深度学习之卷积神经网络（六）

# -*- ecoding:utf-8 -*-

import tensorflow as tf

class model():
    def __init__(self):
        #创建图片节点（占位符模式）形状为784 列
        self.x_image = tf.placeholder(tf.float32, [None, 784])
        #创建期望输出值节点（占位符模式）形状为 10 列
        self.ylabel = tf.placeholder(tf.float32, [None, 10])
        #重新分配排列(-1为自动计算排列值) 图像输入为 （宽 28 高 28 通道 1  -1位图片数量）
        self.x = tf.reshape(self.x_image, [-1,28,28,1])
        #first layer
        #命名域
        with tf.name_scope('conv1'):
            #truncated_normal生成截断正态分布随机数  标准差为0.1
            #创建变量卷积层权值 filter为( 高 5 宽 5 通道数 1 卷积核个数 32)
            self.W_conv1 = tf.Variable(tf.truncated_normal([5,5,1,32], stddev=0.1))
            #创建常数偏移量
            self.b_conv1 = tf.Variable(tf.constant(0.1, shape=[32]))

            #Converlution
            #进行卷积 tf.nn.conv2d(input, filter, strides, padding, use_cudnn_on_gpu=None, name=None)
            #strides 步长（卷积时在图像每一维的步长，这是一个一维的向量，长度4）
            #padding：string类型的量，只能是"SAME","VALID"其中之一，这个值决定了不同的卷积方式
            #use_cudnn_on_gpu:bool类型，是否使用cudnn加速，默认为true
            self.h_conv1 = tf.nn.conv2d(self.x, self.W_conv1, strides=[1,1,1,1], padding='SAME') + self.b_conv1
            #输出值计算 
            # W = ( W - F + 2P)/S + 1
            # H = ( H - F + 2P)/S + 1
            #输出值为（高 24 宽 24 通道数 32）
            #激活函数
            self.h_convRelu1 = tf.nn.relu(self.h_conv1)
        
        #pooling
        with tf.name_scope('pool1'):
            #设定池化
            #tf.nn.max_pool(value, ksize, strides, padding, name=None)
            #ksize:池化窗口的大小，取一个四维向量，一般是[1, height, width, 1],不在batch和channels上做池化
            #strides：和卷积类似，窗口在每一个维度上滑动的步长，一般也是[1, stride,stride, 1]
            #padding：和卷积类似，可以取'VALID' 或者'SAME'
            self.h_pool1 = tf.nn.max_pool(self.h_convRelu1, ksize=[1,2,2,1], strides=[1,2,2,1], padding='SAME')
            #输出值计算
            #SAME模式：
            #W=W/S
            #H=H/S
            #VALID模式：
            # W =（W - F + 1）/S
            # H =（H - F + 1）/S
            #输出值为 （高 12 宽 12 通道数 32）

        #second layer
        with tf.name_scope('conv2'):
            #创建变量卷积层权值 filter为( 高 5 宽 5 通道数 32 卷积核个数 64) 标准差为0.1
            self.W_conv2 = tf.Variable(tf.truncated_normal([5,5,32,64], stddev=0.1))
            #创建常数偏移量
            self.b_conv2 = tf.Variable(tf.constant(0.1, shape=[64]))

            #Converlution
            #进行卷积
            self.h_conv2 = tf.nn.conv2d(self.h_pool1, self.W_conv2, strides=[1,1,1,1], padding='SAME') + self.b_conv2
            #输出为 （高 7 宽 7 通道数 64）
            #激活函数
            self.h_convRelu2 = tf.nn.relu(self.h_conv2)

        
        #pooling
        with tf.name_scope('pool2'):
            #设定池化
            self.h_pool2 = tf.nn.max_pool(self.h_convRelu2, ksize=[1,2,2,1], strides=[1,2,2,1], padding='SAME')

        #卷积部分最后输出为（高 7 宽 7 通道数 64）
        #全连接网络
        with tf.name_scope('fc1'):
            ##构建全连接的输入层权重 输入为 7*7*64个节点  隐藏层为 1024个节点
            self.W_fc1 = tf.Variable(tf.truncated_normal([7*7*64, 1024], stddev=0.1))
            ##构建全连接的输入层偏移量
            self.b_fc1 = tf.Variable(tf.constant(0.1,shape=[1024]))
            #重新设置输入数据的样式
            self.h_pool2_flat = tf.reshape(self.h_pool2, [-1,7*7*64])
            #进行矩阵乘法
            self.h_fc1 = tf.matmul(self.h_pool2_flat, self.W_fc1) + self.b_fc1
            #激活函数
            self.h_fcRelu1 = tf.nn.relu(self.h_fc1)
        
        #防止过拟合
        with tf.name_scope('dropout'):
            self.keep_prob = tf.placeholder(tf.float32)
            #tf.nn.dropout(x, keep_prob, noise_shape=None, seed=None,name=None) 
            #keep_prob: 设置神经元被选中的概率,在初始化时keep_prob是一个占位符,
            self.h_fcdrop1 = tf.nn.dropout(self.h_fcRelu1, self.keep_prob)
        
        #输出层
        with tf.name_scope('fc2'):
            #设定权重 输出层为10个节点
            self.W_fc2 = tf.Variable(tf.truncated_normal([1024, 10], stddev=0.1))
            #偏移量
            self.b_fc2 = tf.Variable(tf.constant(0.1, shape=[10]))
            #矩阵相乘 得到输出
            self.y_conv = tf.matmul(self.h_fcdrop1, self.W_fc2) + self.b_fc2
        
        #计算代价
        with tf.name_scope('loss'):
            #tf.nn.softmax_cross_entropy_with_logits(logits, labels, name=None)
            #logits : 就是神经网络最后一层的输出
            #labels : 实际的标签
            self.cross_entropy = tf.nn.softmax_cross_entropy_with_logits(labels=self.ylabel, logits=self.y_conv)
            #tf.reduce_mean(input_tensor, reduction_indices=None, keep_dims=False, name=None)
            #求取平均值
            self.cross_entropy = tf.reduce_mean(self.cross_entropy)     
        
        #梯度求解
        with tf.name_scope('adam_optimizer'):
            #Adam优化算法：是一个寻找全局最优点的优化算法。
            #寻找误差最小的点
            self.train_step = tf.train.AdamOptimizer(1e-4).minimize(self.cross_entropy)
        
        with tf.name_scope('accuracy'):
            #tf.argmax 返回向量中最大的那个索引号
            #tf.equal  比较两个矩阵或者向量是否相等
            #tf.cast   类型转换
            #tf.reduce_mean 求取平均值
            #计算误差率
            self.correct_prediction = tf.equal(tf.argmax(self.y_conv, 1), tf.argmax(self.ylabel,1))
            self.correct_prediction = tf.cast(self.correct_prediction, tf.float32)
            self.accuracy = tf.reduce_mean(self.correct_prediction)

"""Functions for downloading and reading MNIST data."""
from __future__ import print_function
import gzip
import os
import urllib

import numpy

SOURCE_URL = 'http://yann.lecun.com/exdb/mnist/'


def maybe_download(filename, work_directory):
  """Download the data from Yann's website, unless it's already here."""
  if not os.path.exists(work_directory):
    os.mkdir(work_directory)
  filepath = os.path.join(work_directory, filename)
  if not os.path.exists(filepath):
    filepath, _ = urllib.urlretrieve(SOURCE_URL + filename, filepath)
    statinfo = os.stat(filepath)
    print('Succesfully downloaded', filename, statinfo.st_size, 'bytes.')
  return filepath


def _read32(bytestream):
  dt = numpy.dtype(numpy.uint32).newbyteorder('>')
  return numpy.frombuffer(bytestream.read(4), dtype=dt)[0]


def extract_images(filename):
  """Extract the images into a 4D uint8 numpy array [index, y, x, depth]."""
  print('Extracting', filename)
  with gzip.open(filename) as bytestream:
    magic = _read32(bytestream)
    if magic != 2051:
      raise ValueError(
          'Invalid magic number %d in MNIST image file: %s' %
          (magic, filename))
    num_images = _read32(bytestream)
    rows = _read32(bytestream)
    cols = _read32(bytestream)

    #print num_images
    #print rows
    #print cols

    buf = bytestream.read(rows * cols * num_images)
    data = numpy.frombuffer(buf, dtype=numpy.uint8)
    data = data.reshape(num_images, rows, cols, 1)
    return data


def dense_to_one_hot(labels_dense, num_classes=10):
  """Convert class labels from scalars to one-hot vectors."""
  num_labels = labels_dense.shape[0]
  index_offset = numpy.arange(num_labels) * num_classes
  labels_one_hot = numpy.zeros((num_labels, num_classes))
  labels_one_hot.flat[index_offset + labels_dense.ravel()] = 1
  return labels_one_hot


def extract_labels(filename, one_hot=False):
  """Extract the labels into a 1D uint8 numpy array [index]."""
  print('Extracting', filename)
  with gzip.open(filename) as bytestream:
    magic = _read32(bytestream)
    if magic != 2049:
      raise ValueError(
          'Invalid magic number %d in MNIST label file: %s' %
          (magic, filename))
    num_items = _read32(bytestream)
    buf = bytestream.read(num_items)
    labels = numpy.frombuffer(buf, dtype=numpy.uint8)
    if one_hot:
      return dense_to_one_hot(labels)
    return labels


class DataSet(object):

  def __init__(self, images, labels, fake_data=False):
    if fake_data:
      self._num_examples = 10000
    else:
      assert images.shape[0] == labels.shape[0], (
          "images.shape: %s labels.shape: %s" % (images.shape,
                                                 labels.shape))
      self._num_examples = images.shape[0]

      # Convert shape from [num examples, rows, columns, depth]
      # to [num examples, rows*columns] (assuming depth == 1)
      assert images.shape[3] == 1
      images = images.reshape(images.shape[0],
                              images.shape[1] * images.shape[2])
      # Convert from [0, 255] -> [0.0, 1.0].
      images = images.astype(numpy.float32)
      images = numpy.multiply(images, 1.0 / 255.0)
    self._images = images
    self._labels = labels
    self._epochs_completed = 0
    self._index_in_epoch = 0

  @property
  def images(self):
    return self._images

  @property
  def labels(self):
    return self._labels

  @property
  def num_examples(self):
    return self._num_examples

  @property
  def epochs_completed(self):
    return self._epochs_completed

  def next_batch(self, batch_size, fake_data=False):
    """Return the next `batch_size` examples from this data set."""
    if fake_data:
      fake_image = [1.0 for _ in xrange(784)]
      fake_label = 0
      return [fake_image for _ in xrange(batch_size)], [
          fake_label for _ in xrange(batch_size)]
    start = self._index_in_epoch
    self._index_in_epoch += batch_size
    if self._index_in_epoch > self._num_examples:
      # Finished epoch
      self._epochs_completed += 1
      # Shuffle the data
      perm = numpy.arange(self._num_examples)
      numpy.random.shuffle(perm)
      self._images = self._images[perm]
      self._labels = self._labels[perm]
      # Start next epoch
      start = 0
      self._index_in_epoch = batch_size
      assert batch_size <= self._num_examples
    end = self._index_in_epoch
    return self._images[start:end], self._labels[start:end]


def read_data_sets(train_dir, fake_data=False, one_hot=False):
  class DataSets(object):
    pass
  data_sets = DataSets()

  if fake_data:
    data_sets.train = DataSet([], [], fake_data=True)
    data_sets.validation = DataSet([], [], fake_data=True)
    data_sets.test = DataSet([], [], fake_data=True)
    return data_sets

  TRAIN_IMAGES = 'train-images-idx3-ubyte.gz'
  TRAIN_LABELS = 'train-labels-idx1-ubyte.gz'
  TEST_IMAGES = 't10k-images-idx3-ubyte.gz'
  TEST_LABELS = 't10k-labels-idx1-ubyte.gz'
  VALIDATION_SIZE = 5000

  local_file = maybe_download(TRAIN_IMAGES, train_dir)
  train_images = extract_images(local_file)

  local_file = maybe_download(TRAIN_LABELS, train_dir)
  train_labels = extract_labels(local_file, one_hot=one_hot)

  local_file = maybe_download(TEST_IMAGES, train_dir)
  test_images = extract_images(local_file)

  local_file = maybe_download(TEST_LABELS, train_dir)
  test_labels = extract_labels(local_file, one_hot=one_hot)

  validation_images = train_images[:VALIDATION_SIZE]
  validation_labels = train_labels[:VALIDATION_SIZE]
  train_images = train_images[VALIDATION_SIZE:]
  train_labels = train_labels[VALIDATION_SIZE:]

  data_sets.train = DataSet(train_images, train_labels)
  data_sets.validation = DataSet(validation_images, validation_labels)
  data_sets.test = DataSet(test_images, test_labels)

  return data_sets

import tensorflow as tf
import input_data
from model import model

class TrainCnn():
    def __init__(self):
        #创建模型
        self.network = model()
        #创建图
        self.sess = tf.Session()
        #获取MNIST数据集
        self.data = input_data.read_data_sets('./data/', one_hot=True)
    
    def train(self):
        #每次读取图片大小
        batch_size = 100
        #开始计算
        self.sess.run(tf.global_variables_initializer())
        #计算20000次
        for i in range(20000):
            batch = self.data.train.next_batch(100)
            #输出错误率
            if i%1==0:
                realrate,cross = self.sess.run([self.network.accuracy, self.network.cross_entropy], feed_dict={self.network.x_image:batch[0], self.network.ylabel:batch[1], self.network.keep_prob:1.0})
                print ('step %d, training accuracy:%g ,cross_entropy:%g' %(i, realrate, cross))
            #训练网络
            self.sess.run(self.network.train_step, feed_dict={self.network.x_image:batch[0], self.network.ylabel:batch[1], self.network.keep_prob:0.5})

if __name__ == '__main__':
    app = TrainCnn()
    app.train()