0teEk+wKXAKcCM8DWJFuq6vahOeuBNwMvrKrvJPnZvuqRJC2uzz2FE4GdVXVHVT0EXAGcMWfOa4FLquo7AFV1f4/1SJIW0WcoHAbcM7Q8040N+wXgF5J8LskXkmyYb0NJzk0ynWR6dna2p3IlSX2GQuYZqznLa4D1wCnAWcAHkxz8mF+q2lRVU1U1NTExseyFSpIG+gyFGeDwoeV1wL3zzLm6qn5cVV8DdjAICUnSCPQZCluB9UmOSvJk4Exgy5w5fwn8GkCSQxgcTrqjx5okSQvoLRSq6mHgPOA6YDtwZVXdluSiJBu7adcB30pyO3AD8O+r6lt91SRJWliq5h7mX92mpqZqenp61GVI0lhJsq2qphabt+h1Ckn2A84B/iGw3yPjVfWvHleFkqRVZymHjy4Fngn8BvC3DE4YP9BnUZKk0VhKKPx8VV0A/KCqPgL8U+B5/ZYlSRqFpYTCj7vH7yY5FjgImOytIknSyCzl3kebkjwNeCuDj5QeAFzQa1WSpJFYSihc392b6O+AZwEkOarXqiRJI7GUw0efmGfsquUuRJI0ervcU0jyHAYfQz0oyW8OrXoqQx9NlSTtPRY6fHQ08FLgYOBlQ+MPMLjltSRpL7PLUKiqq4Grk5xcVZ9fwZokSSOylBPNNyd5HV7RLEl7Pa9oliQ1XtEsSWq8olmS1OzOFc0X8OgVzRf2WpUkaSQWDYWq+mD39G/prmiWJO2dFrp47Y0L/WJV/fHylyNJGqWF9hQO7B6PBl7Ao9+v/DIG90GSJO1lFrp47e0ASf4aOL6qHuiW3wZ8fEWqkyStqKV8+ugI4KGh5Yfw00eStFdayqePLgW+mOSTQAGvAD7Sa1WSpJFYyqePLk7yKeBF3dDvVNXN/ZYlSRqFpewpUFU3ATf1XIskacSWck5BkvQEYShIkppFQyHJed1tLiRJe7ml7Ck8E9ia5MokG5Kk76IkSaOxaChU1VuB9cCfA68BvprkD5I8u+faJEkrbEnnFKqqgK93Pw8DTwOuSvLOHmuTJK2wpZxTeH2SbcA7gc8Bz6uqfwucAPyzRX53Q5IdSXYmOX+Bea9MUkmmdrN+SdIyWsp1CocAv1lVdw0PVtVPkrx0V7+UZF/gEuBUYIbBeYktVXX7nHkHAq8Hbtzd4iVJy2sp5xQunBsIQ+u2L/CrJwI7q+qOqnoIuAI4Y555/4nBXsgPl1CvJKlHfV6ncBhwz9DyTDfWJDkOOLyqrlloQ0nOTTKdZHp2dnb5K5UkAf2GwnwfXa22MtkHeDfwpsU2VFWbqmqqqqYmJiaWsURJ0rA+Q2EGOHxoeR1w79DygcCxwGeT3AmcBGzxZLMkjU6fobAVWJ/kqCRPBs7k0W9vo6q+V1WHVNVkVU0CXwA2VtV0jzVJkhbQWyhU1cPAecB1wHbgyqq6LclFSTb29bqSpD23pFtn76mquha4ds7YhbuYe0qftUiSFuddUiVJjaEgSWoMBUlSYyhIkhpDQZLUGAqSpMZQkCQ1hoIkqTEUJEmNoSBJagwFSVJjKEiSGkNBktQYCpKkxlCQJDWGgiSpMRQkSY2hIElqDAVJUmMoSJIaQ0GS1BgKkqTGUJAkNYaCJKkxFCRJjaEgSWoMBUlSYyhIkhpDQZLUGAqSpKbXUEiyIcmOJDuTnD/P+jcmuT3JLUmuT3Jkn/VIkhbWWygk2Re4BDgNOAY4K8kxc6bdDExV1S8CVwHv7KseSdLi+txTOBHYWVV3VNVDwBXAGcMTquqGqnqwW/wCsK7HeiRJi+gzFA4D7hlanunGduUc4FPzrUhybpLpJNOzs7PLWKIkaVifoZB5xmreicmrgCngXfOtr6pNVTVVVVMTExPLWKIkadiaHrc9Axw+tLwOuHfupCQvAd4C/GpV/ajHeiRJi+hzT2ErsD7JUUmeDJwJbBmekOQ44P3Axqq6v8daJElL0FsoVNXDwHnAdcB24Mqqui3JRUk2dtPeBRwAfDzJl5Js2cXmJEkroM/DR1TVtcC1c8YuHHr+kj5fX5K0e7yiWZLUGAqSpMZQkCQ1hoIkqTEUJEmNoSBJagwFSVJjKEiSGkNBktQYCpKkxlCQJDWGgiSpMRQkSY2hIElqDAVJUmMoSJIaQ0GS1BgKkqTGUJAkNYaCJKkxFCRJjaEgSWoMBUlSYyhIkhpDQZLUGAqSpMZQkCQ1hoIkqTEUJElNr6GQZEOSHUl2Jjl/nvU/k+Rj3fobk0z2UsjmzTA5CfvsM3jcvLmXl1kR9iKpR2v62nCSfYFLgFOBGWBrki1VdfvQtHOA71TVzyc5E/gj4LeWtZDNm+Hcc+HBBwfLd901WAY4++xlfane2YuknqWq+tlwcjLwtqr6jW75zQBV9YdDc67r5nw+yRrg68BELVDU1NRUTU9PL72QycnBH5y5jjwS7rxz6dtZDexF0h5Ksq2qphab1+fho8OAe4aWZ7qxeedU1cPA94BnzN1QknOTTCeZnp2d3b0q7r5798ZXM3uR1LM+QyHzjM3dA1jKHKpqU1VNVdXUxMTE7lVxxBG7N76a2YuknvUZCjPA4UPL64B7dzWnO3x0EPDtZa3i4oth7dqfHlu7djA+buxFUs/6DIWtwPokRyV5MnAmsGXOnC3Aq7vnrwQ+s9D5hD1y9tmwadPgWHUyeNy0aTxPZtqLpJ71dqIZIMnpwHuAfYEPVdXFSS4CpqtqS5L9gEuB4xjsIZxZVXcstM3dPtEsSVryiebePpIKUFXXAtfOGbtw6PkPgX/eZw2SpKXzimZJUmMoSJIaQ0GS1BgKkqTGUJAkNYaCJKkxFCRJTa8Xr/UhySwwz+01l+QQ4JvLWM4o2cvqs7f0AfayWj2eXo6sqkVvHjd2ofB4JJleyhV948BeVp+9pQ+wl9VqJXrx8JEkqTEUJEnNEy0UNo26gGVkL6vP3tIH2Mtq1XsvT6hzCpKkhT3R9hQkSQswFCRJzV4bCkk+lOT+JLcOjT09yaeTfLV7fNooa1yKJIcnuSHJ9iS3Jfm9bnwce9kvyReTfLnr5e3d+FFJbux6+Vj3TX1jIcm+SW5Ock23PJa9JLkzyVeSfCnJdDc2ju+xg5NcleT/dP/NnDymfRzd/Vs88vP9JG9YiV722lAAPgxsmDN2PnB9Va0Hru+WV7uHgTdV1XOBk4DXJTmG8ezlR8CLq+qXgOcDG5KcBPwR8O6ul+8A54ywxt31e8D2oeVx7uXXqur5Q5+DH8f32HuB/1lVzwF+icG/zdj1UVU7un+L5wMnAA8Cn2QleqmqvfYHmARuHVreARzaPT8U2DHqGvegp6uBU8e9F2AtcBPwywyu0FzTjZ8MXDfq+pbYw7ruP8wXA9cAGeNe7gQOmTM2Vu8x4KnA1+g+QDOufczT1z8BPrdSvezNewrz+bmqug+ge/zZEdezW5JMMvg+6xsZ0166wy1fAu4HPg38X+C7VfVwN2UGOGxU9e2m9wD/AfhJt/wMxreXAv46ybYk53Zj4/YeexYwC/xFd0jvg0mewvj1MdeZwOXd8957eaKFwthKcgDwCeANVfX9Udezp6rq72uwS7wOOBF47nzTVraq3ZfkpcD9VbVteHieqau+l84Lq+p44DQGhyh/ZdQF7YE1wPHAn1XVccAPGINDRQvpzkltBD6+Uq/5RAuFbyQ5FKB7vH/E9SxJkicxCITNVfXfu+Gx7OURVfVd4LMMzpMcnGRNt2odcO+o6toNLwQ2JrkTuILBIaT3MJ69UFX3do/3Mzh2fSLj9x6bAWaq6sZu+SoGITFufQw7Dbipqr7RLffeyxMtFLYAr+6ev5rB8flVLUmAPwe2V9UfD60ax14mkhzcPd8feAmDE4E3AK/spo1FL1X15qpaV1WTDHbvP1NVZzOGvSR5SpIDH3nO4Bj2rYzZe6yqvg7ck+TobujXgdsZsz7mOItHDx3BSvQy6pMoPZ6cuRy4D/gxg/+DOIfBMd/rga92j08fdZ1L6OMfMzgEcQvwpe7n9DHt5ReBm7tebgUu7MafBXwR2MlgN/lnRl3rbvZ1CnDNuPbS1fzl7uc24C3d+Di+x54PTHfvsb8EnjaOfXS9rAW+BRw0NNZ7L97mQpLUPNEOH0mSFmAoSJIaQ0GS1BgKkqTGUJAkNYaC9DgleVuS319kzsu7GxlKq5qhIK2MlwOGglY9Q0EakuQFSW7pvvvhKd33Phw7z7y3JNmR5G+Ao4fGX5tka/edEZ9IsjbJP2Jw/5p3dffGf/Z881awTWmXvHhNmiPJO4D9gP0Z3EvnD+esP4HB93X8MoObsN0EvK+q/nOSZ1TVt4a2842q+tMkH2Zw1fNV3bp5561Ig9IC1iw+RXrCuQjYCvwQeP08618EfLKqHgRIsmVo3bHdH/mDgQOA63bxGkudJ60oDx9Jj/V0Bn+oD2SwxzCfXe1ifxg4r6qeB7x9gd9f6jxpRRkK0mNtAi4ANjP4es25/g54RZL9u7uLvmxo3YHAfd3tzs8eGn+gW7fYPGmkPHwkDUny28DDVXVZkn2B/53kxVX1mUfmVNVNST7G4I61dwH/a2gTFzD4Zry7gK/waBBcAXwgyesZ3Fp7V/OkkfJEsySp8fCRJKkxFCRJjaEgSWoMBUlSYyhIkhpDQZLUGAqSpOb/A02ojuEmAaS0AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(x, y, color = 'red') \n",
    "plt.title('x vs y')\n",
    "plt.xlabel('x data')\n",
    "plt.ylabel('y data')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "class MyLogisticRegression():\n",
    "    def __init__(self):\n",
    "        print('mylogisitic regression')\n",
    "    def sigmoid(self, z):\n",
    "        return 1 / (1 + np.exp(-z))\n",
    "    \n",
    "    def linRG(self, x, theta):\n",
    "        return np.dot(theta, x.T)\n",
    "\n",
    "    def predict(self, X, theta):\n",
    "      z = self.linRG(X, theta)\n",
    "      return self.sigmoid(z)\n",
    "    def loss(self, h, y):\n",
    "        return (- y * np.log(h) - (1 - y ) * np.log(1 - h) ).mean()\n",
    "    def gradient(self, x, h, y):\n",
    "        n = x.shape[0]\n",
    "        return (1 / n) * np.dot( (h - y), X)\n",
    "    def train(self, X,y, epoch, lr, theta):\n",
    "        for i in range(epoch):\n",
    "            h=self.predict(X, theta)  \n",
    "            gd=self.gradient(X, h, y)\n",
    "            theta = theta-lr * gd \n",
    "        return theta\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mylogisitic regression\n",
      "[-29.28835743   0.84097994]\n"
     ]
    }
   ],
   "source": [
    "mg=MyLogisticRegression()\n",
    "epoch=90000 \n",
    "lr=0.1\n",
    "theta = np.zeros(2)\n",
    "theta=mg.train(X, y, epoch, lr, theta)\n",
    "print(theta)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[8.56165004e-10 3.84490213e-06 1.69738945e-02 9.87268217e-01\n",
      " 9.99997128e-01 9.99999999e-01 1.00000000e+00]\n",
      "[0, 0, 0, 1, 1, 1, 1]\n"
     ]
    }
   ],
   "source": [
    "pd=mg.predict(X, theta)\n",
    "print(pd)\n",
    "prd=[ 1 if ipd>0.5 else 0 for ipd in pd]\n",
    "print(prd)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Iris Dataset [Binary, First two Species (first 100/150 of data) ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "iris =  datasets.load_iris() \n",
    "X = iris.data[:100]  #all columns\n",
    "y = iris.target[:100]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[5.1 3.5 1.4 0.2]\n",
      " [4.9 3.  1.4 0.2]\n",
      " [4.7 3.2 1.3 0.2]]\n",
      "[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n",
      " 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1]\n"
     ]
    }
   ],
   "source": [
    "# print( iris.data[0])\n",
    "print (X[:3])\n",
    "print(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(100, 4)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.  5.1 3.5 1.4 0.2]\n",
      " [1.  4.9 3.  1.4 0.2]\n",
      " [1.  4.7 3.2 1.3 0.2]]\n"
     ]
    }
   ],
   "source": [
    "ix=X[:]\n",
    "\n",
    "ix2=np.ones((ix.shape[0],  1))\n",
    "ix2=np.append(ix2, ix , axis=1)\n",
    "# print(ix2[:3])\n",
    "X=ix2\n",
    "print(X[:3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5\n"
     ]
    }
   ],
   "source": [
    "print(X[0].shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 0. 0. 0. 0.]\n"
     ]
    }
   ],
   "source": [
    "theta = np.zeros(X[0].shape[0])\n",
    "print(theta)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mylogisitic regression\n",
      "[-0.71147732 -1.14034983 -4.04159746  6.31206632  2.96447301]\n"
     ]
    }
   ],
   "source": [
    "mg=MyLogisticRegression()\n",
    "epoch=90000 \n",
    "lr=0.1\n",
    "theta = np.zeros(X[0].shape[0])\n",
    "theta=mg.train(X, y, epoch, lr, theta)\n",
    "print(theta)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]\n"
     ]
    }
   ],
   "source": [
    "pd=mg.predict(X, theta)\n",
    "# print(pd)\n",
    "prd=[ 1 if ipd>0.5 else 0 for ipd in pd]\n",
    "print(prd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "100.0"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " accuracy_score(y, prd)*100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multiclass Logistic Regression with Iris dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import datasets\n",
    "iris =  datasets.load_iris() \n",
    "X = iris.data\n",
    "y = iris.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[5.1 3.5 1.4 0.2]\n",
      " [4.9 3.  1.4 0.2]\n",
      " [4.7 3.2 1.3 0.2]]\n",
      "[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n",
      " 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2\n",
      " 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2\n",
      " 2 2]\n"
     ]
    }
   ],
   "source": [
    "print(X[:3])\n",
    "print(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.  5.1 3.5 1.4 0.2]\n",
      " [1.  4.9 3.  1.4 0.2]\n",
      " [1.  4.7 3.2 1.3 0.2]]\n",
      "[[1.  5.1 3.5 1.4 0.2]\n",
      " [1.  4.9 3.  1.4 0.2]\n",
      " [1.  4.7 3.2 1.3 0.2]]\n"
     ]
    }
   ],
   "source": [
    "ix=X[:]\n",
    "ix2=np.ones((ix.shape[0],  1))\n",
    "ix2=np.append(ix2, ix , axis=1)\n",
    "print(ix2[:3])\n",
    "X=ix2\n",
    "print(X[:3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "k=3  #classes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n",
      " 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2\n",
      " 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2\n",
      " 2 2]\n"
     ]
    }
   ],
   "source": [
    "print(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1\n",
      " 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n",
      " 1 1]\n"
     ]
    }
   ],
   "source": [
    "t= np.array(y==2, dtype = int)\n",
    "print(t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n",
      " 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0]\n",
      "[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n",
      " 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0]\n",
      "[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1\n",
      " 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n",
      " 1 1]\n"
     ]
    }
   ],
   "source": [
    "for ik in range(k):\n",
    "    tmp_y = np.array(y==ik, dtype = int)\n",
    "    print(tmp_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(5,)\n",
      "[[0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0.]]\n"
     ]
    }
   ],
   "source": [
    "print(X[0].shape)\n",
    "thetas=np.zeros( (k, X[0].shape[0]) )\n",
    "print(thetas)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mylogisitic regression\n"
     ]
    }
   ],
   "source": [
    "mg=MyLogisticRegression()\n",
    "epoch=90000 \n",
    "lr=0.1\n",
    "\n",
    "thetas=np.zeros( (k, X[0].shape[0]) )\n",
    "for ik in range(k):\n",
    "    tmp_y = np.array(y==ik, dtype = int)\n",
    "    theta = np.zeros(X[0].shape[0])\n",
    "    theta=mg.train(X, tmp_y, epoch, lr, theta)\n",
    "    thetas[ik]=theta\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[  0.68633772   1.09979165   3.81317717  -5.98736789  -2.83767492]\n",
      " [  7.37790348  -0.24528697  -2.79648084   1.31362024  -2.77833873]\n",
      " [-14.37536249  -3.78415702  -5.20867241   6.96487626  11.13180607]]\n"
     ]
    }
   ],
   "source": [
    "print(thetas)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd=mg.predict(X, thetas)\n",
    "# print(pd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3, 150)\n"
     ]
    }
   ],
   "source": [
    "print(pd.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2]\n"
     ]
    }
   ],
   "source": [
    "prd=[]\n",
    "for i in range(pd.shape[1]):\n",
    "    \n",
    "    vals= [ pd[0][i], pd[1][i], pd[2][i] ]\n",
    "    m=np.argmax(vals)\n",
    "    prd.append(m)\n",
    "    \n",
    "print(prd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1\n",
      " 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2\n",
      " 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2\n",
      " 2 2]\n"
     ]
    }
   ],
   "source": [
    "mm=np.argmax(pd, axis=0)\n",
    "print(mm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "98.0"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " accuracy_score(y, prd) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1.  1.9 3.  2.4 3.2]\n"
     ]
    }
   ],
   "source": [
    "nv=[1, 1.9 , 3.0 ,   2.4 , 3.2]\n",
    "nv=np.array(nv)\n",
    "print(nv)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9.757718227333956e-05 0.0007345045195560041 0.9999997357062428\n",
      "2\n"
     ]
    }
   ],
   "source": [
    "npd=mg.predict(nv, thetas[0])\n",
    "npd1=mg.predict(nv, thetas[1])\n",
    "npd2=mg.predict(nv, thetas[2])\n",
    "print(npd, npd1, npd2)\n",
    "\n",
    "npdr=[npd, npd1, npd2]\n",
    "npv=np.argmax(npdr)\n",
    "print(npv)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[2 2]\n"
     ]
    }
   ],
   "source": [
    "#Helper code\n",
    "vals=[ [3,2,7],\n",
    "       [4,5,6]\n",
    "     ]\n",
    "print( np.argmax(vals, axis=1)  )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[3 2 7]\n",
      " [4 5 6]]\n",
      "[2 5]\n"
     ]
    }
   ],
   "source": [
    "a=np.array(vals)\n",
    "print(a)\n",
    "print(a[:,1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_thd(th, pd):\n",
    "    prd=[ 1 if ipd>th else 0 for ipd in pd]\n",
    "    return prd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import roc_curve\n",
    "from sklearn.metrics import roc_auc_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "prd=get_thd(0.2, pd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "100.0"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " accuracy_score(y, prd)*100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1.30976437e-05, 1.24112699e-04, 3.69594371e-05, 2.19257467e-04,\n",
       "       9.79929923e-06, 2.22022537e-05, 4.66736948e-05, 4.13389626e-05,\n",
       "       3.28757364e-04, 1.15792283e-04, 7.79295279e-06, 9.76144736e-05,\n",
       "       1.03420003e-04, 2.75342119e-05, 2.21128421e-07, 5.91495306e-07,\n",
       "       1.77781965e-06, 1.76171899e-05, 1.75632023e-05, 9.85136753e-06,\n",
       "       9.25757799e-05, 1.98500660e-05, 1.23816797e-06, 4.74970001e-04,\n",
       "       6.48132369e-04, 3.91232146e-04, 1.40584318e-04, 2.19681817e-05,\n",
       "       1.75061578e-05, 2.45484144e-04, 3.28085582e-04, 4.73967695e-05,\n",
       "       1.44514529e-06, 4.90263045e-07, 1.55742697e-04, 1.39646960e-05,\n",
       "       4.41538247e-06, 8.16535491e-06, 1.16765266e-04, 3.68838923e-05,\n",
       "       1.05035216e-05, 2.36715242e-03, 5.20347045e-05, 1.69781632e-04,\n",
       "       1.65459837e-04, 1.87093653e-04, 1.37682566e-05, 7.78685166e-05,\n",
       "       8.73426572e-06, 3.29426105e-05, 9.99994919e-01, 9.99993265e-01,\n",
       "       9.99999363e-01, 9.99997302e-01, 9.99999203e-01, 9.99998911e-01,\n",
       "       9.99998106e-01, 9.99588490e-01, 9.99997578e-01, 9.99986512e-01,\n",
       "       9.99974075e-01, 9.99988727e-01, 9.99992249e-01, 9.99999458e-01,\n",
       "       9.99573378e-01, 9.99983995e-01, 9.99998795e-01, 9.99975237e-01,\n",
       "       9.99999906e-01, 9.99976921e-01, 9.99999764e-01, 9.99959655e-01,\n",
       "       9.99999972e-01, 9.99999346e-01, 9.99987191e-01, 9.99990467e-01,\n",
       "       9.99999573e-01, 9.99999901e-01, 9.99998730e-01, 9.99349475e-01,\n",
       "       9.99974159e-01, 9.99934663e-01, 9.99951631e-01, 9.99999990e-01,\n",
       "       9.99999040e-01, 9.99992879e-01, 9.99998209e-01, 9.99999462e-01,\n",
       "       9.99972767e-01, 9.99993945e-01, 9.99999023e-01, 9.99998475e-01,\n",
       "       9.99982824e-01, 9.99692089e-01, 9.99995691e-01, 9.99978161e-01,\n",
       "       9.99989161e-01, 9.99989803e-01, 9.96188312e-01, 9.99986399e-01])"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}