tensorflow_from_scratch+.+0.1+.+Basics.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Generating data\n",
    ">Numpy vs Tensorflow\n",
    "\n",
    "| Numpy | Tensorflow |\n",
    "| --- | --- |\n",
    "|``` shape = .shape ```| ```shape = .get_shape()```, values are computed and sent over |\n",
    "| output | output has to be computed as the expression is stored as an TF operation |"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(100,)\n",
      "float64\n",
      "_____________________________________________________________________\n",
      "Printing the output for n and x\n",
      "Tensor(\"LinSpace:0\", shape=(100,), dtype=float32)\n",
      "[-5.         -4.8989899  -4.7979798  -4.6969697  -4.5959596  -4.49494949\n",
      " -4.39393939 -4.29292929 -4.19191919 -4.09090909 -3.98989899 -3.88888889\n",
      " -3.78787879 -3.68686869 -3.58585859 -3.48484848 -3.38383838 -3.28282828\n",
      " -3.18181818 -3.08080808 -2.97979798 -2.87878788 -2.77777778 -2.67676768\n",
      " -2.57575758 -2.47474747 -2.37373737 -2.27272727 -2.17171717 -2.07070707\n",
      " -1.96969697 -1.86868687 -1.76767677 -1.66666667 -1.56565657 -1.46464646\n",
      " -1.36363636 -1.26262626 -1.16161616 -1.06060606 -0.95959596 -0.85858586\n",
      " -0.75757576 -0.65656566 -0.55555556 -0.45454545 -0.35353535 -0.25252525\n",
      " -0.15151515 -0.05050505  0.05050505  0.15151515  0.25252525  0.35353535\n",
      "  0.45454545  0.55555556  0.65656566  0.75757576  0.85858586  0.95959596\n",
      "  1.06060606  1.16161616  1.26262626  1.36363636  1.46464646  1.56565657\n",
      "  1.66666667  1.76767677  1.86868687  1.96969697  2.07070707  2.17171717\n",
      "  2.27272727  2.37373737  2.47474747  2.57575758  2.67676768  2.77777778\n",
      "  2.87878788  2.97979798  3.08080808  3.18181818  3.28282828  3.38383838\n",
      "  3.48484848  3.58585859  3.68686869  3.78787879  3.88888889  3.98989899\n",
      "  4.09090909  4.19191919  4.29292929  4.39393939  4.49494949  4.5959596\n",
      "  4.6969697   4.7979798   4.8989899   5.        ]\n"
     ]
    }
   ],
   "source": [
    "n = np.linspace(-5,5,100)\n",
    "print(n.shape)\n",
    "print(n.dtype)\n",
    "\n",
    "t = tf.linspace(-5.0,5.0,100)\n",
    "\n",
    "print(\"_____________________________________________________________________\")\n",
    "print(\"Printing the output for n and x\")\n",
    "print(t)\n",
    "print(n)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    ">As we can see, there is no output for the tensorflow operation here. \n",
    "\n",
    ">Numpy and Tensorflow being almost the same we can see that tensorflow plays the game a little differently. \n",
    "\n",
    ">What tensorflow does is, it takes the given expression and stores it as a TF operation which is executed only when the tensor is in a session.\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Default Graphs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<tensorflow.python.framework.ops.Graph object at 0x0000028844E9E5C0>\n"
     ]
    }
   ],
   "source": [
    "g = tf.get_default_graph()\n",
    "print(g)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['LinSpace/start', 'LinSpace/stop', 'LinSpace/num', 'LinSpace']"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[op.name for op in g.get_operations()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    ">Now we return all the operations which have been added to the graph.\n",
    "\n",
    ">Each of these operations is a Tensor in itself which performs a particular tast.\n",
    "\n",
    ">The last operation takes in the first three operations. We can request the output of any operation which is a tensor, by asking the graph for the Tensor's name."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor 'LinSpace:0' shape=(100,) dtype=float32>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "g.get_tensor_by_name(\"LinSpace:0\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see here, the result of the TF operation is a TF Tensor.\n",
    ">In order to compute anything in TensorFlow we need to create a TF session. The session is responsible for evaluating the TF graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.client.session.Session at 0x2884723b470>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sess = tf.Session()\n",
    "sess"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-5.         -4.89898968 -4.79797983 -4.69696951 -4.59595966 -4.49494934\n",
      " -4.3939395  -4.29292917 -4.19191933 -4.090909   -3.98989916 -3.88888884\n",
      " -3.78787875 -3.68686867 -3.58585858 -3.4848485  -3.38383842 -3.28282833\n",
      " -3.18181825 -3.08080816 -2.97979808 -2.87878799 -2.77777791 -2.67676783\n",
      " -2.5757575  -2.47474742 -2.37373734 -2.27272725 -2.17171717 -2.07070708\n",
      " -1.969697   -1.86868691 -1.76767683 -1.66666675 -1.56565666 -1.46464658\n",
      " -1.36363649 -1.26262641 -1.16161633 -1.06060624 -0.95959616 -0.85858583\n",
      " -0.75757599 -0.65656567 -0.55555582 -0.4545455  -0.35353565 -0.25252533\n",
      " -0.15151501 -0.05050516  0.05050516  0.15151501  0.25252533  0.35353518\n",
      "  0.4545455   0.55555534  0.65656567  0.75757551  0.85858583  0.95959568\n",
      "  1.060606    1.16161585  1.26262617  1.36363602  1.46464634  1.56565666\n",
      "  1.66666651  1.76767683  1.86868668  1.969697    2.07070684  2.17171717\n",
      "  2.27272701  2.37373734  2.47474718  2.5757575   2.67676735  2.77777767\n",
      "  2.87878752  2.97979784  3.08080769  3.18181801  3.28282833  3.38383865\n",
      "  3.48484802  3.58585835  3.68686867  3.78787899  3.88888836  3.98989868\n",
      "  4.090909    4.19191933  4.2929287   4.39393902  4.49494934  4.59595966\n",
      "  4.69696999  4.79797935  4.89898968  5.        ]\n",
      "Tensor(\"LinSpace:0\", shape=(100,), dtype=float32)\n"
     ]
    }
   ],
   "source": [
    "computedT = sess.run(t)\n",
    "print(computedT)\n",
    "print(t)\n",
    "sess.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> By default TensorFlow gets the default graph to run in the session. \n",
    "\n",
    "```\n",
    "sess = tf.Session(graph = tf.get_default_graph())\n",
    "\n",
    "```\n",
    "\n",
    ">But one can change this by selecting which graph they want to import by saying.\n",
    "\n",
    "```\n",
    "sess = tf.Session(graph = g)\n",
    "\n",
    "```\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create Graphs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "g1 = tf.Graph()\n",
    "sess = tf.Session(graph=g1)\n",
    "sess.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Creating an Interactive Session"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sess = tf.InteractiveSession()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Running the operation without specifying the session for the evaluation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-5.        , -4.89898968, -4.79797983, -4.69696951, -4.59595966,\n",
       "       -4.49494934, -4.3939395 , -4.29292917, -4.19191933, -4.090909  ,\n",
       "       -3.98989916, -3.88888884, -3.78787875, -3.68686867, -3.58585858,\n",
       "       -3.4848485 , -3.38383842, -3.28282833, -3.18181825, -3.08080816,\n",
       "       -2.97979808, -2.87878799, -2.77777791, -2.67676783, -2.5757575 ,\n",
       "       -2.47474742, -2.37373734, -2.27272725, -2.17171717, -2.07070708,\n",
       "       -1.969697  , -1.86868691, -1.76767683, -1.66666675, -1.56565666,\n",
       "       -1.46464658, -1.36363649, -1.26262641, -1.16161633, -1.06060624,\n",
       "       -0.95959616, -0.85858583, -0.75757599, -0.65656567, -0.55555582,\n",
       "       -0.4545455 , -0.35353565, -0.25252533, -0.15151501, -0.05050516,\n",
       "        0.05050516,  0.15151501,  0.25252533,  0.35353518,  0.4545455 ,\n",
       "        0.55555534,  0.65656567,  0.75757551,  0.85858583,  0.95959568,\n",
       "        1.060606  ,  1.16161585,  1.26262617,  1.36363602,  1.46464634,\n",
       "        1.56565666,  1.66666651,  1.76767683,  1.86868668,  1.969697  ,\n",
       "        2.07070684,  2.17171717,  2.27272701,  2.37373734,  2.47474718,\n",
       "        2.5757575 ,  2.67676735,  2.77777767,  2.87878752,  2.97979784,\n",
       "        3.08080769,  3.18181801,  3.28282833,  3.38383865,  3.48484802,\n",
       "        3.58585835,  3.68686867,  3.78787899,  3.88888836,  3.98989868,\n",
       "        4.090909  ,  4.19191933,  4.2929287 ,  4.39393902,  4.49494934,\n",
       "        4.59595966,  4.69696999,  4.79797935,  4.89898968,  5.        ], dtype=float32)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t.eval()"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}