Simple visualization for the weights and outputs of convolution layer in Lasagne

mnist.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:a015cb7ff876b2401fc8d24285c3effdb05e3eb9b159129c1fffcdeb566bbfe5"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from __future__ import division, print_function\n",
      "\n",
      "IMAGE_HEIGHT = 28\n",
      "IMAGE_WIDTH = 28\n",
      "\n",
      "import scipy\n",
      "import theano\n",
      "import numpy as np\n",
      "\n",
      "import matplotlib.pyplot as plt\n",
      "from matplotlib import gridspec\n",
      "import seaborn as sns\n",
      "%matplotlib inline\n",
      "\n",
      "from sklearn.datasets import fetch_mldata\n",
      "from sklearn.cross_validation import train_test_split\n",
      "\n",
      "from nolearn.lasagne import NeuralNet\n",
      "from lasagne import layers\n",
      "from lasagne import nonlinearities\n",
      "from lasagne.updates import nesterov_momentum"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "Using gpu device 0: GeForce GTX 670\n"
       ]
      }
     ],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "mnist = fetch_mldata('MNIST original')\n",
      "X = mnist.data.astype(np.float32) / 255\n",
      "y = mnist.target.astype(np.int32)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25, random_state=42)\n",
      "\n",
      "X_train = X_train.reshape(-1, 1, IMAGE_HEIGHT, IMAGE_WIDTH)\n",
      "X_test = X_test.reshape(-1, 1, IMAGE_HEIGHT, IMAGE_WIDTH)\n",
      "\n",
      "print(X_train.shape, X_test.shape)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "(52500, 1, 28, 28) (17500, 1, 28, 28)\n"
       ]
      }
     ],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "clf = NeuralNet(\n",
      "    layers=[\n",
      "        ('input', layers.InputLayer),\n",
      "\n",
      "        ('conv1', layers.Conv2DLayer),\n",
      "        ('pool1', layers.MaxPool2DLayer),\n",
      "        ('dropout1', layers.DropoutLayer),\n",
      "\n",
      "        ('conv2', layers.Conv2DLayer),\n",
      "        ('pool2', layers.MaxPool2DLayer),\n",
      "        ('dropout2', layers.DropoutLayer),\n",
      "\n",
      "        ('hidden3', layers.DenseLayer),\n",
      "        ('dropout3', layers.DropoutLayer),\n",
      "        ('output', layers.DenseLayer),\n",
      "    ],\n",
      "\n",
      "    input_shape=(None, 1, IMAGE_HEIGHT, IMAGE_WIDTH),\n",
      "    output_num_units=10,\n",
      "    output_nonlinearity=nonlinearities.softmax,\n",
      "\n",
      "    conv1_filter_size=(5,5), conv1_num_filters=32,\n",
      "    pool1_ds=(2,2),\n",
      "    dropout1_p=0.1,\n",
      "\n",
      "    conv2_filter_size=(5,5), conv2_num_filters=32,\n",
      "    pool2_ds=(2,2),\n",
      "    dropout2_p=0.2,\n",
      "\n",
      "    hidden3_num_units=256,\n",
      "    dropout3_p=0.5,\n",
      "\n",
      "    update=nesterov_momentum,\n",
      "    update_learning_rate=0.01,\n",
      "    update_momentum=0.9,\n",
      "\n",
      "    max_epochs=50,\n",
      "    verbose=True\n",
      ")\n",
      "\n",
      "clf.fit(X_train, y_train)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "  InputLayer        \t(None, 1, 28, 28)   \tproduces     784 outputs\n",
        "  Conv2DLayer       \t(None, 32, 24, 24)  \tproduces   18432 outputs\n",
        "  MaxPool2DLayer    \t(None, 32, 12, 12)  \tproduces    4608 outputs\n",
        "  DropoutLayer      \t(None, 32, 12, 12)  \tproduces    4608 outputs\n",
        "  Conv2DLayer       \t(None, 32, 8, 8)    \tproduces    2048 outputs\n",
        "  MaxPool2DLayer    \t(None, 32, 4, 4)    \tproduces     512 outputs\n",
        "  DropoutLayer      \t(None, 32, 4, 4)    \tproduces     512 outputs\n",
        "  DenseLayer        \t(None, 256)         \tproduces     256 outputs\n",
        "  DropoutLayer      \t(None, 256)         \tproduces     256 outputs\n",
        "  DenseLayer        \t(None, 10)          \tproduces      10 outputs\n",
        "\n",
        " Epoch  |  Train loss  |  Valid loss  |  Train / Val  |  Valid acc  |  Dur\n",
        "--------|--------------|--------------|---------------|-------------|-------"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "     1  |  \u001b[94m  0.627278\u001b[0m  |  \u001b[32m  0.128318\u001b[0m  |     4.888467  |     96.28%  |  10.8s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "     2  |  \u001b[94m  0.179111\u001b[0m  |  \u001b[32m  0.084250\u001b[0m  |     2.125933  |     97.41%  |  10.8s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "     3  |  \u001b[94m  0.129697\u001b[0m  |  \u001b[32m  0.067983\u001b[0m  |     1.907774  |     97.99%  |  10.9s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "     4  |  \u001b[94m  0.105651\u001b[0m  |  \u001b[32m  0.057781\u001b[0m  |     1.828457  |     98.23%  |  10.9s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "     5  |  \u001b[94m  0.088195\u001b[0m  |  \u001b[32m  0.050079\u001b[0m  |     1.761114  |     98.32%  |  10.9s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "     6  |  \u001b[94m  0.078292\u001b[0m  |  \u001b[32m  0.045419\u001b[0m  |     1.723750  |     98.54%  |  10.9s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "     7  |  \u001b[94m  0.070394\u001b[0m  |  \u001b[32m  0.043152\u001b[0m  |     1.631301  |     98.51%  |  10.9s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "     8  |  \u001b[94m  0.064884\u001b[0m  |  \u001b[32m  0.039497\u001b[0m  |     1.642763  |     98.55%  |  10.9s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "     9  |  \u001b[94m  0.059731\u001b[0m  |  \u001b[32m  0.038294\u001b[0m  |     1.559795  |     98.73%  |  10.9s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "    10  |  \u001b[94m  0.054741\u001b[0m  |  \u001b[32m  0.033703\u001b[0m  |     1.624198  |     98.89%  |  10.9s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "    11  |  \u001b[94m  0.053225\u001b[0m  |    0.034623  |     1.537266  |     98.73%  |  10.9s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "    12  |  \u001b[94m  0.050447\u001b[0m  |  \u001b[32m  0.032541\u001b[0m  |     1.550247  |     99.03%  |  10.9s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "    13  |  \u001b[94m  0.048992\u001b[0m  |  \u001b[32m  0.029131\u001b[0m  |     1.681787  |     99.11%  |  11.0s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "    14  |  \u001b[94m  0.045930\u001b[0m  |  \u001b[32m  0.028829\u001b[0m  |     1.593170  |     99.02%  |  11.0s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "    15  |  \u001b[94m  0.044429\u001b[0m  |  \u001b[32m  0.027750\u001b[0m  |     1.601082  |     99.14%  |  11.0s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        "    16  |  \u001b[94m  0.043024\u001b[0m  |  \u001b[32m  0.027155\u001b[0m  |     1.584405  |     99.08%  |  11.0s"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "\n",
        ">"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "/home/felixlau/devel/Lasagne/src/theano/theano/sandbox/rng_mrg.py:1188: UserWarning: MRG_RandomStreams Can't determine #streams from size (Shape.0), guessing 60*256\n",
        "  nstreams = self.n_streams(size)\n"
       ]
      },
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        " /home/felixlau/anaconda/lib/python2.7/site-packages/nolearn-0.5b2dev-py2.7.egg/nolearn/lasagne.py(153)fit()\n",
        "-> return self\n"
       ]
      },
      {
       "ename": "KeyboardInterrupt",
       "evalue": "",
       "output_type": "pyerr",
       "traceback": [
        "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m\n\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
        "\u001b[1;32m<ipython-input-4-f1e2095815ef>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m     39\u001b[0m )\n\u001b[0;32m     40\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 41\u001b[1;33m \u001b[0mclf\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX_train\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_train\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
        "\u001b[1;32m/home/felixlau/anaconda/lib/python2.7/site-packages/nolearn-0.5b2dev-py2.7.egg/nolearn/lasagne.pyc\u001b[0m in \u001b[0;36mfit\u001b[1;34m(self, X, y)\u001b[0m\n\u001b[0;32m    151\u001b[0m         \u001b[1;32mexcept\u001b[0m \u001b[0mKeyboardInterrupt\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    152\u001b[0m             \u001b[0mpdb\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mset_trace\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 153\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    154\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    155\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mtrain_loop\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/home/felixlau/anaconda/lib/python2.7/site-packages/nolearn-0.5b2dev-py2.7.egg/nolearn/lasagne.pyc\u001b[0m in \u001b[0;36mfit\u001b[1;34m(self, X, y)\u001b[0m\n\u001b[0;32m    151\u001b[0m         \u001b[1;32mexcept\u001b[0m \u001b[0mKeyboardInterrupt\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    152\u001b[0m             \u001b[0mpdb\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mset_trace\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 153\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    154\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    155\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mtrain_loop\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/home/felixlau/anaconda/lib/python2.7/bdb.pyc\u001b[0m in \u001b[0;36mtrace_dispatch\u001b[1;34m(self, frame, event, arg)\u001b[0m\n\u001b[0;32m     47\u001b[0m             \u001b[1;32mreturn\u001b[0m \u001b[1;31m# None\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     48\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mevent\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;34m'line'\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 49\u001b[1;33m             \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdispatch_line\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     50\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mevent\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;34m'call'\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     51\u001b[0m             \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdispatch_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0marg\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/home/felixlau/anaconda/lib/python2.7/bdb.pyc\u001b[0m in \u001b[0;36mdispatch_line\u001b[1;34m(self, frame)\u001b[0m\n\u001b[0;32m     65\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mdispatch_line\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mframe\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     66\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstop_here\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mor\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbreak_here\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 67\u001b[1;33m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0muser_line\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     68\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mquitting\u001b[0m\u001b[1;33m:\u001b[0m \u001b[1;32mraise\u001b[0m \u001b[0mBdbQuit\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     69\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtrace_dispatch\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/home/felixlau/anaconda/lib/python2.7/pdb.pyc\u001b[0m in \u001b[0;36muser_line\u001b[1;34m(self, frame)\u001b[0m\n\u001b[0;32m    156\u001b[0m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_wait_for_mainpyfile\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    157\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbp_commands\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 158\u001b[1;33m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0minteraction\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mNone\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    159\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    160\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mbp_commands\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mframe\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/home/felixlau/anaconda/lib/python2.7/pdb.pyc\u001b[0m in \u001b[0;36minteraction\u001b[1;34m(self, frame, traceback)\u001b[0m\n\u001b[0;32m    208\u001b[0m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msetup\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mframe\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtraceback\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    209\u001b[0m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mprint_stack_entry\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstack\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcurindex\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 210\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcmdloop\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    211\u001b[0m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mforget\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    212\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/home/felixlau/anaconda/lib/python2.7/cmd.pyc\u001b[0m in \u001b[0;36mcmdloop\u001b[1;34m(self, intro)\u001b[0m\n\u001b[0;32m    128\u001b[0m                     \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0muse_rawinput\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    129\u001b[0m                         \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 130\u001b[1;33m                             \u001b[0mline\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mraw_input\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mprompt\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    131\u001b[0m                         \u001b[1;32mexcept\u001b[0m \u001b[0mEOFError\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    132\u001b[0m                             \u001b[0mline\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;34m'EOF'\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/home/felixlau/anaconda/lib/python2.7/site-packages/IPython/kernel/zmq/ipkernel.pyc\u001b[0m in \u001b[0;36m<lambda>\u001b[1;34m(prompt)\u001b[0m\n\u001b[0;32m    361\u001b[0m         \u001b[1;31m# raw_input in the user namespace.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    362\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mcontent\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'allow_stdin'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 363\u001b[1;33m             \u001b[0mraw_input\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mlambda\u001b[0m \u001b[0mprompt\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m''\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_raw_input\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mprompt\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mident\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mparent\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    364\u001b[0m             \u001b[0minput\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mlambda\u001b[0m \u001b[0mprompt\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m''\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0meval\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mraw_input\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mprompt\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    365\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;32m/home/felixlau/anaconda/lib/python2.7/site-packages/IPython/kernel/zmq/ipkernel.pyc\u001b[0m in \u001b[0;36m_raw_input\u001b[1;34m(self, prompt, ident, parent)\u001b[0m\n\u001b[0;32m    763\u001b[0m             \u001b[1;32mexcept\u001b[0m \u001b[0mKeyboardInterrupt\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    764\u001b[0m                 \u001b[1;31m# re-raise KeyboardInterrupt, to truncate traceback\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 765\u001b[1;33m                 \u001b[1;32mraise\u001b[0m \u001b[0mKeyboardInterrupt\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    766\u001b[0m             \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    767\u001b[0m                 \u001b[1;32mbreak\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
        "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
       ]
      }
     ],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "clf.score(X_test, y_test)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 5,
       "text": [
        "0.9906285714285714"
       ]
      }
     ],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "clf_layers = clf.get_all_layers()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 6
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "layer = clf_layers[1]\n",
      "W = layer.W.get_value()\n",
      "b = layer.b.get_value()\n",
      "f = [w + bb for w, bb in zip(W, b)]\n",
      "\n",
      "gs = gridspec.GridSpec(6, 6)\n",
      "for i in range(layer.num_filters):\n",
      "    g = gs[i]\n",
      "    ax = plt.subplot(g)\n",
      "    ax.grid()\n",
      "    ax.set_xticks([])\n",
      "    ax.set_yticks([])\n",
      "    ax.imshow(f[i][0])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAADtCAYAAAA7myeeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\njIwTkEkzIyMj4wRk0szIyMg4Af8PD7tB8QelhHMAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f6bee7ccf50>"
       ]
      }
     ],
     "prompt_number": 7
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "output = layer.get_output_for(X_test[:1]).eval()[0]\n",
      "gs = gridspec.GridSpec(6, 6)\n",
      "for i in range(layer.num_filters):\n",
      "    g = gs[i]\n",
      "    ax = plt.subplot(g)\n",
      "    ax.grid()\n",
      "    ax.set_xticks([])\n",
      "    ax.set_yticks([])\n",
      "    ax.imshow(output[i])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAADtCAYAAAA7myeeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f6bef9ed690>"
       ]
      }
     ],
     "prompt_number": 8
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "layer = clf_layers[4]\n",
      "W = layer.W.get_value()\n",
      "b = layer.b.get_value()\n",
      "f = [w + bb for w, bb in zip(W, b)]\n",
      "\n",
      "gs = gridspec.GridSpec(6, 6)\n",
      "for i in range(layer.num_filters):\n",
      "    g = gs[i]\n",
      "    ax = plt.subplot(g)\n",
      "    ax.grid()\n",
      "    ax.set_xticks([])\n",
      "    ax.set_yticks([])\n",
      "    ax.imshow(f[i][0])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAADtCAYAAAA7myeeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f6befa27690>"
       ]
      }
     ],
     "prompt_number": 9
    }
   ],
   "metadata": {}
  }
 ]
}