02.03tensorflow学习笔记三：实例数据下载与读取.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "http://www.cnblogs.com/denny402/p/5852689.html\n",
    "\n",
    "Tensorflow从文件读取数据http://blog.csdn.net/zengxyuyu/article/details/53289906\n",
    "\n",
    "### tensorflow版本：V1.3\n",
    "\n",
    "\n",
    "### 一、mnist数据\n",
    "\n",
    "深度学习的入门实例，一般就是mnist手写数字分类识别，因此我们应该先下载这个数据集。\n",
    "\n",
    "tensorflow提供一个input_data.py文件，专门用于下载mnist数据，我们直接调用就可以了，代码如下：\n",
    "\n",
    "TensorFlow下MNIST数据集下载脚本input_data.py\n",
    "http://blog.csdn.net/lwplwf/article/details/54896959"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting data/train-images-idx3-ubyte.gz\n",
      "Extracting data/train-labels-idx1-ubyte.gz\n",
      "Extracting data/t10k-images-idx3-ubyte.gz\n",
      "Extracting data/t10k-labels-idx1-ubyte.gz\n"
     ]
    }
   ],
   "source": [
    "#下载用于训练和测试的mnist数据集的源码\n",
    "\n",
    "import input_data # 调用input_data\n",
    "mnist = input_data.read_data_sets('data/', one_hot=True)"
   ]
  },
  {
   "attachments": {
    "image.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "执行完成后，会在当前目录下新建一个文件夹MNIST_data, 下载的数据将放入这个文件夹内。下载的四个文件为：\n",
    "![image.png](attachment:image.png)\n",
    "input_data文件会调用一个maybe_download函数，确保数据下载成功。这个函数还会判断数据是否已经下载，如果已经下载好了，就不再重复下载。\n",
    "\n",
    "下载下来的数据集被分三个子集：5.5W行的训练数据集（mnist.train），5千行的验证数据集（mnist.validation)和1W行的测试数据集（mnist.test）。因为每张图片为28x28的黑白图片，所以每行为784维的向量。\n",
    "\n",
    "每个子集都由两部分组成：图片部分（images)和标签部分(labels), 我们可以用下面的代码来查看 ："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(55000, 784)\n",
      "(55000, 10)\n",
      "(5000, 784)\n",
      "(5000, 10)\n",
      "(10000, 784)\n",
      "(10000, 10)\n"
     ]
    }
   ],
   "source": [
    "print( mnist.train.images.shape)\n",
    "print(mnist.train.labels.shape) \n",
    "print(mnist.validation.images.shape) \n",
    "print(mnist.validation.labels.shape) \n",
    "print(mnist.test.images.shape) \n",
    "print(mnist.test.labels.shape) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 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.]]\n",
      "[[ 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.]]\n"
     ]
    }
   ],
   "source": [
    "# 如果想在spyder编辑器中查看具体数值，可以将这些数据提取为变量来查看，如：\n",
    "val_data=mnist.validation.images\n",
    "val_label=mnist.validation.labels\n",
    "\n",
    "print(val_data)\n",
    "print(val_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 二、CSV数据\n",
    "除了mnist手写字体图片数据，tf还提供了几个csv的数据供大家练习，存放路径为：\n",
    "`/home/xxx/anaconda3/lib/python3.5/site-packages/tensorflow/contrib/learn/python/learn/datasets/data/text_train.csv`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "module 'tensorflow.contrib.learn' has no attribute 'python'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-5-7c87ad52616a>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[1;32mimport\u001b[0m \u001b[0mtensorflow\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontrib\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlearn\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpython\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlearn\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdatasets\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbase\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0mbase\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[0miris_data\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0miris_label\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mbase\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mload_iris\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[0mhouse_data\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mhouse_label\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mbase\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mload_boston\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mAttributeError\u001b[0m: module 'tensorflow.contrib.learn' has no attribute 'python'"
     ]
    }
   ],
   "source": [
    "import tensorflow.contrib.learn.python.learn.datasets.base as base\n",
    "iris_data,iris_label=base.load_iris()\n",
    "house_data,house_label=base.load_boston()"
   ]
  },
  {
   "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": "3.5.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}