Assignment2_175.ipynb

{
 "cells": [
  {
   "cell_type": "raw",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#Write a program using tensorflow to build a stochastic gradient descent model for linear regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Part 1\n",
    "1. Co-effiecients m &c for y=mx+c are calculated for given values from previous assignment\n",
    "2. Start with 0 (zero ) value for  m & c .\n",
    "3. Using cost function \n",
    "$$J(\\beta) = \\frac{1}{2m}\\sum_{i=1}^m(h_\\beta(x^{(i)})-y^{(i)})^2$$\n",
    "where $(h_\\beta(x^{(i)})$ is prediction for present iteration\n",
    "$y^{(i)}$ is the prediction obtained from y=mx+c value\n",
    "4. Find the values of m & c by updating the values with learning rate of 0.005, 0.0005 etc and epoch value 100,500,1000\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Part 2\n",
    "1. Display the graph of decreasing cost function wrt learning rate and epoch\n",
    "2. Compare time complexity of tensorflow api and user defined function\n",
    "3. Start with random values of m & c . Analyse best approach for initial values of m & c.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "rng = np.random\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "learning_rate = 0.0001\n",
    "training_epochs = 2000\n",
    "display_step = 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-0.3049\n",
      "1.494\n"
     ]
    }
   ],
   "source": [
    "with tf.name_scope(\"array_creation\"):\n",
    "    x = tf.constant(np.asarray([1.0,9.4,3.32,0.88,-2.23,1.11,0.57,-2.25,-3.31,6.45]),dtype = tf.float32,name = \"x_array\")\n",
    "    y = tf.constant(np.asarray([1.22,0.24,-0.08,2.25,4.41,3.09,-6.66,-9.77,0.001,2.25]),dtype = tf.float32, name= \"y_array\")\n",
    "with tf.name_scope(\"Calculation_of_means\"):\n",
    "    y_mean = tf.reduce_mean(y, name = \"y_mean\")\n",
    "    with tf.Session() as sess:\n",
    "        result_y = sess.run(y_mean)\n",
    "        print(result_y)\n",
    "    x_mean = tf.reduce_mean(x, name = \"x_mean\")\n",
    "    with tf.Session() as sess:\n",
    "        result_x = sess.run(x_mean)\n",
    "        print(result_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.494\n",
      "14.2987\n"
     ]
    }
   ],
   "source": [
    "\n",
    "with tf.name_scope(\"Calculating_x_mean_and_x_variance\"):\n",
    "    mean_x, variance = tf.nn.moments(x, [0], name = \"mean_x_and_variance_x\")\n",
    "    with tf.Session() as sess:\n",
    "        m, v = sess.run([mean_x, variance])\n",
    "        print(m)\n",
    "        print(v)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.60367\n"
     ]
    }
   ],
   "source": [
    "#Calculation of covariance\n",
    "with tf.name_scope(\"Calculation_of_covariance\"):\n",
    "    def tensorflow_covariance(x_array,y_array,x_mean,y_mean):\n",
    "        cov = 0.0\n",
    "        for i in range(0,10):\n",
    "            x_val = tf.subtract(x_array[i],x_mean, name=\"x_i_minus_x_bar\")\n",
    "            y_val = tf.subtract(y_array[i],y_mean, name=\"y_i_minus_y_bar\")\n",
    "            total_val = tf.multiply(x_val,y_val, name=\"Multiplying_found_values\")\n",
    "            cov = tf.add(cov,total_val, name=\"Summation_done\")\n",
    "        return cov/10.0\n",
    "    with tf.Session() as sess:\n",
    "        covar = sess.run(tensorflow_covariance(x,y,x_mean,y_mean))\n",
    "        print(covar)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.252027\n",
      "-0.681428\n"
     ]
    }
   ],
   "source": [
    "#calculation of slope and intercept\n",
    "#insted of taking a random value this is done\n",
    "with tf.name_scope(\"Calculation_of_slope_m_and_intercept_c\"):\n",
    "    slope = tf.div(covar,v,name=\"Finding_slope\")\n",
    "    intm = tf.multiply(slope,m,name = \"Intermediate_step\")\n",
    "    c_intm = tf.subtract(result_y,intm,name = \"Finding_c\")\n",
    "\n",
    "    with tf.Session() as sess:\n",
    "        m_slope = sess.run(slope)\n",
    "        c = sess.run(c_intm)\n",
    "        print(m_slope)\n",
    "        print(c)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "###Part-2: Plotting graph for actual values against predicted values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0100 cost= 8.864309311 W= 0.0593288 b= -0.00540942\n",
      "Epoch: 0200 cost= 8.822266579 W= 0.0794198 b= -0.00942419\n",
      "Epoch: 0300 cost= 8.791221619 W= 0.0965057 b= -0.0136748\n",
      "Epoch: 0400 cost= 8.768070221 W= 0.111048 b= -0.0181178\n",
      "Epoch: 0500 cost= 8.750589371 W= 0.123435 b= -0.0227163\n",
      "Epoch: 0600 cost= 8.737184525 W= 0.134 b= -0.0274393\n",
      "Epoch: 0700 cost= 8.726714134 W= 0.14302 b= -0.0322606\n",
      "Epoch: 0800 cost= 8.718362808 W= 0.150732 b= -0.0371581\n",
      "Epoch: 0900 cost= 8.711541176 W= 0.157337 b= -0.042113\n",
      "Epoch: 1000 cost= 8.705831528 W= 0.163003 b= -0.0471098\n",
      "Epoch: 1100 cost= 8.700933456 W= 0.167875 b= -0.0521349\n",
      "Epoch: 1200 cost= 8.696626663 W= 0.172073 b= -0.0571774\n",
      "Epoch: 1300 cost= 8.692759514 W= 0.175702 b= -0.0622278\n",
      "Epoch: 1400 cost= 8.689220428 W= 0.178847 b= -0.0672781\n",
      "Epoch: 1500 cost= 8.685929298 W= 0.181581 b= -0.0723218\n",
      "Epoch: 1600 cost= 8.682827950 W= 0.183969 b= -0.0773534\n",
      "Epoch: 1700 cost= 8.679876328 W= 0.186062 b= -0.0823682\n",
      "Epoch: 1800 cost= 8.677042961 W= 0.187905 b= -0.0873623\n",
      "Epoch: 1900 cost= 8.674306870 W= 0.189536 b= -0.0923323\n",
      "Epoch: 2000 cost= 8.671654701 W= 0.190986 b= -0.0972756\n",
      "Epoch: 2100 cost= 8.669071198 W= 0.192283 b= -0.10219\n",
      "Epoch: 2200 cost= 8.666550636 W= 0.19345 b= -0.107075\n",
      "Epoch: 2300 cost= 8.664084435 W= 0.194506 b= -0.111927\n",
      "Epoch: 2400 cost= 8.661669731 W= 0.195468 b= -0.116745\n",
      "Epoch: 2500 cost= 8.659303665 W= 0.196349 b= -0.12153\n",
      "Epoch: 2600 cost= 8.656983376 W= 0.197162 b= -0.126279\n",
      "Epoch: 2700 cost= 8.654705048 W= 0.197916 b= -0.130993\n",
      "Epoch: 2800 cost= 8.652467728 W= 0.19862 b= -0.135671\n",
      "Epoch: 2900 cost= 8.650270462 W= 0.199281 b= -0.140312\n",
      "Epoch: 3000 cost= 8.648110390 W= 0.199904 b= -0.144917\n",
      "Epoch: 3100 cost= 8.645990372 W= 0.200496 b= -0.149484\n",
      "Epoch: 3200 cost= 8.643905640 W= 0.20106 b= -0.154016\n",
      "Epoch: 3300 cost= 8.641856194 W= 0.201601 b= -0.158509\n",
      "Epoch: 3400 cost= 8.639842987 W= 0.20212 b= -0.162967\n",
      "Epoch: 3500 cost= 8.637863159 W= 0.202622 b= -0.167388\n",
      "Epoch: 3600 cost= 8.635918617 W= 0.203108 b= -0.171772\n",
      "Epoch: 3700 cost= 8.634006500 W= 0.20358 b= -0.176119\n",
      "Epoch: 3800 cost= 8.632127762 W= 0.204039 b= -0.18043\n",
      "Epoch: 3900 cost= 8.630280495 W= 0.204487 b= -0.184705\n",
      "Epoch: 4000 cost= 8.628462791 W= 0.204927 b= -0.188944\n",
      "Epoch: 4100 cost= 8.626678467 W= 0.205356 b= -0.193148\n",
      "Epoch: 4200 cost= 8.624923706 W= 0.205779 b= -0.197316\n",
      "Epoch: 4300 cost= 8.623197556 W= 0.206194 b= -0.201449\n",
      "Epoch: 4400 cost= 8.621502876 W= 0.206603 b= -0.205546\n",
      "Epoch: 4500 cost= 8.619836807 W= 0.207005 b= -0.209609\n",
      "Epoch: 4600 cost= 8.618199348 W= 0.207402 b= -0.213637\n",
      "Epoch: 4700 cost= 8.616587639 W= 0.207793 b= -0.217631\n",
      "Epoch: 4800 cost= 8.615004539 W= 0.20818 b= -0.221592\n",
      "Epoch: 4900 cost= 8.613449097 W= 0.208562 b= -0.225518\n",
      "Epoch: 5000 cost= 8.611920357 W= 0.208939 b= -0.229411\n",
      "Optimization Finished!\n",
      "Training cost= 8.61192 W= 0.208939 b= -0.229411 \n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAD8CAYAAABw1c+bAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd7d4122dd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "with tf.name_scope(\"Plotting\"):\n",
    "    n_samples = x_array.shape[0]\n",
    "    X = tf.placeholder(\"float\")\n",
    "    Y = tf.placeholder(\"float\")\n",
    "\n",
    "    # Set model weights\n",
    "    W = tf.Variable(rng.randn(), name=\"weight\")\n",
    "    b = tf.Variable(rng.randn(), name=\"bias\")\n",
    "\n",
    "    # Construct a linear model\n",
    "    pred = tf.add(tf.multiply(X, W), b)\n",
    "\n",
    "\n",
    "    # Mean squared error\n",
    "    cost = tf.reduce_sum(tf.pow(pred-Y, 2))/(2*n_samples)\n",
    "    # Gradient descent\n",
    "    optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)\n",
    "\n",
    "    # Initializing the variables\n",
    "    init = tf.global_variables_initializer()\n",
    "\n",
    "    # Launch the graph\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 (p, r) in zip(x_array, y_array):\n",
    "                sess.run(optimizer, feed_dict={X: p, Y: r})\n",
    "\n",
    "            # Display logs per epoch step\n",
    "            if (epoch+1) % display_step == 0:\n",
    "                c = sess.run(cost, feed_dict={X: x_array, Y:y_array})\n",
    "                print(\"Epoch:\", '%04d' % (epoch+1), \"cost=\", \"{:.9f}\".format(c), \\\n",
    "                    \"W=\", sess.run(W), \"b=\", sess.run(b))\n",
    "\n",
    "        print(\"Optimization Finished!\")\n",
    "        training_cost = sess.run(cost, feed_dict={X: x_array, Y: y_array})\n",
    "        print(\"Training cost=\", training_cost, \"W=\", sess.run(W), \"b=\", sess.run(b), '\\n')\n",
    "\n",
    "        # Graphic display\n",
    "        plt.plot(x_array, y_array, 'ro', label='Original data')\n",
    "        plt.plot(x_array, sess.run(W) * x_array + sess.run(b), label='Fitted line')\n",
    "        plt.legend()\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5.3181474312\n"
     ]
    }
   ],
   "source": [
    "#rms error\n",
    "with tf.name_scope(\"RMSerror\"):\n",
    "    rms = tf.sqrt(tf.reduce_mean(tf.squared_difference(x_array, y_array,name = \"difference_square\"),name=\"Finding_mean\"),name = \"Finding_square_root\")\n",
    "    with tf.Session() as sess:\n",
    "        rmse=sess.run(rms)\n",
    "        print(rmse)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.252027\n"
     ]
    }
   ],
   "source": [
    "with tf.name_scope(\"Finding_teta_1\"): \n",
    "    y_var = tf.subtract(y,result_y,name = \"Subtract_y_array_with_y_mean\")\n",
    "    x_var = tf.subtract(x,m,name = \"Subtract_x_array_with_x_mean\")\n",
    "    mult = tf.multiply(x_var,y_var,name = \"Multiply_calculated_arrays\")\n",
    "    sumn = tf.reduce_sum(mult,name = \"Find_sum_of_x_i_minus_mean_x_and_y_i_minus_mean_y\")\n",
    "    x_var2 = tf.multiply(x_var,x_var,name = \"Squaring_found_arrray_values\")\n",
    "    sumd = tf.reduce_sum(x_var2,name = \"Find_sum_of_array_of_x_i_minus_mean_x\")\n",
    "    val = sumn/sumd\n",
    "\n",
    "    with tf.Session() as sess:\n",
    "        res = sess.run(val)\n",
    "        print(res)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-0.681428\n"
     ]
    }
   ],
   "source": [
    "with tf.name_scope(\"Finding_teta_0\"):    \n",
    "    temp = tf.multiply(res,m,name = \"Multiply_res_with_slope\")\n",
    "    teta = tf.subtract(result_y,temp,name=\"Sub_obtained_res_with_mean_y\")\n",
    "    with tf.Session() as sess:\n",
    "        teta0 = sess.run(teta)\n",
    "        print(teta0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-0.42940134  1.68762565  0.15530139 -0.45964459 -1.24344873 -0.40167835\n",
      " -0.53777295 -1.24848914 -1.51563787  0.94414598]\n"
     ]
    }
   ],
   "source": [
    "with tf.name_scope(\"Finding_predictions\"):\n",
    "    mx = tf.multiply(res,x,name = \"Multiply_res_with_x_array\")\n",
    "    y_temp = tf.add(mx,teta0,name = \"Add_m_multiplied_x_array_with_c\")\n",
    "    with tf.Session() as sess:\n",
    "        y_new = sess.run(y_temp)\n",
    "        print(y_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8.52351\n"
     ]
    }
   ],
   "source": [
    "t_minus = tf.subtract(y_new,y,name = \"Sub_new_preds_with_original_y\")\n",
    "t_squared = tf.multiply(t_minus,t_minus,name= \"Square_obtained_res\")\n",
    "t_sum = tf.reduce_sum(t_squared,name=\"Find_array_sum\")\n",
    "j_teta = tf.div(t_sum,20,name=\"Divide_by_no_of_elements\")\n",
    "with tf.Session() as sess:\n",
    "    print(sess.run(j_teta))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8.52351\n"
     ]
    }
   ],
   "source": [
    "with tf.Session() as sess:\n",
    "    writer = tf.summary.FileWriter(\"/tmp/tboard/output_regg2\", sess.graph)\n",
    "    print(sess.run(j_teta))\n",
    "    writer.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "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
}