Linear_Regression_Gradient_Descent.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Gradient Descent Used for optimization of the Error Function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/Users/jmandivara/Documents/TensorFlow'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os\n",
    "os.getcwd()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from numpy import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  32.50234527,   31.70700585],\n",
       "       [  53.42680403,   68.77759598],\n",
       "       [  61.53035803,   62.5623823 ],\n",
       "       [  47.47563963,   71.54663223],\n",
       "       [  59.81320787,   87.23092513],\n",
       "       [  55.14218841,   78.21151827],\n",
       "       [  52.21179669,   79.64197305],\n",
       "       [  39.29956669,   59.17148932],\n",
       "       [  48.10504169,   75.3312423 ],\n",
       "       [  52.55001444,   71.30087989],\n",
       "       [  45.41973014,   55.16567715],\n",
       "       [  54.35163488,   82.47884676],\n",
       "       [  44.1640495 ,   62.00892325],\n",
       "       [  58.16847072,   75.39287043],\n",
       "       [  56.72720806,   81.43619216],\n",
       "       [  48.95588857,   60.72360244],\n",
       "       [  44.68719623,   82.89250373],\n",
       "       [  60.29732685,   97.37989686],\n",
       "       [  45.61864377,   48.84715332],\n",
       "       [  38.81681754,   56.87721319],\n",
       "       [  66.18981661,   83.87856466],\n",
       "       [  65.41605175,  118.5912173 ],\n",
       "       [  47.48120861,   57.25181946],\n",
       "       [  41.57564262,   51.39174408],\n",
       "       [  51.84518691,   75.38065167],\n",
       "       [  59.37082201,   74.76556403],\n",
       "       [  57.31000344,   95.45505292],\n",
       "       [  63.61556125,   95.22936602],\n",
       "       [  46.73761941,   79.05240617],\n",
       "       [  50.55676015,   83.43207142],\n",
       "       [  52.22399609,   63.35879032],\n",
       "       [  35.56783005,   41.4128853 ],\n",
       "       [  42.43647694,   76.61734128],\n",
       "       [  58.16454011,   96.76956643],\n",
       "       [  57.50444762,   74.08413012],\n",
       "       [  45.44053073,   66.58814441],\n",
       "       [  61.89622268,   77.76848242],\n",
       "       [  33.09383174,   50.71958891],\n",
       "       [  36.43600951,   62.12457082],\n",
       "       [  37.67565486,   60.81024665],\n",
       "       [  44.55560838,   52.68298337],\n",
       "       [  43.31828263,   58.56982472],\n",
       "       [  50.07314563,   82.90598149],\n",
       "       [  43.87061265,   61.4247098 ],\n",
       "       [  62.99748075,  115.2441528 ],\n",
       "       [  32.66904376,   45.57058882],\n",
       "       [  40.16689901,   54.0840548 ],\n",
       "       [  53.57507753,   87.99445276],\n",
       "       [  33.86421497,   52.72549438],\n",
       "       [  64.70713867,   93.57611869],\n",
       "       [  38.11982403,   80.16627545],\n",
       "       [  44.50253806,   65.10171157],\n",
       "       [  40.59953838,   65.56230126],\n",
       "       [  41.72067636,   65.28088692],\n",
       "       [  51.08863468,   73.43464155],\n",
       "       [  55.0780959 ,   71.13972786],\n",
       "       [  41.37772653,   79.10282968],\n",
       "       [  62.49469743,   86.52053844],\n",
       "       [  49.20388754,   84.74269781],\n",
       "       [  41.10268519,   59.35885025],\n",
       "       [  41.18201611,   61.68403752],\n",
       "       [  50.18638949,   69.84760416],\n",
       "       [  52.37844622,   86.09829121],\n",
       "       [  50.13548549,   59.10883927],\n",
       "       [  33.64470601,   69.89968164],\n",
       "       [  39.55790122,   44.86249071],\n",
       "       [  56.13038882,   85.49806778],\n",
       "       [  57.36205213,   95.53668685],\n",
       "       [  60.26921439,   70.25193442],\n",
       "       [  35.67809389,   52.72173496],\n",
       "       [  31.588117  ,   50.39267014],\n",
       "       [  53.66093226,   63.64239878],\n",
       "       [  46.68222865,   72.24725107],\n",
       "       [  43.10782022,   57.81251298],\n",
       "       [  70.34607562,  104.25710159],\n",
       "       [  44.49285588,   86.64202032],\n",
       "       [  57.5045333 ,   91.486778  ],\n",
       "       [  36.93007661,   55.23166089],\n",
       "       [  55.80573336,   79.55043668],\n",
       "       [  38.95476907,   44.84712424],\n",
       "       [  56.9012147 ,   80.20752314],\n",
       "       [  56.86890066,   83.14274979],\n",
       "       [  34.3331247 ,   55.72348926],\n",
       "       [  59.04974121,   77.63418251],\n",
       "       [  57.78822399,   99.05141484],\n",
       "       [  54.28232871,   79.12064627],\n",
       "       [  51.0887199 ,   69.58889785],\n",
       "       [  50.28283635,   69.51050331],\n",
       "       [  44.21174175,   73.68756432],\n",
       "       [  38.00548801,   61.36690454],\n",
       "       [  32.94047994,   67.17065577],\n",
       "       [  53.69163957,   85.66820315],\n",
       "       [  68.76573427,  114.85387123],\n",
       "       [  46.2309665 ,   90.12357207],\n",
       "       [  68.31936082,   97.91982104],\n",
       "       [  50.03017434,   81.53699078],\n",
       "       [  49.23976534,   72.11183247],\n",
       "       [  50.03957594,   85.23200734],\n",
       "       [  48.14985889,   66.22495789],\n",
       "       [  25.12848465,   53.45439421]])"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#df=pd.read_csv('data.csv',sep=',',header=None)\n",
    "points=genfromtxt('data.csv',delimiter=',')\n",
    "points"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "#learning Rate---Tuning nob for the linearRegression\n",
    "#you need to choose a low learning rate intials a min but not too low..\n",
    "#as too learning rate will make your model to convege slow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "leaning_rate=0.0001\n",
    "intial_b=0 #we gonna learn so intial must be 0\n",
    "intial_m=0\n",
    "#y=mx+b "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#no of iterations to run until to reach solution.\n",
    "num_iterations=1000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def gradient_descent(points,leaningrate,starting_b,starting_m,iterations):\n",
    "    print(leaningrate,starting_b,starting_m,iterations)\n",
    "    b=starting_b\n",
    "    m=starting_m\n",
    "    for i in range(iterations):\n",
    "        b,m=step_gradient(b,m,array(points),leaningrate)\n",
    "    print(b)\n",
    "    print(m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "def step_gradient(b_current,m_current,points,learning_rate):\n",
    "    b_gradient=0\n",
    "    m_gradient=0\n",
    "    N=float(len(points))\n",
    "    for i in range(0,len(points)):\n",
    "        x=points[i,0]\n",
    "        y=points[i,1]\n",
    "        b_gradient += -(2/N)*(y-((m_current*x)+b_current))\n",
    "        m_gradient += (2/N)*x*(y-((m_current*x)+b_current))\n",
    "    new_b=b_current-(learning_rate*b_gradient)\n",
    "    #print(new_b)\n",
    "    new_m=m_current-(leaning_rate*m_gradient)\n",
    "    return [new_b,new_m]\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_errorr():\n",
    "    error=0\n",
    "    for i in range(0,len(df)):\n",
    "        x=df[i,0]\n",
    "        y=df[i,1]\n",
    "        totalerr+=(y-(mx+b))**2\n",
    "    return totalerr/float(len(df))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "53.426804033275019"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[1,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0001 0 0 1000\n",
      "0.0889365199374\n",
      "1.47774408519\n"
     ]
    },
    {
     "ename": "TypeError",
     "evalue": "'NoneType' object is not iterable",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-100-40136352a886>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0;34m[\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mm\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mgradient_descent\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpoints\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mleaning_rate\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mintial_b\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mintial_m\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mnum_iterations\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m: 'NoneType' object is not iterable"
     ]
    }
   ],
   "source": [
    "[b,m]=gradient_descent(points,leaning_rate,intial_b,intial_m,num_iterations)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.088936519937413458"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.4777440851894448"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.6.1"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 1,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}