Lab5 9Sep2021.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "cca1374f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aaf02fff",
   "metadata": {},
   "source": [
    "### Read the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "8b733597",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(506, 13)\n",
      "(506,)\n"
     ]
    }
   ],
   "source": [
    "data = pd.read_csv('./TrainingData/housing.csv', header=None, delimiter=\"\\s+\")\n",
    "data = data.values\n",
    "X = data[:, :13]\n",
    "Y = data[:,13]\n",
    "print(X.shape)\n",
    "print(Y.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0c715885",
   "metadata": {},
   "source": [
    "### Normalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "f97ae523",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(13,) (13,)\n"
     ]
    }
   ],
   "source": [
    "u = np.mean(X,axis=0)\n",
    "std = np.std(X,axis=0)\n",
    "print(u.shape, std.shape)\n",
    "X = (X-u)/ std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "c7558b5d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x1203d4b20>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.style.use('seaborn')\n",
    "plt.scatter(X[:,5],Y)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5471544b",
   "metadata": {},
   "source": [
    "### Add X0 to X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "e011a512",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(506, 14)\n"
     ]
    }
   ],
   "source": [
    "ones = np.ones((X.shape[0],1))\n",
    "X = np.hstack((ones,X))\n",
    "print(X.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "5d8a14c1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 1.         -0.41978194  0.28482986 ... -1.45900038  0.44105193\n",
      "  -1.0755623 ]\n",
      " [ 1.         -0.41733926 -0.48772236 ... -0.30309415  0.44105193\n",
      "  -0.49243937]\n",
      " [ 1.         -0.41734159 -0.48772236 ... -0.30309415  0.39642699\n",
      "  -1.2087274 ]\n",
      " ...\n",
      " [ 1.         -0.41344658 -0.48772236 ...  1.17646583  0.44105193\n",
      "  -0.98304761]\n",
      " [ 1.         -0.40776407 -0.48772236 ...  1.17646583  0.4032249\n",
      "  -0.86530163]\n",
      " [ 1.         -0.41500016 -0.48772236 ...  1.17646583  0.44105193\n",
      "  -0.66905833]]\n"
     ]
    }
   ],
   "source": [
    "print(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "4eca5a0e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def hypothesis(x,theta) :\n",
    "    n = x.shape[0]\n",
    "    y_ = 0\n",
    "    \n",
    "    for i in range(n) :\n",
    "        y_ += theta[i] * x[i]\n",
    "        \n",
    "    return y_\n",
    "        \n",
    "def error(X,Y,theta) :\n",
    "    \n",
    "    e = 0.0\n",
    "    m = X.shape[0]\n",
    "    \n",
    "    for i in range(m) :\n",
    "        y = Y[i]\n",
    "        y_ = hypothesis(X[i],theta)\n",
    "        \n",
    "        e += (y - y_) ** 2\n",
    "        \n",
    "    return e / m\n",
    "\n",
    "def gradient(X,Y,theta) :\n",
    "    \n",
    "    m,n = X.shape\n",
    "    grad = np.zeros((n,))\n",
    "    \n",
    "    for j in range(n) :\n",
    "        \n",
    "        res = 0.0\n",
    "        for i in range(m) :\n",
    "            \n",
    "            y = Y[i]\n",
    "            y_ = hypothesis(X[i],theta)\n",
    "            \n",
    "            res += (y_ - y)*X[i][j]\n",
    "            \n",
    "        grad[j] = res\n",
    "        \n",
    "    return grad/m\n",
    "\n",
    "def gradient_descent(X,Y,lr=0.1,max_epoch=300) :\n",
    "    \n",
    "    m,n = X.shape\n",
    "    theta = np.zeros((n,))\n",
    "    error_list = []\n",
    "    \n",
    "    for i in range(max_epoch) :\n",
    "        \n",
    "        e = error(X,Y,theta)\n",
    "        error_list.append(e)\n",
    "        \n",
    "        grad = gradient(X,Y,theta) \n",
    "        \n",
    "        for j in range(n) :\n",
    "            theta[j] = theta[j] - lr * grad[j]\n",
    "    \n",
    "    \n",
    "    \n",
    "    return theta, error_list"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8ebc8504",
   "metadata": {},
   "source": [
    "### Obtain theta "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "d6c9be00",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 2.25328063e+01 -9.03091692e-01  1.03815625e+00  1.53477685e-02\n",
      "  6.99554920e-01 -2.02101672e+00  2.70014278e+00 -1.93085233e-03\n",
      " -3.10234837e+00  2.34354753e+00 -1.72031485e+00 -2.04614394e+00\n",
      "  8.47845679e-01 -3.73089521e+00]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x120679550>]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "theta,error_list = gradient_descent(X,Y)\n",
    "print(theta)\n",
    "plt.plot(error_list)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "30fac0a5",
   "metadata": {},
   "source": [
    "### Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "896b7428",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_list = []\n",
    "\n",
    "m = X.shape[0]\n",
    "\n",
    "for i in range(m) :\n",
    "    pred = hypothesis(X[i],theta)\n",
    "    y_pred_list.append(pred)\n",
    "    \n",
    "y_pred_list = np.array(y_pred_list)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0271c8e5",
   "metadata": {},
   "source": [
    "### R2 Score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "26a71b7b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def r2_score(Y, Y_) :\n",
    "    num = np.sum((Y - Y_)**2)\n",
    "    denom = np.sum((Y - Y.mean())**2)\n",
    "    score = 1 - num/denom\n",
    "    \n",
    "    return score * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "3e78ff78",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "74.04541323942743"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r2_score(Y,y_pred_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cae14777",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.8.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}