IFmodel_SNN-Copy0.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:358b2d686ad6eb9eec9c624ab2ea6a840792c88f9166495e46b331111b52860b"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import h5py\n",
      "import theanets\n",
      "import time\n",
      "import random"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 15
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#  Converting deep ANNs to SNNs:  \n",
      "\n",
      "# 1) Use ReLUs for all units of the network.\n",
      "\n",
      "# 2) Fix the bias to zero throughout training, and train with backpropagation.\n",
      "\n",
      "# 3) Directly map the weights from the ReLU network to a network of IF units.\n",
      "\n",
      "# 4) Use weight normalization to obtain near-lossless accuracy and faster convergence."
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from brian import *\n",
      "from brian.library.synapses import *\n",
      "from brian.clock import reinit_default_clock\n",
      "from brian.stdunits import kHz, Hz, ms\n",
      "from brian.network import Network\n",
      "from brian.monitor import StateMonitor, SpikeMonitor\n",
      "from brian.globalprefs import set_global_preferences\n",
      "from brian import clear\n",
      "\n",
      "import cPickle\n",
      "%matplotlib inline"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "/usr/local/lib/python2.7/site-packages/brian-1.4.1-py2.7-linux-x86_64.egg/brian/utils/sparse_patch/__init__.py:39: UserWarning: Couldn't find matching sparse matrix patch for scipy version 0.15.1, but in most cases this shouldn't be a problem.\n",
        "  warnings.warn(\"Couldn't find matching sparse matrix patch for scipy version %s, but in most cases this shouldn't be a problem.\" % scipy.__version__)\n"
       ]
      }
     ],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "f=h5py.File(\"data/data.hdf5\")\n",
      "f.keys()\n",
      "f['data'].keys()\n",
      "\n",
      "global input_matrix\n",
      "global target_matrix\n",
      "global decorator_matrix\n",
      "\n",
      "input_matrix     = f['data']['input'].value\n",
      "target_matrix    = f['data']['target'].value\n",
      "decorator_matrix = f['data']['deco'].value\n",
      "f.close()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "nInputs  = input_matrix.shape[1]\n",
      "nOutputs = target_matrix.shape[1]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "print \"Samples in input_matrix=\", len(input_matrix), \"       Inputs=\", nInputs, \"       Outputs=\",nOutputs"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Samples in input_matrix= 982867        Inputs= 82        Outputs= 1\n"
       ]
      }
     ],
     "prompt_number": 6
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "truncate=100000          \n",
      "nHiddenNeurons=200\n",
      "iLayers=1\n",
      "\n",
      "experiment=theanets.feedforward.Regressor(layers=(nInputs,)+tuple([nHiddenNeurons for l in range(iLayers), 'relu'])+ ( (nOutputs, 'relu'),))\n",
      "\n",
      "# Create Training & Validation Set\n",
      "\n",
      "IN_matrix     = input_matrix[0:truncate,...]      \n",
      "TARGET_matrix = target_matrix[0:truncate,...]\n",
      "DEC_matrix    = decorator_matrix[0:truncate,...] \n",
      "    \n",
      "cut   =int(IN_matrix.shape[0]*0.8)\n",
      "train =IN_matrix[:cut], TARGET_matrix[:cut]    \n",
      "valid =IN_matrix[cut:], TARGET_matrix[cut:]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 23
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# No negative inputs\n",
      "\n",
      "for i, iSample in enumerate(train[0]):\n",
      "    minInput= min(iSample)\n",
      "    if(minInput < 0):\n",
      "        for j, jInput in enumerate(iSample):\n",
      "            if (jInput != 0.) :\n",
      "                train[0][i][j]=jInput - minInput"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 24
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "Weights_hid1= experiment.find('hid1', 'w').get_value()\n",
      "Weights_out = experiment.find('out','w').get_value() \n",
      "#Bias_hid1   = experiment.find('hid1', 'b').get_value()\n",
      "#Bias_out    = experiment.find('hid1', 'b').get_value()\n",
      "\n",
      "\n",
      "New_Weights     =zeros(Weights_hid1.shape)\n",
      "#New_bias        =zeros(Bias_hid1.shape)\n",
      "#New_bias_out    =zeros(Bias_out.shape)\n",
      "New_Weights_out =zeros(Weights_out.shape)\n",
      "\n",
      "for i, iNeuron in enumerate(Weights_hid1):\n",
      "    for iw,iWeight in enumerate(iNeuron):\n",
      "        New_Weights[i][iw]=random.expovariate(-15)\n",
      "#        New_bias[i][iw]   =random.expovariate(2)\n",
      "    \n",
      "\n",
      "for i, iNeuron in enumerate(Weights_out):\n",
      "    for iw,iWeight in enumerate(iNeuron):\n",
      "        New_Weights_out[i][iw]=random.expovariate(-15)\n",
      "#        New_bias_out[i][iw]   =random.expovariate(2)\n",
      "        \n",
      "experiment.find('hid1', 'w').set_value([iNeuron[1] for iNeuron in enumerate(New_Weights)])\n",
      "experiment.find('hid1', 'b').set_value([iNeuron[1] for iNeuron in enumerate(New_bias)])\n",
      "#experiment.find('out' , 'w').set_value([iNeuron[1] for iNeuron in enumerate(New_Weights_out)])\n",
      "#experiment.find('hid1', 'b').set_value([iNeuron[1] for iNeuron in enumerate(New_bias_out)])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 34
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "fom_train   = 1.0\n",
      "fom_valid   = 1.0\n",
      "\n",
      "LearningRate = 1e-4\n",
      "Epochs       = 0\n",
      "maxEpochs    = 1500\n",
      "Dropout      = 0.5\n",
      "momentum     = 0.5\n",
      "\n",
      "timing_start = time.mktime(time.gmtime())\n",
      "    \n",
      "for TrainError, ValError in experiment.itertrain(train,valid, \n",
      "                                                 learning_rate  = LearningRate,\n",
      "                                                 algo='rprop',\n",
      "                                                 hidden_dropout = Dropout,\n",
      "                                                 iteration_size = 1,\n",
      "                                                 batch_size     = 10, \n",
      "                                                 rms_halflife   = 0.99, \n",
      "                                                 rms_regularizer= 1e-09):\n",
      "    print 'Epoch=', Epochs+1\n",
      "    if Epochs%1 == 0:\n",
      "        print 'Training   Error =',TrainError['err']\n",
      "        print 'Validation Error =',ValError['err']\n",
      "    Epochs+=1\n",
      "    if Epochs == maxEpochs :\n",
      "        break\n",
      "\n",
      "experiment.save('network-pickle.pkl.gz')        \n",
      "prediction = experiment.predict(train[0])\n",
      "goodness = np.asarray([[cut,(list(train[1][np.where(prediction>cut)]==1).count(True)+list(train[1][np.where(prediction<cut)]==0).count(True))/ float(train[0].shape[0])]  for cut in np.arange(0,1,0.05)])   \n",
      "fom_train = max( goodness[:,1])    \n",
      "\n",
      "prediction = experiment.predict(valid[0])\n",
      "goodness   = np.asarray([[cut,(list(valid[1][np.where(prediction>cut)]==1).count(True)+list(valid[1][np.where(prediction<cut)]==0).count(True))/ float(valid[0].shape[0])]  for cut in np.arange(0,1,0.05)])\n",
      "fom_valid  = max( goodness[:,1])     \n",
      "\n",
      "MaxCut= goodness[np.where(goodness[:,1]==goodness[:,1].max())][0][0]\n",
      "timing_end=time.mktime(time.gmtime())\n",
      "timing=timing_end-timing_start\n",
      "    \n",
      "print 'Samples=', truncate,'    Hid. Layers=',iLayers, \"   Hid. Neurons=\", nHiddenNeurons, \"   F.O.M. train set=\",fom_train, \"   FOM valid. set=\", fom_valid, \" Time=\", timing, \" Max Cut=\", MaxCut      "
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Epoch= 1\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.7\n",
        "Epoch= 2\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.7\n",
        "Epoch= 3\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.7\n",
        "Epoch= 4\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.7\n",
        "Epoch= 5\n",
        "Training   Error = 0.4\n",
        "Validation Error = 0.7\n",
        "Epoch= 6\n",
        "Training   Error = 0.3\n",
        "Validation Error = 0.7\n",
        "Epoch= 7\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.7\n",
        "Epoch= 8\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.7\n",
        "Epoch= 9\n",
        "Training   Error = 0.4\n",
        "Validation Error = 0.7\n",
        "Epoch= 10\n",
        "Training   Error = 0.4\n",
        "Validation Error = 0.7\n",
        "Epoch= 11\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.6\n",
        "Epoch= 12\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.6\n",
        "Epoch= 13\n",
        "Training   Error = 0.4\n",
        "Validation Error = 0.6\n",
        "Epoch= 14\n",
        "Training   Error = 0.9\n",
        "Validation Error = 0.6\n",
        "Epoch= 15\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.6\n",
        "Epoch= 16\n",
        "Training   Error = 0.3\n",
        "Validation Error = 0.6\n",
        "Epoch= 17\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.6\n",
        "Epoch= 18\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.6\n",
        "Epoch= 19\n",
        "Training   Error = 0.3\n",
        "Validation Error = 0.6\n",
        "Epoch= 20\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.6\n",
        "Epoch="
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " 21\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.4\n",
        "Epoch= 22\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.4\n",
        "Epoch= 23\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.4\n",
        "Epoch= 24\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.4\n",
        "Epoch= 25\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.4\n",
        "Epoch= 26\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.4\n",
        "Epoch= 27\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.4\n",
        "Epoch= 28\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.4\n",
        "Epoch= 29\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.4\n",
        "Epoch= 30\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.4\n",
        "Epoch= 31\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.7\n",
        "Epoch= 32\n",
        "Training   Error = 0.4\n",
        "Validation Error = 0.7\n",
        "Epoch= 33\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.7\n",
        "Epoch= 34\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.7\n",
        "Epoch= 35\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.7\n",
        "Epoch= 36\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.7\n",
        "Epoch= 37\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.7\n",
        "Epoch= 38\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.7\n",
        "Epoch= 39\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.7\n",
        "Epoch= 40\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.7\n",
        "Epoch= 41\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.8\n",
        "Epoch="
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " 42\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.8\n",
        "Epoch= 43\n",
        "Training   Error = 0.9\n",
        "Validation Error = 0.8\n",
        "Epoch= 44\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.8\n",
        "Epoch= 45\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.8\n",
        "Epoch= 46\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.8\n",
        "Epoch= 47\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.8\n",
        "Epoch= 48\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.8\n",
        "Epoch= 49\n",
        "Training   Error = 0.9\n",
        "Validation Error = 0.8\n",
        "Epoch= 50\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.8\n",
        "Epoch= 51\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.5\n",
        "Epoch= 52\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.5\n",
        "Epoch= 53\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.5\n",
        "Epoch= 54\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.5\n",
        "Epoch= 55\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.5\n",
        "Epoch= 56\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.5\n",
        "Epoch= 57\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.5\n",
        "Epoch= 58\n",
        "Training   Error = 0.2\n",
        "Validation Error = 0.5\n",
        "Epoch= 59\n",
        "Training   Error = 0.9\n",
        "Validation Error = 0.5\n",
        "Epoch= 60\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.5\n",
        "Epoch= 61\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.8\n",
        "Epoch="
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " 62\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.8\n",
        "Epoch= 63\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.8\n",
        "Epoch= 64\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.8\n",
        "Epoch= 65\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.8\n",
        "Epoch= 66\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.8\n",
        "Epoch= 67\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.8\n",
        "Epoch= 68\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.8\n",
        "Epoch= 69\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.8\n",
        "Epoch= 70\n",
        "Training   Error = 0.3\n",
        "Validation Error = 0.8\n",
        "Epoch= 71\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.3\n",
        "Epoch= 72\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.3\n",
        "Epoch= 73\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.3\n",
        "Epoch= 74\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.3\n",
        "Epoch= 75\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.3\n",
        "Epoch= 76\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.3\n",
        "Epoch= 77\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.3\n",
        "Epoch= 78\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.3\n",
        "Epoch= 79\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.3\n",
        "Epoch= 80\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.3\n",
        "Epoch= 81\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.6\n",
        "Epoch="
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " 82\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.6\n",
        "Epoch= 83\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.6\n",
        "Epoch= 84\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.6\n",
        "Epoch= 85\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.6\n",
        "Epoch= 86\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.6\n",
        "Epoch= 87\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.6\n",
        "Epoch= 88\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.6\n",
        "Epoch= 89\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.6\n",
        "Epoch= 90\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.6\n",
        "Epoch= 91\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.3\n",
        "Epoch= 92\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.3\n",
        "Epoch= 93\n",
        "Training   Error = 0.9\n",
        "Validation Error = 0.3\n",
        "Epoch= 94\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.3\n",
        "Epoch= 95\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.3\n",
        "Epoch= 96\n",
        "Training   Error = 0.3\n",
        "Validation Error = 0.3\n",
        "Epoch= 97\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.3\n",
        "Epoch= 98\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.3\n",
        "Epoch= 99\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.3\n",
        "Epoch= 100\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.3\n",
        "Epoch= 101\n",
        "Training   Error = 1.0\n",
        "Validation Error = 0.6\n",
        "Epoch= 102\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.6\n",
        "Epoch="
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " 103\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.6\n",
        "Epoch= 104\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.6\n",
        "Epoch= 105\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.6\n",
        "Epoch= 106\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.6\n",
        "Epoch= 107\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.6\n",
        "Epoch= 108\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.6\n",
        "Epoch= 109\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.6\n",
        "Epoch= 110\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.6\n",
        "Epoch= 111\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.5\n",
        "Epoch= 112\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.5\n",
        "Epoch= 113\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.5\n",
        "Epoch= 114\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.5\n",
        "Epoch= 115\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.5\n",
        "Epoch= 116\n",
        "Training   Error = 0.9\n",
        "Validation Error = 0.5\n",
        "Epoch= 117\n",
        "Training   Error = 0.4\n",
        "Validation Error = 0.5\n",
        "Epoch= 118\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.5\n",
        "Epoch= 119\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.5\n",
        "Epoch= 120\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.5\n",
        "Epoch= 121\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.4\n",
        "Epoch= 122\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.4\n",
        "Epoch="
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " 123\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.4\n",
        "Epoch= 124\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.4\n",
        "Epoch= 125\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.4\n",
        "Epoch= 126\n",
        "Training   Error = 0.8\n",
        "Validation Error = 0.4\n",
        "Epoch= 127\n",
        "Training   Error = 0.7\n",
        "Validation Error = 0.4\n",
        "Epoch= 128\n",
        "Training   Error = 0.5\n",
        "Validation Error = 0.4\n",
        "Epoch= 129\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.4\n",
        "Epoch= 130\n",
        "Training   Error = 0.6\n",
        "Validation Error = 0.4\n",
        "Samples="
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " 100000     Hid. Layers= 1    Hid. Neurons= 200    F.O.M. train set= 0.3909875    FOM valid. set= 0.3925  Time= 10.0  Max Cut= 0.05\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING (theano.gof.cmodule): Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpOywr8J/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING:theano.gof.cmodule:Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpOywr8J/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING (theano.gof.cmodule): Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpEYOWAH/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING:theano.gof.cmodule:Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpEYOWAH/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING (theano.gof.cmodule): Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpJJ0xhj/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING:theano.gof.cmodule:Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpJJ0xhj/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING (theano.gof.cmodule): Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmp5l0J74/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING:theano.gof.cmodule:Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmp5l0J74/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING (theano.gof.cmodule): Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpuZkOp7/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING:theano.gof.cmodule:Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpuZkOp7/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING (theano.gof.cmodule): Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpbsLHpR/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING:theano.gof.cmodule:Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpbsLHpR/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING (theano.gof.cmodule): Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpTJeULh/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING:theano.gof.cmodule:Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmpTJeULh/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING (theano.gof.cmodule): Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmp2HE8lt/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "WARNING:theano.gof.cmodule:Removing key file /root/.theano/compiledir_Linux-2.6-el6.x86_64-x86_64-with-redhat-6.6-Carbon-x86_64-2.7.8-64/tmp2HE8lt/key.pkl because the corresponding module is gone from the file system.\n"
       ]
      }
     ],
     "prompt_number": 39
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Load network\n",
      "\n",
      "Layers_Weights =[]                                 #list of tuples: [(nNeurons, weights), ( tuple of the 3rd layer ), (  4th layer), .... ]\n",
      "\n",
      "for Layer in experiment.layers[1:]:\n",
      "    nNeurons  = Layer.size\n",
      "    Weights1  = experiment.find(Layer.name, 'w').get_value()\n",
      "    Layers_Weights.append((nNeurons, Weights1))                    # Layers_Weights -> only Hidden Layer(s) & Output Layer"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "abs_max=0\n",
      "for layer_num, weights in enumerate(Layers_Weights):   \n",
      "    abs_max = max(abs_max, np.max(weights[1]))\n",
      "print abs_max\n",
      "for layer_num, weights in enumerate(Layers_Weights):   \n",
      "    for i, neuron in enumerate(weights[1]):\n",
      "        for j,iW in enumerate(neuron):\n",
      "            Layers_Weights[layer_num][1][i][j]=float(iW) / float(abs_max)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Step 4      : Weight Normalization using algorithm: Model-Based Normalization\n",
      "\n",
      "for layer_num, weights in enumerate(Layers_Weights):   \n",
      "    max_pos_input = 0\n",
      "    for i, neuron in enumerate(weights[1]):\n",
      "        input_sum = 0\n",
      "        for iW in neuron:\n",
      "            input_sum += max(0, iW)\n",
      "        max_pos_input = max(max_pos_input, input_sum)\n",
      "    print max_pos_input,\"for layer\",layer_num\n",
      "    for i, neuron in enumerate(weights[1]):\n",
      "        j=0                                   \n",
      "        for iW in neuron:\n",
      "            #Layers_Weights[layer_num][1][i][j]=float(iW) / float(max_pos_input)    # Rescale all weights\n",
      "            j+=1"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "10.956948922 for layer 0\n",
        "0.024085800402 for layer 1\n"
       ]
      }
     ],
     "prompt_number": 63
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "reinit_default_clock()\n",
      "clear(True)\n",
      "# Spiking Neural Network :\n",
      "\n",
      "# IF Model : http://www.briansimulator.org/docs/tutorial_1a_the_simplest_brian_program.html\n",
      "#          : http://icwww.epfl.ch/~gerstner/SPNM/node26.html\n",
      "\n",
      "# Membrane time constant:  tau_m = 10 * ms\n",
      "# Reset potential       :  v = 0 * mV\n",
      "# Threshold potential   :  v = 15 * mV\n",
      "# Equation              :  dv/dt =  -v/tau_m : volt \n",
      "\n",
      "tau_m = 5000 * ms \n",
      "v_r   = 0    * mV \n",
      "v_th  = 1000 * mV # threshold potential\n",
      "\n",
      "eqs = '''\n",
      "dv/dt = -v/tau_m : volt\n",
      "'''\n",
      "\n",
      "# Build populations\n",
      "pops=[]\n",
      "conns=[]\n",
      "\n",
      "pops.append(PoissonGroup(82))            # Input Layer\n",
      "\n",
      "for (layer_num, layer_size) in enumerate(Layers_Weights):\n",
      "    pops.append(NeuronGroup(layer_size[0], eqs, threshold=v_th ,reset=v_r))   # Hidden Layer(s) + Output\n",
      "for (layer_num, weights) in enumerate(Layers_Weights):\n",
      "    c = Connection(pops[layer_num], pops[layer_num+1], 'v', weight=weights[1]*1000*mV)   # , sparseness=0.1)\n",
      "    conns.append(c)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 12
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "pops[1].v =0\n",
      "pops[1].v"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 13,
       "text": [
        "array([ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,\n",
        "        0.,  0.,  0.,  0.,  0.])"
       ]
      }
     ],
     "prompt_number": 13
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "conns[0][0,0]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 14,
       "text": [
        "-0.0068417279909194096"
       ]
      }
     ],
     "prompt_number": 14
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Model-Based Normalization"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 15
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def emulate(index):\n",
      "    in_value = train[0][index]\n",
      "    target = train[1][index]\n",
      "\n",
      "    pops[0].rate = in_value*Hz\n",
      "\n",
      "    Mv   = StateMonitor(pops[-1], 'v', record=True)\n",
      "    Cout = SpikeCounter(pops[-1])\n",
      "    Cmid = SpikeCounter(pops[1])\n",
      "    Cin  = SpikeCounter(pops[0])\n",
      "\n",
      "    Mout = SpikeMonitor(pops[-1])\n",
      "    Min  = SpikeMonitor(pops[0])\n",
      "\n",
      "    run(5.0*second)\n",
      "\n",
      "    print \" ------------- Sample: \",index, \" -------------\"\n",
      "    print \"Given Inputs:\",  \"                                                              Given Target:\", target\n",
      "    print in_value\n",
      "    \n",
      "    print \"\"\n",
      "    print \"Spiking\"\n",
      "    print \"Input Layer=\"         ,Cin.count\n",
      "    print \"First Hidden Layer=\"  ,Cmid.count\n",
      "    print \"Output Layer =\"       ,Cout.count\n",
      "\n",
      "    plot(Mv.times/ms, Mv[0]/mV)\n",
      "    plt.xlabel(\"Time (ms)\")\n",
      "    plt.ylabel(\"Voltage (mV)\")\n",
      "    show()\n",
      "    \n",
      "    \n",
      "    #raster_plot(Mout)\n",
      "    #show()\n",
      "    #raster_plot(Min)\n",
      "    #show()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 15
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "emulate(0)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " ------------- Sample:  0  -------------\n",
        "Given Inputs:                                                               Given Target: [ 0.]\n",
        "[ 16.80667114  11.27456379  10.51134396  13.86901093  16.37549973\n",
        "   6.03183746  11.30953884   1.27181816  18.33466339   9.54668045\n",
        "   7.23787355  18.04201508  15.045228     8.47550011   7.20095205   0.\n",
        "  15.18261719   8.3663702    7.65065527  12.85144901   0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           9.82641602   5.16712093]\n",
        "\n",
        "Spiking\n",
        "Input Layer= [92 62 50 61 88 40 71 14 91 49 37 89 73 52 35  0 79 42 34 59  0  0  0  0  0\n",
        "  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
        "  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
        "  0  0  0  0  0 52 43]\n",
        "First Hidden Layer= [ 36 381  30  31 383 381 381 382  37  41  51  27  32  35  35  30 385  30\n",
        "  37  33  29  31 382 381 381 381 381  30  31  31  34  46 398  35  26 395\n",
        "  32  43  31  30  27 384  29  30  32 385 381  34  26  35  30  32 388 381\n",
        "  27  35 381 388 385  30  33  29 381  29  28 382  28  30 381 381 386 381\n",
        " 387 381 381 384 381  31  43 385 381  32 388 384 392  48 381  30  30 385\n",
        " 381 387  37  46  51 381 382  30 382 386  29 381 387  44  32  32 381 389\n",
        " 384 395  31  49 389 381  35  32 381  27  30 387  30 381  31 386  37  51\n",
        "  32 381 381  51  37  29 381  32  29 381 381  32 381  31 381 383 382  30\n",
        "  45  33  45 383 381 382 382  36 381 384 381  30  32 384  28 384 382  33\n",
        "  49  58  33 381 386 381 385 382 381  38  28 383 384 393 382  29 382 389\n",
        "  29  34  45 386  33 381  30  45  30  32 383 384  40  31 381 381  52  27\n",
        "  48 381]\n",
        "Output Layer = [0]\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEVCAYAAADKN2OaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7fdcf8ffbc50>"
       ]
      }
     ],
     "prompt_number": 16
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "emulate(1)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " ------------- Sample:  1  -------------\n",
        "Given Inputs:                                                               Given Target: [ 0.]\n",
        "[ 20.18263435  11.30976295  15.59121037   0.          19.60983658\n",
        "   4.29410553   8.04539871  17.52112007  12.94768524   8.72885609\n",
        "   5.22983932   7.35402584  19.36964035   7.41294479   8.23366261\n",
        "   0.24747849   0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           8.08284569   8.95333004]\n",
        "\n",
        "Spiking\n",
        "Input Layer= [100  57  69   0 103  17  37  90  67  43  24  37  91  29  35   1   0   0\n",
        "   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
        "   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
        "   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
        "   0   0   0   0   0   0   0   0  38  45]\n",
        "First Hidden Layer= [  0 293   0   0 292 292 293 292   0   0   5   0   0   0   0   0 295   0\n",
        "   1   0   0   0 292 292 292 295 292   0   0   0   0   5 330   0   0 295\n",
        "   0   3   0   0   0 295   0   0   0 298 292   0   0   0   0   0 293 292\n",
        "   0   5 292 300 298   0   0   0 292   0   0 292   0   0 292 292 296 292\n",
        " 296 292 292 292 292   0   1 292 292   0 295 301 303   3 292   0   0 295\n",
        " 292 297   0   0   0 295 292   0 292 292   0 293 293   0   0   0 292 294\n",
        " 310 302   0   6 293 292   0   0 292   0   0 292   0 292   0 294   8   3\n",
        "   0 292 292   5   0   0 292   0   0 292 292   0 292   0 292 292 293   0\n",
        "   2   0   4 298 292 292 292   0 292 300 292   0   0 292   0 298 294   0\n",
        "   3   7   2 292 294 293 292 293 292   1   0 293 292 293 293   0 295 292\n",
        "   0   1   0 293   0 293   0   7   0   0 296 296   0   0 292 293   3   0\n",
        "   2 292]\n",
        "Output Layer = [0]\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZUAAAEVCAYAAAAo63jjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7fdce07e2d90>"
       ]
      }
     ],
     "prompt_number": 17
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "emulate(2)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " ------------- Sample:  2  -------------\n",
        "Given Inputs:                                                               Given Target: [ 1.]\n",
        "[  8.48667526   5.24193239   6.90290213   3.30880976  10.72842979\n",
        "   5.88240814   3.04137874   9.5026741    6.25680923   3.89672327\n",
        "   4.31376314   0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   1.67394209   5.04522991]\n",
        "\n",
        "Spiking\n",
        "Input Layer= [45 23 30 10 58 31 14 58 25 15 23  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
        "  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
        "  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
        "  0  0  0  0  0  3 21]\n",
        "First Hidden Layer= [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1\n",
        " 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 0 1 0 0 0 0\n",
        " 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0\n",
        " 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0\n",
        " 0 2 0 6 0 0 1 0 0 0 2 0 0 0 2 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
        " 0 0 0 0 1 0 0 0 0 2 0 0 0 0 0]\n",
        "Output Layer = [0]\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEPCAYAAABoekJnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJCIikhUlEhERycr/B5fRjDx4R6e8AAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f81ba80dcd0>"
       ]
      }
     ],
     "prompt_number": 22
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "                                                \n",
      "emulate(25)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " ------------- Sample:  25  -------------\n",
        "Given Inputs:                                                               Given Target: [ 1.]\n",
        "[ 17.38028526   0.39839554   9.40573692   0.          16.09911346\n",
        "   2.76607704  12.03481674   5.29946566  22.51858902   6.11639071\n",
        "  10.98604012   4.91439295  11.23464298   5.66585398  10.67743587\n",
        "   7.25254631  17.08579826  10.32454967   9.29169559   1.84202099\n",
        "  22.25771713  10.48590469   5.81002235   6.79015589  15.29386139\n",
        "   6.01898956   6.89363766   5.61283064  13.84298897   7.67221403\n",
        "   7.69472408  13.27347183   0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           0.           0.           0.           0.           0.           0.\n",
        "   0.           9.6785965    8.85594082]\n",
        "\n",
        "Spiking\n",
        "Input Layer= [ 92   3  41   0  83  12  54  23 111  25  51  21  49  28  46  41  81  60\n",
        "  62  12  97  45  29  35  65  30  29  31  71  28  55  60   0   0   0   0\n",
        "   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
        "   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
        "   0   0   0   0   0   0   0   0  54  62]\n",
        "First Hidden Layer= [ 0  0  0  0  0  0  1  0  0  0  0  0  0  0  0  4  4  0  0  0  2  1  0  0  0\n",
        "  0  0  0  0  3  0  1  0  0  2  0  0  0  0  0  0  2  1  0  0  0  5  0  0  0\n",
        "  1  0  0  0  3  0  1  0  0  0  1  1  0  0  5  0  2  8  0  0  0  0  0  0  0\n",
        "  0  4  4  0  0  0  0  0  2  0  0  0  1  0  0  6  0  0  3  0  0  0  0  0  0\n",
        "  1  7  0  0  0  0  0  0  0  0  0  0  0  0  0  4  5  3  0  1  1  0  0  0  2\n",
        "  0  0  3  0  9  0  0  2  0  0  0  0  0  0  0  1  0  0  0  0  0  4  4  2  2\n",
        "  0  7  0  0  0 23  0  0  5  0  1  5  0  0  3  2  0  1  0  5  0  1  0  0  0\n",
        "  0  0  0  0  0  0  1  0  0  0  0  1  0  1  2  2  0  0  4  3  0  0  2  0  0]\n",
        "Output Layer = [0]\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZgAAAEPCAYAAAB/WNKuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f81c21618d0>"
       ]
      }
     ],
     "prompt_number": 23
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}