1-3. basic mnist.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MNIST"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "https://en.wikipedia.org/wiki/MNIST_database"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/parksurk/.virtualenvs/cv/lib/python3.5/site-packages/matplotlib/font_manager.py:273: UserWarning: Matplotlib is building the font cache using fc-list. This may take a moment.\n",
      "  warnings.warn('Matplotlib is building the font cache using fc-list. This may take a moment.')\n",
      "/home/parksurk/.virtualenvs/cv/lib/python3.5/site-packages/matplotlib/font_manager.py:273: UserWarning: Matplotlib is building the font cache using fc-list. This may take a moment.\n",
      "  warnings.warn('Matplotlib is building the font cache using fc-list. This may take a moment.')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Packages are loaded!!!\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "\n",
    "%matplotlib inline\n",
    "print (\"Packages are loaded!!!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Download and Extract MNIST dataset\n",
      "Successfully downloaded train-images-idx3-ubyte.gz 9912422 bytes.\n",
      "Extracting data/train-images-idx3-ubyte.gz\n",
      "Successfully downloaded train-labels-idx1-ubyte.gz 28881 bytes.\n",
      "Extracting data/train-labels-idx1-ubyte.gz\n",
      "Successfully downloaded t10k-images-idx3-ubyte.gz 1648877 bytes.\n",
      "Extracting data/t10k-images-idx3-ubyte.gz\n",
      "Successfully downloaded t10k-labels-idx1-ubyte.gz 4542 bytes.\n",
      "Extracting data/t10k-labels-idx1-ubyte.gz\n",
      "type of 'mnist' is <class 'tensorflow.contrib.learn.python.learn.datasets.base.Datasets'>\n",
      "number of train data is 55000\n",
      "number of test data is 10000\n"
     ]
    }
   ],
   "source": [
    "print (\"Download and Extract MNIST dataset\")\n",
    "mnist = input_data.read_data_sets('data/', one_hot=True)\n",
    "print\n",
    "print (\"type of 'mnist' is %s\" % (type(mnist)))\n",
    "print (\"number of train data is %d\" %(mnist.train.num_examples))\n",
    "print (\"number of test data is %d\" %(mnist.test.num_examples))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "What does the data of MNIST look like?\n",
      " type of 'trainimg' is <class 'numpy.ndarray'>\n",
      " type of 'trainlabel' is <class 'numpy.ndarray'>\n",
      " type of 'testimg' is <class 'numpy.ndarray'>\n",
      " type of 'testlabel' is <class 'numpy.ndarray'>\n",
      " shape of 'trainimg' is (55000, 784)\n",
      " shape of 'trainlabel' is (55000, 10)\n",
      " shape of 'testimg' is (10000, 784)\n",
      " shape of 'testlabel' is (10000, 10)\n"
     ]
    }
   ],
   "source": [
    "# What does the data of MNIST look like? \n",
    "print (\"What does the data of MNIST look like?\")\n",
    "trainimg   = mnist.train.images\n",
    "trainlabel = mnist.train.labels\n",
    "testimg    = mnist.test.images\n",
    "testlabel  = mnist.test.labels\n",
    "print\n",
    "print (\" type of 'trainimg' is %s\"    % (type(trainimg)))\n",
    "print (\" type of 'trainlabel' is %s\"  % (type(trainlabel)))\n",
    "print (\" type of 'testimg' is %s\"     % (type(testimg)))\n",
    "print (\" type of 'testlabel' is %s\"   % (type(testlabel)))\n",
    "print (\" shape of 'trainimg' is %s\"   % (trainimg.shape,))\n",
    "print (\" shape of 'trainlabel' is %s\" % (trainlabel.shape,))\n",
    "print (\" shape of 'testimg' is %s\"    % (testimg.shape,))\n",
    "print (\" shape of 'testlabel' is %s\"  % (testlabel.shape,))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "How does the training data look like?\n",
      "40849th Training Data Label is 8\n",
      "24739th Training Data Label is 4\n",
      "26520th Training Data Label is 5\n",
      "23951th Training Data Label is 5\n",
      "30072th Training Data Label is 8\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWEAAAFuCAYAAABQqakCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f990ac0a4a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWEAAAFuCAYAAABQqakCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJBVkCEtSQYawJBVkCEtSQYawJBX0/wH7gzi4K/W4jgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f99082e7978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWEAAAFuCAYAAABQqakCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGcKSVJEhLEkVGcKSVJEhLEkV/Q+btys9tp2RKQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f99082c0828>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWEAAAFuCAYAAABQqakCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f9908293d30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWEAAAFuCAYAAABQqakCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f9908258a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# How does the training data look like?\n",
    "print (\"How does the training data look like?\")\n",
    "nsample = 5\n",
    "randidx = np.random.randint(trainimg.shape[0], size=nsample)\n",
    "\n",
    "for i in randidx:\n",
    "    curr_img   = np.reshape(trainimg[i, :], (28, 28)) # 28 by 28 matrix \n",
    "    curr_label = np.argmax(trainlabel[i, :] ) # Label\n",
    "    plt.matshow(curr_img, cmap=plt.get_cmap('gray'))\n",
    "    plt.title(\"\" + str(i) + \"th Training Data \" \n",
    "              + \"Label is \" + str(curr_label))\n",
    "    print (\"\" + str(i) + \"th Training Data \" \n",
    "           + \"Label is \" + str(curr_label))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Batch Learning? \n",
      "type of 'batch_xs' is <class 'numpy.ndarray'>\n",
      "type of 'batch_ys' is <class 'numpy.ndarray'>\n",
      "shape of 'batch_xs' is (100, 784)\n",
      "shape of 'batch_ys' is (100, 10)\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# Batch Learning? \n",
    "print (\"Batch Learning? \")\n",
    "batch_size = 100\n",
    "batch_xs, batch_ys = mnist.train.next_batch(batch_size)\n",
    "print (\"type of 'batch_xs' is %s\" % (type(batch_xs)))\n",
    "print (\"type of 'batch_ys' is %s\" % (type(batch_ys)))\n",
    "print (\"shape of 'batch_xs' is %s\" % (batch_xs.shape,))\n",
    "print (\"shape of 'batch_ys' is %s\" % (batch_ys.shape,))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5. Get Random Batch with 'np.random.randint'\n",
      "type of 'batch_xs2' is <class 'numpy.ndarray'>\n",
      "type of 'batch_ys2' is <class 'numpy.ndarray'>\n",
      "shape of 'batch_xs2' is (100, 784)\n",
      "shape of 'batch_ys2' is (100, 10)\n"
     ]
    }
   ],
   "source": [
    "# Get Random Batch with 'np.random.randint'\n",
    "print (\"5. Get Random Batch with 'np.random.randint'\")\n",
    "randidx   = np.random.randint(trainimg.shape[0], size=batch_size)\n",
    "batch_xs2 = trainimg[randidx, :]\n",
    "batch_ys2 = trainlabel[randidx, :]\n",
    "print (\"type of 'batch_xs2' is %s\" % (type(batch_xs2)))\n",
    "print (\"type of 'batch_ys2' is %s\" % (type(batch_ys2)))\n",
    "print (\"shape of 'batch_xs2' is %s\" % (batch_xs2.shape,))\n",
    "print (\"shape of 'batch_ys2' is %s\" % (batch_ys2.shape,))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([14424, 18266, 51424, 45126, 41166, 15003, 54676, 30006, 45970,\n",
       "       52095,  3522, 44936, 21949, 44842, 41837, 40668,  9667, 41646,\n",
       "        3600, 15251, 26337, 27460, 52678,  6449, 48053, 20193, 40260,\n",
       "       29480, 20567, 45904, 30837, 45905, 45180,  7602,  2576, 23432,\n",
       "       44468, 37605, 25818, 37354,  7794, 13354, 10975, 50969, 16164,\n",
       "       16209, 47054, 53515, 52012, 28380,  4587, 14285,  6080, 50722,\n",
       "       36188, 17257,  1648, 27785,  4299, 36106, 54172,  4019, 49182,\n",
       "       51365, 48105,  5694, 18919,  4503,  7214, 43934, 41313, 24234,\n",
       "       35871,  5711, 38938, 24403, 20709, 49222,  4379, 48905, 33369,\n",
       "       27893, 11668, 14801,  7086, 42162, 12329, 22984, 42546, 37712,\n",
       "       54478, 49055, 10352, 26655, 49978, 53378, 53715, 43047, 20590,  8454])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "randidx"
   ]
  },
  {
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}