fashionmnist1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ilayda/Desktop/anaconda3/lib/python3.6/site-packages/h5py/__init__.py:34: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting data/fashion/train-images-idx3-ubyte.gz\n",
      "Extracting data/fashion/train-labels-idx1-ubyte.gz\n",
      "Extracting data/fashion/t10k-images-idx3-ubyte.gz\n",
      "Extracting data/fashion/t10k-labels-idx1-ubyte.gz\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "fashion_mnist = input_data.read_data_sets('data/fashion', source_url='http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/',one_hot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training set (images) shape: (55000, 784)\n",
      "Training set (labels) shape: (55000, 10)\n",
      "Test set (images) shape: (10000, 784)\n",
      "Test set (labels) shape: (10000, 10)\n"
     ]
    }
   ],
   "source": [
    "# Shapes of training set\n",
    "print(\"Training set (images) shape: {shape}\".format(shape=fashion_mnist.train.images.shape))\n",
    "print(\"Training set (labels) shape: {shape}\".format(shape=fashion_mnist.train.labels.shape))\n",
    "\n",
    "# Shapes of test set\n",
    "print(\"Test set (images) shape: {shape}\".format(shape=fashion_mnist.test.images.shape))\n",
    "print(\"Test set (labels) shape: {shape}\".format(shape=fashion_mnist.test.labels.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Create dictionary of target classes\n",
    "label_dict = {\n",
    " 0: \"T-shirt/top\",\n",
    " 1: \"Trouser\",\n",
    " 2: \"Pullover\",\n",
    " 3: \"Dress\",\n",
    " 4: \"Coat\",\n",
    " 5: \"Sandal\",\n",
    " 6: \"Shirt\",\n",
    " 7: \"Sneaker\",\n",
    " 8: \"Bag\",\n",
    " 9: \"Ankle boot\"\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y = 3 (Dress)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngGYPHMVwIwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c1d2b2898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y = 7 (Sneaker)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c1ffca5f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sample_1 = fashion_mnist.train.images[47].reshape(28,28)\n",
    "# Get corresponding integer label from one-hot encoded data\n",
    "sample_label_1 = np.where(fashion_mnist.train.labels[47] == 1)[0][0]\n",
    "# Plot sample\n",
    "print(\"y = {label_index} ({label})\".format(label_index=sample_label_1, label=label_dict[sample_label_1]))\n",
    "plt.imshow(sample_1, cmap='Greys')\n",
    "plt.show()\n",
    "\n",
    "# Sample 2\n",
    "\n",
    "# Get 28x28 image\n",
    "sample_2 = fashion_mnist.train.images[23].reshape(28,28)\n",
    "# Get corresponding integer label from one-hot encoded data\n",
    "sample_label_2 = np.where(fashion_mnist.train.labels[23] == 1)[0][0]\n",
    "# Plot sample\n",
    "print(\"y = {label_index} ({label})\".format(label_index=sample_label_2, label=label_dict[sample_label_2]))\n",
    "plt.imshow(sample_2, cmap='Greys')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def var(name, shape, init=None, std=None):\n",
    "    if init is None:\n",
    "        if std is None:\n",
    "            std = (2./shape[0])**0.5\n",
    "        init = tf.truncated_normal_initializer(stddev=std)\n",
    "    return tf.get_variable(name=name, shape=shape,\n",
    "                                  dtype=tf.float32, initializer=init)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From <ipython-input-7-4b2f21eae19d>:23: softmax_cross_entropy_with_logits (from tensorflow.python.ops.nn_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "\n",
      "Future major versions of TensorFlow will allow gradients to flow\n",
      "into the labels input on backprop by default.\n",
      "\n",
      "See tf.nn.softmax_cross_entropy_with_logits_v2.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "g = tf.Graph()\n",
    "with g.as_default():\n",
    "    X = tf.placeholder(dtype=tf.float32, shape=[None, 28*28])\n",
    "    Y = tf.placeholder(dtype=tf.float32, shape=[None, 10])\n",
    "    \n",
    "    #hidden layer\n",
    "    W1 = var('W1',[28*28, 100])\n",
    "    b1 = var('b1',[100],tf.constant_initializer(0.1))\n",
    "    out1 = tf.nn.relu(tf.matmul(X, W1) + b1)\n",
    "    \n",
    "    #output layer\n",
    "    W2 = var('W2',[100,10])\n",
    "    b2 = var('b2',[10],tf.constant_initializer(0.1))\n",
    "    logits =tf.matmul(out1, W2) + b2\n",
    "    \n",
    "    #accuracy\n",
    "    pred = tf.argmax(logits, axis=1)\n",
    "    truth = tf.argmax(Y, axis=1)\n",
    "    match = tf.cast(tf.equal(pred,truth), tf.float32)\n",
    "    acc = tf.reduce_mean(match)\n",
    "    \n",
    "    loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(\n",
    "    logits=logits, labels=Y))\n",
    "    step = tf.train.AdamOptimizer().minimize(loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sess = tf.InteractiveSession(graph=g)\n",
    "tf.global_variables_initializer().run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9999) acc: 0.8800\r"
     ]
    }
   ],
   "source": [
    "bsize = 100\n",
    "for I in range(10000):\n",
    "    x,y = fashion_mnist.train.next_batch(bsize)\n",
    "    _, acc_ = sess.run([step, acc], feed_dict={X:x, Y:y})\n",
    "    print('%d) acc: %2.4f' % (I,acc_), end = '\\r')\n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9197636"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sess.run(acc, feed_dict={X:fashion_mnist.train.images, Y:fashion_mnist.train.labels})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8791"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sess.run(acc, feed_dict={X:fashion_mnist.test.images, Y:fashion_mnist.test.labels})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0) loss: 0.2475\n",
      "100) loss: 0.0086\n",
      "200) loss: 0.0045\n",
      "300) loss: 0.0031\n",
      "400) loss: 0.0023\n",
      "500) loss: 0.0018\n",
      "600) loss: 0.0014\n",
      "700) loss: 0.0012\n",
      "800) loss: 0.0010\n",
      "900) loss: 0.0009\n"
     ]
    }
   ],
   "source": [
    "loss_list = []\n",
    "for i in range(1000):\n",
    "    _,loss_ = sess.run([step, loss],feed_dict={X:x, Y:y})\n",
    "    loss_list.append(loss_)\n",
    "    if not i%100:\n",
    "        print('%d) loss: %2.4f'  %(i, loss_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD8CAYAAACb4nSYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nWJPchtSiAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c26697470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(loss_list[10:1000],c='r')\n",
    "plt.legend(['loss function'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}