01_TF_basics_and_linear_regression/mnist_data_introduction_tf.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MNIST 数据集 简介\n",
    "作为最经典的深度学习数据集，MNISIT包含65,000个灰度书写数字图片，尺寸均为`28x28`，其中55,000个用于训练，10,000个用于测试\n",
    "所有图片已归一化与中心化，像素值从0到1。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 使用\n",
    "我们使用TensorFlow的input_data函数进行数据准备与输入，它能帮助：\n",
    "* 自动下载数据\n",
    "* 将数据集load成numpy array的形式"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "nbpresent": {
     "id": "2d65d9e3-ece6-4ad1-8b37-234fd268f59a"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting ./mnist/train-images-idx3-ubyte.gz\n",
      "Extracting ./mnist/train-labels-idx1-ubyte.gz\n",
      "Extracting ./mnist/t10k-images-idx3-ubyte.gz\n",
      "Extracting ./mnist/t10k-labels-idx1-ubyte.gz\n"
     ]
    }
   ],
   "source": [
    "from __future__ import print_function\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "mnist = input_data.read_data_sets(\"./mnist/\", one_hot=True)\n",
    "\n",
    "# Load data\n",
    "x_train = mnist.train.images\n",
    "y_train = mnist.train.labels\n",
    "x_test = mnist.test.images\n",
    "y_test = mnist.test.labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 数据维度\n",
    "每张图片维度为784 (28x28x1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "nbpresent": {
     "id": "b754ce65-be9c-43e0-b903-1c43fd546d2e"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x_train:  (55000, 784)\n",
      "y_train:  (55000, 10)\n",
      "x_test:  (10000, 784)\n",
      "y_test:  (10000, 10)\n"
     ]
    }
   ],
   "source": [
    "print(\"x_train: \", x_train.shape)\n",
    "print(\"y_train: \", y_train.shape)\n",
    "print(\"x_test: \", x_test.shape)\n",
    "print(\"y_test: \", y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 数据可视化 \n",
    "使用matplotlib可视化MNIST:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "nbpresent": {
     "id": "bb5ff0c6-7c2c-416c-b4f0-387ace2502ba"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXEAAAEACAYAAABF+UbAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEOb/ALg2K/SnNEmAAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fad9c04c850>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "%matplotlib inline\n",
    "\n",
    "def plot_mnist(data, classes):\n",
    "    \n",
    "    for i in range(10):\n",
    "        idxs = (classes == i)\n",
    "        \n",
    "        # get 10 images for class i\n",
    "        images = data[idxs][0:10]\n",
    "            \n",
    "        for j in range(5):   \n",
    "            plt.subplot(5, 10, i + j*10 + 1)\n",
    "            plt.imshow(images[j].reshape(28, 28), cmap='gray')\n",
    "            # print a title only once for each class\n",
    "            if j == 0:\n",
    "                plt.title(i)\n",
    "            plt.axis('off')\n",
    "    plt.show()\n",
    "\n",
    "classes = np.argmax(y_train, 1)\n",
    "plot_mnist(x_train, classes)"
   ]
  }
 ],
 "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": "3.6.1"
  },
  "nbpresent": {
   "slides": {
    "16e84804-693b-4682-9a66-f4c709ffc6b2": {
     "id": "16e84804-693b-4682-9a66-f4c709ffc6b2",
     "prev": null,
     "regions": {
      "f8bef982-3cf0-4fb1-b40b-a3fcb2111dfd": {
       "attrs": {
        "height": 0.8,
        "width": 0.8,
        "x": 0.1,
        "y": 0.1
       },
       "content": {
        "cell": "3791f06d-5d42-441c-b534-0598fd5e106c",
        "part": "whole"
       },
       "id": "f8bef982-3cf0-4fb1-b40b-a3fcb2111dfd"
      }
     }
    },
    "2aa3c177-15ae-49b6-a806-e8ec49964d2d": {
     "id": "2aa3c177-15ae-49b6-a806-e8ec49964d2d",
     "prev": "7d29e2be-9b3e-4f52-93d1-ea7ed02b8104",
     "regions": {
      "14cb355f-770b-4c86-90a9-98271f5bce14": {
       "attrs": {
        "height": 0.8,
        "width": 0.8,
        "x": 0.1,
        "y": 0.1
       },
       "content": {
        "cell": "9786c91a-8f16-4daf-b2de-58bb29bd3c31",
        "part": "whole"
       },
       "id": "14cb355f-770b-4c86-90a9-98271f5bce14"
      }
     }
    },
    "48f2456b-2a85-4392-91e4-9a0f2af709d0": {
     "id": "48f2456b-2a85-4392-91e4-9a0f2af709d0",
     "prev": "16e84804-693b-4682-9a66-f4c709ffc6b2",
     "regions": {
      "b467c71e-ba2f-4259-a17d-c2e776f918d0": {
       "attrs": {
        "height": 0.8,
        "width": 0.8,
        "x": 0.1,
        "y": 0.1
       },
       "content": {
        "cell": "2d65d9e3-ece6-4ad1-8b37-234fd268f59a",
        "part": "whole"
       },
       "id": "b467c71e-ba2f-4259-a17d-c2e776f918d0"
      }
     }
    },
    "59b2b4d4-13f2-431a-bf86-33c91bb93af6": {
     "id": "59b2b4d4-13f2-431a-bf86-33c91bb93af6",
     "prev": "2aa3c177-15ae-49b6-a806-e8ec49964d2d",
     "regions": {
      "e002466d-bd43-407d-b5c9-f92631d2db94": {
       "attrs": {
        "height": 0.8,
        "width": 0.8,
        "x": 0.1,
        "y": 0.1
       },
       "content": {
        "cell": "bb5ff0c6-7c2c-416c-b4f0-387ace2502ba",
        "part": "whole"
       },
       "id": "e002466d-bd43-407d-b5c9-f92631d2db94"
      }
     }
    },
    "6aaaeb70-a7ab-4bc3-a234-5a95f3c4708f": {
     "id": "6aaaeb70-a7ab-4bc3-a234-5a95f3c4708f",
     "prev": "48f2456b-2a85-4392-91e4-9a0f2af709d0",
     "regions": {
      "03c87ea1-b2bc-4787-b6e7-43387e17f617": {
       "attrs": {
        "height": 0.8,
        "width": 0.8,
        "x": 0.1,
        "y": 0.1
       },
       "content": {
        "cell": "b754ce65-be9c-43e0-b903-1c43fd546d2e",
        "part": "whole"
       },
       "id": "03c87ea1-b2bc-4787-b6e7-43387e17f617"
      }
     }
    },
    "7d29e2be-9b3e-4f52-93d1-ea7ed02b8104": {
     "id": "7d29e2be-9b3e-4f52-93d1-ea7ed02b8104",
     "prev": "6aaaeb70-a7ab-4bc3-a234-5a95f3c4708f",
     "regions": {
      "6e76ee26-c96a-4604-9cae-de137f528c50": {
       "attrs": {
        "height": 0.8,
        "width": 0.8,
        "x": 0.1,
        "y": 0.1
       },
       "content": {
        "cell": "8ea67423-e641-48b3-9938-b2982eccf6f1",
        "part": "whole"
       },
       "id": "6e76ee26-c96a-4604-9cae-de137f528c50"
      }
     }
    },
    "e1e5cf35-fb45-413c-a7c5-8876ba42486d": {
     "id": "e1e5cf35-fb45-413c-a7c5-8876ba42486d",
     "prev": "59b2b4d4-13f2-431a-bf86-33c91bb93af6",
     "regions": {
      "1ff4d857-af53-46e4-b760-5fe65b6ae929": {
       "attrs": {
        "height": 0.8,
        "width": 0.8,
        "x": 0.1,
        "y": 0.1
       },
       "content": {
        "cell": "705e05ab-b637-43e7-a55d-411dc2bcaf2c",
        "part": "whole"
       },
       "id": "1ff4d857-af53-46e4-b760-5fe65b6ae929"
      }
     }
    },
    "e959e439-6fb7-4a60-92d4-81338251bd92": {
     "id": "e959e439-6fb7-4a60-92d4-81338251bd92",
     "prev": "e1e5cf35-fb45-413c-a7c5-8876ba42486d",
     "regions": {
      "a1e7cf2b-8ff1-4c04-82b8-4fbd3f9d7683": {
       "attrs": {
        "height": 0.8,
        "width": 0.8,
        "x": 0.1,
        "y": 0.1
       },
       "content": {
        "cell": "802b2fbd-81e9-4024-bbad-ca46fc9f7514",
        "part": "whole"
       },
       "id": "a1e7cf2b-8ff1-4c04-82b8-4fbd3f9d7683"
      }
     }
    }
   },
   "themes": {}
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}