keras/convmnist/keras_cnn_mnist.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CNN (keras example) on mnist (from keras's examples) with convolution visualization\n",
    "\n",
    "This is https://github.com/fchollet/keras/blob/master/examples/mnist_cnn.py"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## For keras < v1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "os.environ['THEANO_FLAGS']='mode=FAST_RUN,device=gpu,floatX=float32'\n",
    "# This gives a 'perform' error in compile\n",
    "#os.environ['THEANO_FLAGS']='mode=FAST_COMPILE,device=gpu1,floatX=float32'\n",
    "#os.environ['THEANO_FLAGS']='device=gpu0'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Couldn't import dot_parser, loading of dot files will not be possible.\n",
      "gpu\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using gpu device 0: Tesla M2075\n"
     ]
    }
   ],
   "source": [
    "import theano\n",
    "print theano.config.device"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\n    Train a simple convnet on the MNIST dataset.\\n\\n    Run on GPU: THEANO_FLAGS=mode=FAST_RUN,device=gpu,floatX=float32 python mnist_cnn.py\\n\\n    Get to 99.25% test accuracy after 12 epochs (there is still a lot of margin for parameter tuning).\\n    16 seconds per epoch on a GRID K520 GPU.\\n'"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from __future__ import absolute_import\n",
    "from __future__ import print_function\n",
    "import pylab as pl\n",
    "import matplotlib.cm as cm\n",
    "import numpy as np\n",
    "np.random.seed(1337) # for reproducibility\n",
    "\n",
    "from keras.datasets import mnist\n",
    "from keras.models import Sequential\n",
    "from keras.layers.core import Dense, Dropout, Activation, Flatten\n",
    "from keras.layers.convolutional import Convolution2D, MaxPooling2D\n",
    "from keras.utils import np_utils\n",
    "\n",
    "'''\n",
    "    Train a simple convnet on the MNIST dataset.\n",
    "\n",
    "    Run on GPU: THEANO_FLAGS=mode=FAST_RUN,device=gpu,floatX=float32 python mnist_cnn.py\n",
    "\n",
    "    Get to 99.25% test accuracy after 12 epochs (there is still a lot of margin for parameter tuning).\n",
    "    16 seconds per epoch on a GRID K520 GPU.\n",
    "'''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "np.set_printoptions(precision=5, suppress=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train shape: (60000, 1, 28, 28)\n",
      "60000 train samples\n",
      "10000 test samples\n"
     ]
    }
   ],
   "source": [
    "nb_classes = 10\n",
    "\n",
    "# the data, shuffled and split between tran and test sets\n",
    "(X_train, y_train), (X_test, y_test) = mnist.load_data()\n",
    "\n",
    "X_train = X_train.reshape(X_train.shape[0], 1, 28, 28)\n",
    "X_test = X_test.reshape(X_test.shape[0], 1, 28, 28)\n",
    "X_train = X_train.astype(\"float32\")\n",
    "X_test = X_test.astype(\"float32\")\n",
    "X_train /= 255\n",
    "X_test /= 255\n",
    "print('X_train shape:', X_train.shape)\n",
    "print(X_train.shape[0], 'train samples')\n",
    "print(X_test.shape[0], 'test samples')\n",
    "\n",
    "# convert class vectors to binary class matrices\n",
    "Y_train = np_utils.to_categorical(y_train, nb_classes)\n",
    "Y_test = np_utils.to_categorical(y_test, nb_classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "label :  [ 0.  0.  0.  0.  0.  1.  0.  0.  0.  0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPwAAAD8CAYAAABTq8lnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f86381fee50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "i = 4600\n",
    "pl.imshow(X_train[i, 0], interpolation='nearest', cmap=cm.binary)\n",
    "print(\"label : \", Y_train[i,:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "\n",
    "model.add(Convolution2D(32, 1, 3, 3, border_mode='full')) \n",
    "convout1 = Activation('relu')\n",
    "model.add(convout1)\n",
    "model.add(Convolution2D(32, 32, 3, 3))\n",
    "\n",
    "convout2 = Activation('relu')\n",
    "model.add(convout2)\n",
    "model.add(MaxPooling2D(poolsize=(2, 2)))\n",
    "model.add(Dropout(0.25))\n",
    "\n",
    "model.add(Flatten())\n",
    "model.add(Dense(32*196, 128))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Dropout(0.5))\n",
    "\n",
    "model.add(Dense(128, nb_classes))\n",
    "model.add(Activation('softmax'))\n",
    "\n",
    "model.compile(loss='categorical_crossentropy', optimizer='adadelta')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 60000 samples, validate on 10000 samples\n",
      "Epoch 0\n",
      "60000/60000 [==============================] - 79s - loss: 0.2596 - acc: 0.9200 - val_loss: 0.0548 - val_acc: 0.9826\n",
      "Epoch 1\n",
      "60000/60000 [==============================] - 79s - loss: 0.0961 - acc: 0.9713 - val_loss: 0.0441 - val_acc: 0.9861\n",
      "Epoch 2\n",
      "60000/60000 [==============================] - 79s - loss: 0.0735 - acc: 0.9782 - val_loss: 0.0426 - val_acc: 0.9860\n",
      "Epoch 3\n",
      "60000/60000 [==============================] - 79s - loss: 0.0617 - acc: 0.9816 - val_loss: 0.0330 - val_acc: 0.9885\n",
      "Epoch 4\n",
      "60000/60000 [==============================] - 79s - loss: 0.0513 - acc: 0.9844 - val_loss: 0.0277 - val_acc: 0.9913\n",
      "Epoch 5\n",
      "60000/60000 [==============================] - 79s - loss: 0.0467 - acc: 0.9859 - val_loss: 0.0305 - val_acc: 0.9895\n",
      "Epoch 6\n",
      "60000/60000 [==============================] - 79s - loss: 0.0416 - acc: 0.9868 - val_loss: 0.0291 - val_acc: 0.9911\n",
      "Epoch 7\n",
      "60000/60000 [==============================] - 79s - loss: 0.0377 - acc: 0.9882 - val_loss: 0.0289 - val_acc: 0.9909\n",
      "Epoch 8\n",
      "60000/60000 [==============================] - 79s - loss: 0.0349 - acc: 0.9889 - val_loss: 0.0263 - val_acc: 0.9914\n",
      "Epoch 9\n",
      "60000/60000 [==============================] - 79s - loss: 0.0333 - acc: 0.9900 - val_loss: 0.0258 - val_acc: 0.9916\n",
      "Epoch 10\n",
      "60000/60000 [==============================] - 79s - loss: 0.0305 - acc: 0.9906 - val_loss: 0.0240 - val_acc: 0.9924\n",
      "Epoch 11\n",
      "60000/60000 [==============================] - 79s - loss: 0.0273 - acc: 0.9918 - val_loss: 0.0299 - val_acc: 0.9906\n",
      "Test score: 0.0299028589356\n",
      "Test accuracy: 0.9906\n"
     ]
    }
   ],
   "source": [
    "WEIGHTS_FNAME = 'mnist_cnn_weights.hdf'\n",
    "if True and os.path.exists(WEIGHTS_FNAME):\n",
    "    # Just change the True to false to force re-training\n",
    "    print('Loading existing weights')\n",
    "    model.load_weights(WEIGHTS_FNAME)\n",
    "else:\n",
    "    batch_size = 128\n",
    "    nb_epoch = 12\n",
    "    model.fit(X_train, Y_train, batch_size=batch_size, nb_epoch=nb_epoch,\n",
    "              show_accuracy=True, verbose=1, validation_data=(X_test, Y_test))\n",
    "    model.save_weights(WEIGHTS_FNAME)\n",
    "score = model.evaluate(X_test, Y_test, show_accuracy=True, verbose=0)\n",
    "print('Test score:', score[0])\n",
    "print('Test accuracy:', score[1])\n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.  0.  1.  0.  0.  0.  0.  0.  0.  0.]\n",
      " [ 0.  1.  0.  0.  0.  0.  0.  0.  0.  0.]\n",
      " [ 1.  0.  0.  0.  0.  0.  0.  0.  0.  0.]\n",
      " [ 0.  0.  0.  0.  1.  0.  0.  0.  0.  0.]]\n",
      "[[ 0.  0.  1.  0.  0.  0.  0.  0.  0.  0.]\n",
      " [ 0.  1.  0.  0.  0.  0.  0.  0.  0.  0.]\n",
      " [ 1.  0.  0.  0.  0.  0.  0.  0.  0.  0.]\n",
      " [ 0.  0.  0.  0.  1.  0.  0.  0.  0.  0.]]\n"
     ]
    }
   ],
   "source": [
    "print(model.predict(X_test[1:5]))\n",
    "print(Y_test[1:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "Y_pred = model.predict(X_test)\n",
    "# Convert one-hot to index\n",
    "y_pred = np.argmax(Y_pred, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.99      0.99      0.99       980\n",
      "          1       0.99      1.00      1.00      1135\n",
      "          2       0.99      0.99      0.99      1032\n",
      "          3       0.98      1.00      0.99      1010\n",
      "          4       0.99      0.99      0.99       982\n",
      "          5       0.99      0.99      0.99       892\n",
      "          6       0.99      0.99      0.99       958\n",
      "          7       0.99      0.99      0.99      1028\n",
      "          8       1.00      0.98      0.99       974\n",
      "          9       0.99      0.98      0.99      1009\n",
      "\n",
      "avg / total       0.99      0.99      0.99     10000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Convolution visualizations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "convout1_f = theano.function([model.get_input(train=False)], convout1.get_output(train=False))\n",
    "#convout2_f = theano.function([model.get_input(train=False)], convout2.get_output(train=False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# utility functions\n",
    "from mpl_toolkits.axes_grid1 import make_axes_locatable\n",
    "\n",
    "def nice_imshow(ax, data, vmin=None, vmax=None, cmap=None):\n",
    "    \"\"\"Wrapper around pl.imshow\"\"\"\n",
    "    if cmap is None:\n",
    "        cmap = cm.jet\n",
    "    if vmin is None:\n",
    "        vmin = data.min()\n",
    "    if vmax is None:\n",
    "        vmax = data.max()\n",
    "    divider = make_axes_locatable(ax)\n",
    "    cax = divider.append_axes(\"right\", size=\"5%\", pad=0.05)\n",
    "    im = ax.imshow(data, vmin=vmin, vmax=vmax, interpolation='nearest', cmap=cmap)\n",
    "    pl.colorbar(im, cax=cax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAR4AAAEKCAYAAADEjjpbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f8617f06ed0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "i = 4600\n",
    "\n",
    "# Visualize the first layer of convolutions on an input image\n",
    "X = X_test[i:i+1]\n",
    "\n",
    "pl.figure()\n",
    "pl.title('input')\n",
    "nice_imshow(pl.gca(), np.squeeze(X), vmin=0, vmax=1, cmap=cm.binary)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy.ma as ma\n",
    "def make_mosaic(imgs, nrows, ncols, border=1):\n",
    "    \"\"\"\n",
    "    Given a set of images with all the same shape, makes a\n",
    "    mosaic with nrows and ncols\n",
    "    \"\"\"\n",
    "    nimgs = imgs.shape[0]\n",
    "    imshape = imgs.shape[1:]\n",
    "    \n",
    "    mosaic = ma.masked_all((nrows * imshape[0] + (nrows - 1) * border,\n",
    "                            ncols * imshape[1] + (ncols - 1) * border),\n",
    "                            dtype=np.float32)\n",
    "    \n",
    "    paddedh = imshape[0] + border\n",
    "    paddedw = imshape[1] + border\n",
    "    for i in xrange(nimgs):\n",
    "        row = int(np.floor(i / ncols))\n",
    "        col = i % ncols\n",
    "        \n",
    "        mosaic[row * paddedh:row * paddedh + imshape[0],\n",
    "               col * paddedw:col * paddedw + imshape[1]] = imgs[i]\n",
    "    return mosaic\n",
    "\n",
    "#pl.imshow(make_mosaic(np.random.random((9, 10, 10)), 3, 3, border=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "W shape :  (32, 3, 3)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA30AAANFCAYAAADPo5V5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n2wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f8618629250>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize weights\n",
    "W = model.layers[0].W.get_value(borrow=True)\n",
    "W = np.squeeze(W)\n",
    "print(\"W shape : \", W.shape)\n",
    "\n",
    "pl.figure(figsize=(15, 15))\n",
    "pl.title('conv1 weights')\n",
    "nice_imshow(pl.gca(), make_mosaic(W, 6, 6), cmap=cm.binary)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C1 shape :  (32, 30, 30)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3sAAAOiCAYAAADJ/M4EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nPQAAAABIEJ09AAAAAEjQ/we/2grx17f8JQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f8617c10250>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize convolution result (after activation)\n",
    "C1 = convout1_f(X)\n",
    "C1 = np.squeeze(C1)\n",
    "print(\"C1 shape : \", C1.shape)\n",
    "\n",
    "pl.figure(figsize=(15, 15))\n",
    "pl.suptitle('convout1')\n",
    "nice_imshow(pl.gca(), make_mosaic(C1, 6, 6), cmap=cm.binary)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}