001-Loading-Data-MNIST-v2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Scipy - Loading MNIST Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### download MNIST mat file from this link\n",
    "https://github.com/amplab/datascience-sp14/blob/master/lab7/mldata/mnist-original.mat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Alternative local file loader (due to mldata.org being down)\n",
    "from scipy.io import loadmat\n",
    "\n",
    "mnist_raw = loadmat(\"mldata/mnist-original.mat\")\n",
    "mnist = {\n",
    "    \"data\": mnist_raw[\"data\"].T,\n",
    "    \"target\": mnist_raw[\"label\"][0],\n",
    "    \"COL_NAMES\": [\"label\", \"data\"],\n",
    "    \"DESCR\": \"mldata.org dataset: mnist-original\",\n",
    "    }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'COL_NAMES': ['label', 'data'],\n",
       " 'DESCR': 'mldata.org dataset: mnist-original',\n",
       " 'data': array([[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]], dtype=uint8),\n",
       " 'target': array([0., 0., 0., ..., 9., 9., 9.])}"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 70K images, 28x28 pixels/image, each pixel = 0 (white) to 255 (black)\n",
    "mnist # a dict object"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((70000, 784), (70000,))"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# take a peek\n",
    "X,y = mnist['data'], mnist['target']\n",
    "\n",
    "X.shape, y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdb3da080d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "5.0"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# display example image\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "some_digit = X[36000]\n",
    "some_digit_image = some_digit.reshape(28, 28)\n",
    "\n",
    "plt.imshow(\n",
    "    some_digit_image, \n",
    "    cmap = plt.cm.binary,\n",
    "    interpolation=\"nearest\")\n",
    "\n",
    "plt.axis(\"off\")\n",
    "plt.show()\n",
    "\n",
    "# looks like a \"five\". What's the corresponding label?\n",
    "y[36000]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "def plot_digit(X, y):\n",
    "    digit_image = X.reshape(28, 28)\n",
    "\n",
    "    plt.imshow(\n",
    "        digit_image, \n",
    "        cmap = plt.cm.binary,\n",
    "        interpolation=\"nearest\")\n",
    "\n",
    "    plt.axis(\"off\")\n",
    "    plt.show()\n",
    "    \n",
    "    print(\"Label\", y)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdb3d99d0d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('Label', 2.0)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdb3d99d110>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('Label', 0.0)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdb3d7f5f10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('Label', 5.0)\n"
     ]
    }
   ],
   "source": [
    "plot_digit(X[15000], y[15000])\n",
    "plot_digit(X[2000], y[2000])\n",
    "plot_digit(X[36000], y[36000])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "((60000, 784), (10000, 784), (60000,), (10000,))\n"
     ]
    }
   ],
   "source": [
    "# dataset already split into training (1st 60K) & test (last 10K) images.\n",
    "# shuffle training set for cross-validation quality\n",
    "\n",
    "X_train, X_test, y_train, y_test = X[:60000], X[60000:], y[:60000], y[60000:]\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "shuffle_index = np.random.permutation(60000)\n",
    "X_train, y_train = X_train[shuffle_index], y_train[shuffle_index]\n",
    "\n",
    "print(X_train.shape, X_test.shape, y_train.shape, y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exploring MNIST Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdb34128ed0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(10, 10, figsize=(8, 8),\n",
    "                        subplot_kw={'xticks':[], 'yticks':[]},\n",
    "                        gridspec_kw=dict(hspace=0.1, wspace=0.1))\n",
    "\n",
    "for i, ax in enumerate(axes.flat):\n",
    "    ax.imshow(X_train[i].reshape((28,28)), cmap='binary', interpolation='nearest')\n",
    "    ax.text(0.05, 0.05, str(int(y_train[i])),\n",
    "           transform=ax.transAxes, color='blue')"
   ]
  }
 ],
 "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": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}