{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use shift-enter to execute a code block and move to the next one."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 1.1 Import tensorflow and other libraries.\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "\n",
    "%matplotlib inline\n",
    "import pylab"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# 1.2 Create input data using NumPy. y = x * 0.1 + 0.3 + noise\n",
    "x_data = np.random.rand(100).astype(np.float32)\n",
    "noise = np.random.normal(scale=0.01, size=len(x_data))\n",
    "y_data = x_data * 0.1 + 0.3 + noise\n",
    "\n",
    "# Uncomment the following line to plot our input data.\n",
    "# pylab.plot(x_data, y_data, '.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# 1.3 Buld inference graph.\n",
    "# Create Variables W and b that compute y_data = W * x_data + b\n",
    "W = tf.Variable(tf.random_uniform([1], 0.0, 1.0))\n",
    "b = tf.Variable(tf.zeros([1]))\n",
    "y = W * x_data + b\n",
    "\n",
    "# Uncomment the following lines to see W and b are.\n",
    "# print(W)\n",
    "# print(b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# 1.4 Build training graph.\n",
    "loss = tf.reduce_mean(tf.square(y - y_data))  # Create an operation that calculates loss.\n",
    "optimizer = tf.train.GradientDescentOptimizer(0.5)  # Create an optimizer.\n",
    "train = optimizer.minimize(loss)  # Create an operation that minimizes loss.\n",
    "init = tf.initialize_all_variables()  # Create an operation initializes all the variables.\n",
    "\n",
    "# Uncomment the following 3 lines to see what 'loss', 'optimizer' and 'train' are.\n",
    "# print(\"loss:\", loss)\n",
    "# print(\"optimizer:\", optimizer)\n",
    "# print(\"train:\", train)\n",
    "# print(init)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# 1.5 Uncomment the following line to see what we have built.\n",
    "# print(tf.get_default_graph().as_graph_def())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# 1.6 Create a session and launch the graph.\n",
    "sess = tf.Session()\n",
    "sess.run(init)\n",
    "y_initial_values = sess.run(y)  # Save initial values for plotting later.\n",
    "\n",
    "# Uncomment the following line to see the initial W and b values.\n",
    "# print(sess.run([W, b]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# 1.7 Perform training.\n",
    "for step in range(201):\n",
    "    sess.run(train)\n",
    "    # Uncomment the following two lines to watch training happen real time.\n",
    "    # if step % 20 == 0:\n",
    "    #    print(step, sess.run([W, b]))\n",
    "\n",
    "print(sess.run([W, b]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# 1.8 Uncomment the following lines to compare.\n",
    "# pylab.plot(x_data, y_data, '.', label=\"target_values\")\n",
    "# pylab.plot(x_data, y_initial_values, \".\", label=\"initial_values\")\n",
    "# pylab.plot(x_data, sess.run(y), \".\", label=\"trained_values\")\n",
    "# pylab.legend()\n",
    "# pylab.ylim(0, 1.0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>Back to [0_tf_hello_world.ipynb](0_tf_hello_world.ipynb).</p>\n",
    "<p>Next to [2_mnist.ipynb](2_mnist.ipynb).</p>"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}