fnn.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sharing Variables\n",
    "\n",
    "What is wrong with the following code?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "theta1 [[ 0.44360071 -0.55896199]\n",
      " [-0.30600622 -0.5037573 ]\n",
      " [-0.53044277  0.16552085]\n",
      " [ 0.31216997  0.17150772]\n",
      " [ 0.04960573  0.1461049 ]\n",
      " [ 0.01075457  0.17645364]\n",
      " [ 0.06395692  0.14987926]\n",
      " [ 0.00446167 -0.56886625]\n",
      " [-0.07550567 -0.2945618 ]\n",
      " [ 0.39035973  0.01515578]]\n",
      "theta2 [[-0.82923883 -0.39798287]\n",
      " [-0.43532887 -0.27481681]\n",
      " [-0.06774535  0.59454221]\n",
      " [ 0.2501682  -0.871499  ]\n",
      " [-0.20009017 -0.29256272]\n",
      " [-0.14484291  0.09395451]\n",
      " [ 0.20101026 -0.46657833]\n",
      " [-0.69620067 -0.72480702]\n",
      " [-0.09735443 -0.78983605]\n",
      " [ 0.19394319  0.02787798]]\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "\n",
    "def fnn(x, output_dim):\n",
    "    #weights and biases\n",
    "    w1 = tf.Variable(tf.random_normal([10, 2], stddev=0.35), name=\"weights1\")\n",
    "    b1 = tf.Variable(tf.zeros([2]), name=\"biases1\")\n",
    "\n",
    "    w2 = tf.Variable(tf.random_normal([2, output_dim], stddev=0.35), name=\"weights2\")\n",
    "    b2 = tf.Variable(tf.zeros([2]), name=\"biases2\")\n",
    "\n",
    "    # nn operators\n",
    "    y1 = tf.nn.relu(tf.matmul(x, w1) + b1)\n",
    "    y2 = tf.nn.sigmoid(tf.matmul(y1,w2) + b2)\n",
    "    return y2, [w1, w2]\n",
    "\n",
    "# Defining the computational graph\n",
    "x1 = tf.placeholder(tf.float32, shape=(1, 10))\n",
    "y1, theta1 = fnn(x1, 1)\n",
    "\n",
    "# The second network has different weights and biases\n",
    "x2 = tf.placeholder(tf.float32, shape=(1, 10))\n",
    "y2, theta2 = fnn(x2, 1)\n",
    "\n",
    "# Initializing the session\n",
    "with tf.Session() as sess:\n",
    "    tf.initialize_all_variables().run()\n",
    "    # Feeding and Fetching data\n",
    "    theta1, theta2 = sess.run([theta1, theta2], {x1: np.random.random([1, 10]), x2: np.random.random([1, 10])})\n",
    "    print(\"theta1\", theta1[0])\n",
    "    print(\"theta2\", theta2[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. It is hard to add a layer to the network.\n",
    "2. We cannot share the weights and biases. (Each function call creates a new set of variables)\n",
    "\n",
    "Creating networks using tf.Variable can get complicated. Instead, use tf.get_variable() to create/return variables and tf.variable_scope() to manage namespaces."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "Variable ffn/h/weights already exists, disallowed. Did you mean to set reuse=True in VarScope? Originally defined at:\n\n  File \"<ipython-input-2-0cf02566ad7f>\", line 5, in linear\n    w = tf.get_variable(name='weights', shape=[x.get_shape()[1], out_dim], dtype=tf.float32, initializer=tf.random_normal_initializer())\n  File \"<ipython-input-2-0cf02566ad7f>\", line 15, in <module>\n    y11, theta11 = linear(x1, 10, name=\"h\", activation_fn=tf.nn.relu)\n  File \"/Users/aida/anaconda3/lib/python3.5/site-packages/IPython/core/interactiveshell.py\", line 2885, in run_code\n    exec(code_obj, self.user_global_ns, self.user_ns)\n",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-2-0cf02566ad7f>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     19\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     20\u001b[0m     \u001b[0mx2\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplaceholder\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfloat32\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mshape\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m10\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 21\u001b[0;31m     \u001b[0my21\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtheta21\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlinear\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m10\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mname\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"h\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mactivation_fn\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrelu\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     22\u001b[0m     \u001b[0my22\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtheta22\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlinear\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mname\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"out\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mactivation_fn\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msigmoid\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     23\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-2-0cf02566ad7f>\u001b[0m in \u001b[0;36mlinear\u001b[0;34m(x, out_dim, name, activation_fn)\u001b[0m\n\u001b[1;32m      3\u001b[0m     \u001b[0;32mwith\u001b[0m \u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvariable_scope\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m         \u001b[0;31m#look for name/weights\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m         \u001b[0mw\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_variable\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'weights'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mshape\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_shape\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mout_dim\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfloat32\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minitializer\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandom_normal_initializer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      6\u001b[0m         \u001b[0mb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_variable\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'biases'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mshape\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mout_dim\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfloat32\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minitializer\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mconstant_initializer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0.0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m         \u001b[0mout\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmatmul\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mw\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/Users/aida/anaconda3/lib/python3.5/site-packages/tensorflow/python/ops/variable_scope.py\u001b[0m in \u001b[0;36mget_variable\u001b[0;34m(name, shape, dtype, initializer, regularizer, trainable, collections, caching_device, partitioner, validate_shape, custom_getter)\u001b[0m\n\u001b[1;32m    828\u001b[0m       \u001b[0mcollections\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcollections\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcaching_device\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcaching_device\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    829\u001b[0m       \u001b[0mpartitioner\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mpartitioner\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalidate_shape\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mvalidate_shape\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 830\u001b[0;31m       custom_getter=custom_getter)\n\u001b[0m\u001b[1;32m    831\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    832\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/Users/aida/anaconda3/lib/python3.5/site-packages/tensorflow/python/ops/variable_scope.py\u001b[0m in \u001b[0;36mget_variable\u001b[0;34m(self, var_store, name, shape, dtype, initializer, regularizer, trainable, collections, caching_device, partitioner, validate_shape, custom_getter)\u001b[0m\n\u001b[1;32m    671\u001b[0m           \u001b[0mcollections\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcollections\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcaching_device\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcaching_device\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    672\u001b[0m           \u001b[0mpartitioner\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mpartitioner\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalidate_shape\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mvalidate_shape\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 673\u001b[0;31m           custom_getter=custom_getter)\n\u001b[0m\u001b[1;32m    674\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    675\u001b[0m   def _get_partitioned_variable(\n",
      "\u001b[0;32m/Users/aida/anaconda3/lib/python3.5/site-packages/tensorflow/python/ops/variable_scope.py\u001b[0m in \u001b[0;36mget_variable\u001b[0;34m(self, name, shape, dtype, initializer, regularizer, reuse, trainable, collections, caching_device, partitioner, validate_shape, custom_getter)\u001b[0m\n\u001b[1;32m    215\u001b[0m           \u001b[0mreuse\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mreuse\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrainable\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtrainable\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcollections\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcollections\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    216\u001b[0m           \u001b[0mcaching_device\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcaching_device\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpartitioner\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mpartitioner\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 217\u001b[0;31m           validate_shape=validate_shape)\n\u001b[0m\u001b[1;32m    218\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    219\u001b[0m   def _get_partitioned_variable(\n",
      "\u001b[0;32m/Users/aida/anaconda3/lib/python3.5/site-packages/tensorflow/python/ops/variable_scope.py\u001b[0m in \u001b[0;36m_true_getter\u001b[0;34m(name, shape, dtype, initializer, regularizer, reuse, trainable, collections, caching_device, partitioner, validate_shape)\u001b[0m\n\u001b[1;32m    200\u001b[0m           \u001b[0minitializer\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0minitializer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mregularizer\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mregularizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreuse\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mreuse\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    201\u001b[0m           \u001b[0mtrainable\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtrainable\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcollections\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcollections\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 202\u001b[0;31m           caching_device=caching_device, validate_shape=validate_shape)\n\u001b[0m\u001b[1;32m    203\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    204\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mcustom_getter\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/Users/aida/anaconda3/lib/python3.5/site-packages/tensorflow/python/ops/variable_scope.py\u001b[0m in \u001b[0;36m_get_single_variable\u001b[0;34m(self, name, shape, dtype, initializer, regularizer, reuse, trainable, collections, caching_device, validate_shape)\u001b[0m\n\u001b[1;32m    492\u001b[0m                          \u001b[0;34m\" Did you mean to set reuse=True in VarScope? \"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    493\u001b[0m                          \"Originally defined at:\\n\\n%s\" % (\n\u001b[0;32m--> 494\u001b[0;31m                              name, \"\".join(traceback.format_list(tb))))\n\u001b[0m\u001b[1;32m    495\u001b[0m       \u001b[0mfound_var\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_vars\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    496\u001b[0m       \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mshape\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mis_compatible_with\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfound_var\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_shape\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: Variable ffn/h/weights already exists, disallowed. Did you mean to set reuse=True in VarScope? Originally defined at:\n\n  File \"<ipython-input-2-0cf02566ad7f>\", line 5, in linear\n    w = tf.get_variable(name='weights', shape=[x.get_shape()[1], out_dim], dtype=tf.float32, initializer=tf.random_normal_initializer())\n  File \"<ipython-input-2-0cf02566ad7f>\", line 15, in <module>\n    y11, theta11 = linear(x1, 10, name=\"h\", activation_fn=tf.nn.relu)\n  File \"/Users/aida/anaconda3/lib/python3.5/site-packages/IPython/core/interactiveshell.py\", line 2885, in run_code\n    exec(code_obj, self.user_global_ns, self.user_ns)\n"
     ]
    }
   ],
   "source": [
    "# function for creating nn layers\n",
    "def linear(x, out_dim, name, activation_fn=None):\n",
    "    with tf.variable_scope(name):\n",
    "        #look for name/weights\n",
    "        w = tf.get_variable(name='weights', shape=[x.get_shape()[1], out_dim], dtype=tf.float32, initializer=tf.random_normal_initializer())\n",
    "        b = tf.get_variable(name='biases', shape=[out_dim], dtype=tf.float32, initializer=tf.constant_initializer(0.0))\n",
    "        out = tf.matmul(x, w) + b\n",
    "        if activation_fn != None:\n",
    "            out = activation_fn(out)\n",
    "    return out, [w, b]\n",
    "\n",
    "# Computational Graph\n",
    "with tf.variable_scope(\"ffn\") as scope:\n",
    "    x1 = tf.placeholder(tf.float32, shape=(1, 10))\n",
    "    y11, theta11 = linear(x1, 10, name=\"h\", activation_fn=tf.nn.relu)\n",
    "    y12, theta12 = linear(y1, 1, name=\"out\", activation_fn=tf.nn.sigmoid)\n",
    "\n",
    "    #scope.reuse_variables()\n",
    "\n",
    "    x2 = tf.placeholder(tf.float32, shape=(1, 10))\n",
    "    y21, theta21 = linear(x2, 10, name=\"h\", activation_fn=tf.nn.relu)\n",
    "    y22, theta22 = linear(y1, 1, name=\"out\", activation_fn=tf.nn.sigmoid)\n",
    "\n",
    "\n",
    "# Initializing the session\n",
    "with tf.Session() as sess:\n",
    "    print(\"session\")\n",
    "    tf.initialize_all_variables().run()\n",
    "    # Feeding and Fetching data\n",
    "    theta1, theta2 = sess.run([theta12, theta22], {x1: np.random.random([1, 10]), x2: np.random.random([1, 10])})\n",
    "    print(theta1[0])\n",
    "    print(theta2[0])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Running the above code raises an error because the variable names are reused: \"Variable h/weights already exists, disallowed. Did you mean to set reuse=True in VarScope.\"\n",
    "\n",
    "If you need to share the variables, you need to explictly specify it using reuse_variables(). In the above code, you need to uncomment scope.reuse_variables().\n",
    "\n"
   ]
  }
 ],
 "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": "3.5.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}