Assignment2_Lab2_Part1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#the graph of decreasing cost function wrt learning rate and epoch\n",
    "#Start with random values of theta1 & theta2 . Analyse best approach for initial values of θ1 & θ2."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train_X=numpy.asarray([1,2,4,3,5])\n",
    "train_Y=numpy.asarray([1,3,3,2,5])\n",
    "n_samples = train_X.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X = tf.placeholder(\"float\")\n",
    "Y = tf.placeholder(\"float\")\n",
    "W = tf.Variable(np.random.randn(), name=\"weight\")\n",
    "b = tf.Variable(np.random.randn(), name=\"bias\")\n",
    "pred = tf.add(tf.multiply(X, W), b)\n",
    "cost = tf.reduce_sum(tf.pow(pred-Y, 2))/(2*n_samples)\n",
    "w_rnd=[]\n",
    "b_rnd=[]\n",
    "ep_rnd=[]\n",
    "cst_rnd=[]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LEARNING RATE= 0.05 \t Epoch= 1000 \n",
      "\n",
      "\n",
      "Epoch: 0050 cost= 0.240961939 W= 0.81534 b= 0.392101\n",
      "Epoch: 0100 cost= 0.240978047 W= 0.819704 b= 0.375231\n",
      "Epoch: 0150 cost= 0.241016150 W= 0.82252 b= 0.364346\n",
      "Epoch: 0200 cost= 0.241052344 W= 0.824337 b= 0.357321\n",
      "Epoch: 0250 cost= 0.241080478 W= 0.82551 b= 0.352788\n",
      "Epoch: 0300 cost= 0.241100669 W= 0.826266 b= 0.349863\n",
      "Epoch: 0350 cost= 0.241114587 W= 0.826755 b= 0.347975\n",
      "Epoch: 0400 cost= 0.241123751 W= 0.82707 b= 0.346757\n",
      "Epoch: 0450 cost= 0.241129920 W= 0.827273 b= 0.345971\n",
      "Epoch: 0500 cost= 0.241134048 W= 0.827404 b= 0.345463\n",
      "Epoch: 0550 cost= 0.241136596 W= 0.827489 b= 0.345136\n",
      "Epoch: 0600 cost= 0.241138384 W= 0.827544 b= 0.344925\n",
      "Epoch: 0650 cost= 0.241139412 W= 0.827579 b= 0.344789\n",
      "Epoch: 0700 cost= 0.241140157 W= 0.827602 b= 0.344701\n",
      "Epoch: 0750 cost= 0.241140559 W= 0.827616 b= 0.344644\n",
      "Epoch: 0800 cost= 0.241140917 W= 0.827626 b= 0.344607\n",
      "Epoch: 0850 cost= 0.241141051 W= 0.827632 b= 0.344584\n",
      "Epoch: 0900 cost= 0.241141245 W= 0.827636 b= 0.344568\n",
      "Epoch: 0950 cost= 0.241141230 W= 0.827639 b= 0.344559\n",
      "Epoch: 1000 cost= 0.241141319 W= 0.82764 b= 0.344553\n",
      "Optimization Finished!\n",
      "Training cost= 0.241141 W= 0.82764 b= 0.344553 \n",
      "\n"
     ]
    }
   ],
   "source": [
    "    \n",
    "    learning_rate=0.05\n",
    "    training_epochs=1000\n",
    "    display_step = 50\n",
    "    print(\"LEARNING RATE=\",learning_rate,\"\\t Epoch=\",training_epochs,\"\\n\\n\")\n",
    "    optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)\n",
    "\n",
    "    init = tf.global_variables_initializer()\n",
    "    with tf.Session() as sess:\n",
    "        sess.run(init)\n",
    "\n",
    "    # Fit all training data\n",
    "        for epoch in range(training_epochs):\n",
    "            for (x, y) in zip(train_X, train_Y):\n",
    "                sess.run(optimizer, feed_dict={X: x, Y: y})\n",
    "\n",
    "        #Display logs per epoch step\n",
    "            if (epoch+1) % display_step == 0:\n",
    "                c = sess.run(cost, feed_dict={X: train_X, Y:train_Y})\n",
    "                print (\"Epoch:\", '%04d' % (epoch+1), \"cost=\", \"{:.9f}\".format(c), \\\n",
    "                \"W=\", sess.run(W), \"b=\", sess.run(b))\n",
    "                w_rnd.append(sess.run(W))\n",
    "                b_rnd.append(sess.run(b))\n",
    "                ep_rnd.append(epoch)\n",
    "                cst_rnd.append(c)\n",
    "\n",
    "        print (\"Optimization Finished!\")\n",
    "        training_cost = sess.run(cost, feed_dict={X: train_X, Y: train_Y})\n",
    "        print (\"Training cost=\", training_cost, \"W=\", sess.run(W), \"b=\", sess.run(b), '\\n')\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LEARNING RATE= 0.05 \t Epoch= 1000 \n",
      "\n",
      "\n",
      "Epoch: 0050 cost= 0.242040664 W= 0.844209 b= 0.280501\n",
      "Epoch: 0100 cost= 0.241635278 W= 0.838333 b= 0.303215\n",
      "Epoch: 0150 cost= 0.241424233 W= 0.834541 b= 0.317873\n",
      "Epoch: 0200 cost= 0.241308883 W= 0.832095 b= 0.327332\n",
      "Epoch: 0250 cost= 0.241243333 W= 0.830516 b= 0.333436\n",
      "Epoch: 0300 cost= 0.241204545 W= 0.829497 b= 0.337375\n",
      "Epoch: 0350 cost= 0.241181165 W= 0.828839 b= 0.339916\n",
      "Epoch: 0400 cost= 0.241166592 W= 0.828415 b= 0.341556\n",
      "Epoch: 0450 cost= 0.241157413 W= 0.828141 b= 0.342615\n",
      "Epoch: 0500 cost= 0.241151616 W= 0.827965 b= 0.343298\n",
      "Epoch: 0550 cost= 0.241148070 W= 0.827851 b= 0.343739\n",
      "Epoch: 0600 cost= 0.241145760 W= 0.827777 b= 0.344023\n",
      "Epoch: 0650 cost= 0.241144150 W= 0.82773 b= 0.344207\n",
      "Epoch: 0700 cost= 0.241143256 W= 0.827699 b= 0.344325\n",
      "Epoch: 0750 cost= 0.241142586 W= 0.827679 b= 0.344402\n",
      "Epoch: 0800 cost= 0.241142109 W= 0.827666 b= 0.344451\n",
      "Epoch: 0850 cost= 0.241141945 W= 0.827658 b= 0.344482\n",
      "Epoch: 0900 cost= 0.241141796 W= 0.827653 b= 0.344503\n",
      "Epoch: 0950 cost= 0.241141632 W= 0.827649 b= 0.344516\n",
      "Epoch: 1000 cost= 0.241141558 W= 0.827647 b= 0.344525\n",
      "Optimization Finished!\n",
      "Training cost= 0.241142 W= 0.827647 b= 0.344525 \n",
      "\n"
     ]
    }
   ],
   "source": [
    "    W = tf.Variable(0.0, name=\"weight\")\n",
    "    b = tf.Variable(0.0, name=\"bias\")\n",
    "    pred = tf.add(tf.multiply(X, W), b)\n",
    "    cost = tf.reduce_sum(tf.pow(pred-Y, 2))/(2*n_samples)\n",
    "    epo=[]\n",
    "    cst=[]\n",
    "    w1=[]\n",
    "    b1=[]\n",
    "    learning_rate=0.05\n",
    "    training_epochs=1000\n",
    "    display_step = 50\n",
    "    print(\"LEARNING RATE=\",learning_rate,\"\\t Epoch=\",training_epochs,\"\\n\\n\")\n",
    "    optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)\n",
    "\n",
    "    init = tf.global_variables_initializer()\n",
    "    with tf.Session() as sess:\n",
    "        sess.run(init)\n",
    "\n",
    "    # Fit all training data\n",
    "        for epoch in range(training_epochs):\n",
    "            for (x, y) in zip(train_X, train_Y):\n",
    "                sess.run(optimizer, feed_dict={X: x, Y: y})\n",
    "\n",
    "        #Display logs per epoch step\n",
    "            if (epoch+1) % display_step == 0:\n",
    "                c = sess.run(cost, feed_dict={X: train_X, Y:train_Y})\n",
    "                print (\"Epoch:\", '%04d' % (epoch+1), \"cost=\", \"{:.9f}\".format(c), \\\n",
    "                \"W=\", sess.run(W), \"b=\", sess.run(b))\n",
    "                epo.append(epoch)\n",
    "                cst.append(c)\n",
    "                w1.append(sess.run(W))\n",
    "                b1.append(sess.run(b))\n",
    "\n",
    "\n",
    "        print (\"Optimization Finished!\")\n",
    "        training_cost = sess.run(cost, feed_dict={X: train_X, Y: train_Y})\n",
    "        print (\"Training cost=\", training_cost, \"W=\", sess.run(W), \"b=\", sess.run(b), '\\n')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAD8CAYAAABpcuN4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc36cf23a58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(epo,cst, label='Epoch wrt to Cost θ1 & θ2=0')\n",
    "plt.plot(ep_rnd,cst_rnd, label='θ1 & θ2=Random')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAD8CAYAAAB6paOMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKQhERCKcgkBEJMIpCEREIpyCQEQkwv0fml6f8MPX2hIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fc36d97a470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#2\n",
    "#Start with random values of theta1 & theta2 . Analyse best approach for initial values of θ1 & θ2.\n",
    "plt.plot(w_rnd,b_rnd,label='θ1 & θ2=Random')\n",
    "plt.plot(w1,b1, label='θ1 & θ2=0')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#From the Above both graph it can be analysed that when θ1 & θ2=0 or random the θ1 & θ2 values for the plot is not \n",
    "#affected much \n",
    "#But when Analysed the Graph of Epoch and Cost when θ1 & θ2 =0 and random respectively it seems that there\n",
    "#is variation in both the graph\n",
    "#from the graph it can be said that θ1 & θ2= some random value is favorable because from random value to some stable \n",
    "#value may take or may require more no. of iterations but since much more learning takes place it makes our learning \n",
    "#algorithm more efficient "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Compare time complexity of tensorflow api and user defined function\n",
    "\n",
    "#the TIme complexity of the user defined functions is less when compared to the TensorFlow API and user defined functions \n",
    "#are comparitively faster however when it comes to the real time Data Sets or the real world data then user defined functions\n",
    "#tend to fail and Tensor Flow Api in such cases perform functions very faster irrespective of Data size and the Tensor Flow \n",
    "#API have inbuilt modules to check over false Data. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#2. Implement 3 cost functions\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Root Mean Square ERROR=  2.13889\n"
     ]
    }
   ],
   "source": [
    "ar2=tf.constant([2,2,4,6,5,4],name=\"Array_Y\",dtype=\"float\") \n",
    "mean_y=tf.reduce_mean(ar2)\n",
    "ses=tf.Session()\n",
    "dif=tf.square(ar2-mean_y)\n",
    "rms=(tf.reduce_sum(dif))/6\n",
    "print (\"Root Mean Square ERROR= \",ses.run(rms))\n"
   ]
  },
  {
   "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"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}