1TensorFlow实现自编码器-1.8为自编码器计算图添加标量，图像等汇总节点

#1.8为自编码器计算图添加标量，图像等汇总节点

import numpy as np
import sklearn.preprocessing as prep
import tensorflow as tf
import matplotlib.pyplot as plt
from tensorflow.examples.tutorials.mnist import input_data
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'

#控制训练过程的参数
learning_rate = 0.01
training_epochs = 20
batch_size = 256
display_step = 1
examples_to_show = 10

#网络模型参数
n_hidden1_units = 256 #编码器第一隐藏层神经元数量（让编码器和解码器都有同样规模的隐藏层）
n_hidden2_units = 128 #编码器第二隐藏层神经元数量（让编码器和解码器都有同样规模的隐藏层）
n_input_units = 784 #输入层神经元数量MNIST data input(img shape :28*28)
n_output_units = n_input_units #解码器输出晨神经元数量必须等于输入数据的units数量

#对一个张量进行全面的汇总（均值，标准差，最大最小值，直方图）
#用于TensorBoard可视化
def variable_summaries(var):
    with tf.name_scope('summaries'):
        mean = tf.reduce_mean(var)
        tf.summary.scalar('mean',mean)
        with tf.name_scope('stddev'):
            stddev = tf.sqrt(tf.reduce_mean(tf.square(var-mean)))
            tf.summary.scalar('stddev',stddev)
            tf.summary.scalar('max', tf.reduce_max(var))
            tf.summary.scalar('min', tf.reduce_min(var))
            tf.summary.histogram('histogram', var)

#根据输入输出节点数量返回初始化好的指定名称的权重Variable
def WeightsVariable(n_in,n_out,name_str):
    return tf.Variable(tf.random_normal([n_in,n_out]),dtype=tf.float32,name=name_str)

#根据输出节点数量返回初始化好的指定名称的偏置Variable
def BiasesVariable(n_out,name_str):
    return tf.Variable(tf.random_normal([n_out]),dtype=tf.float32,name=name_str)

#构建编码器
def Encoder(x_origin,activate_func=tf.nn.sigmoid):
    #编码器第一隐藏层
    with tf.name_scope('Layer1'):
        weights = WeightsVariable(n_input_units,n_hidden1_units,'weights')
        biases = BiasesVariable(n_hidden1_units,'biases')
        x_code1 = activate_func(tf.add(tf.matmul(x_origin,weights),biases))

    # 编码器第二隐藏层
    with tf.name_scope('Layer2'):
        weights = WeightsVariable(n_hidden1_units, n_hidden2_units, 'weights')
        biases = BiasesVariable(n_hidden2_units, 'biases')
        x_code2 = activate_func(tf.add(tf.matmul(x_code1, weights), biases))

        variable_summaries(weights)
        variable_summaries(biases)

    return x_code2

#构建解码器
def Decoder(x_code,activate_func):
    #解码器第一隐藏层
    with tf.name_scope('Layer1'):
        weights = WeightsVariable(n_hidden2_units, n_hidden1_units, 'weights')
        biases = BiasesVariable(n_hidden1_units, 'biases')
        x_decode1 = activate_func(tf.add(tf.matmul(x_code, weights), biases))

    # 解码器第二隐藏层
    with tf.name_scope('Layer2'):
        weights = WeightsVariable(n_hidden1_units, n_output_units, 'weights')
        biases = BiasesVariable(n_output_units, 'biases')
        x_decode2 = activate_func(tf.add(tf.matmul(x_decode1, weights), biases))

        variable_summaries(weights)
        variable_summaries(biases)

    return x_decode2

#调用上面写的函数构造计算图
with tf.Graph().as_default():
    #计算图输入
    with tf.name_scope('X_origin'):
        X_Origin = tf.placeholder(tf.float32,[None,n_input_units])

    #构建编码器模型
    with tf.name_scope('Encoder'):
        X_code = Encoder(X_Origin,activate_func=tf.nn.sigmoid)

    # 构建解码器模型
    with tf.name_scope('Decoder'):
        X_decode = Decoder(X_code,activate_func=tf.nn.sigmoid)

    # 定义损失节点：重构数据与原始数据的误差平方和损失
    with tf.name_scope('Loss'):
        Loss = tf.reduce_mean(tf.pow(X_Origin - X_decode , 2))

    # 定义优化器，训练节点
    with tf.name_scope('Train'):
        Optimizer = tf.train.RMSPropOptimizer(learning_rate)
        Train = Optimizer.minimize(Loss)

    # 为计算图添加损失节点的标量汇总（scalar summary）
    with tf.name_scope('LossSummary'):
        tf.summary.scalar('loss',Loss)
        tf.summary.scalar('learning_rate', learning_rate)

    # 为计算图添加图像汇总（image summary）
    with tf.name_scope('ImageSummaries'):
        image_original = tf.reshape(X_Origin,[-1,28,28,1])
        image_reconstructed = tf.reshape(X_decode,[-1,28,28,1])
        tf.summary.image('image_original', image_original,9)
        tf.summary.image('image_reconstructed', image_reconstructed,9)

    #聚合所有汇总节点
    merged_summaries = tf.summary.merge_all()

    # 为所有变量添加初始化节点
    Init = tf.global_variables_initializer()

    print('把计算图写入事件文件，在Tesorboard里面查看')
    summary_write = tf.summary.FileWriter(logdir='logs',graph=tf.get_default_graph())
    summary_write.flush()

    # 读取数据集
    mnist = input_data.read_data_sets('MNIST_data/', one_hot=True)

    with tf.Session() as sess:
        sess.run(Init)

        total_batch = int(mnist.train.num_examples / batch_size)
        # 训练指定轮数，每一轮包含若干个批次
        for epoch in range(training_epochs):
            #每一轮（回合）都要把所有batch跑一遍
            for i in range(total_batch):
                batch_xs ,batch_ys = mnist.train.next_batch(batch_size)
                #运行优化器Train节点（backprop)和Loss节点（获取损失值）
                _,loss = sess.run([Train,Loss],feed_dict={X_Origin:batch_xs})
            #每一轮训完之后，输出logs
            if epoch % display_step == 0:
                print("Epoch:",'%04d' % (epoch + 1),"Loss=","{:.9f}".format(loss))
                #调用sess.run()方法运行汇总节点，更新事件文件
                summary_str = sess.run(merged_summaries,feed_dict={X_Origin:batch_xs})
                summary_write.add_summary(summary_str,epoch)
                summary_write.flush()

        #关闭summary_write
        summary_write.close()
        print("模型训练完毕")

        #把训练好的编码器-解码器模型用在测试集上，输出重建后的样本数据
        reconstructions = sess.run(X_decode,feed_dict={X_Origin:mnist.test.images[:examples_to_show]})
        #比较原始图像与重建后的图像
        f,a = plt.subplots(2,10,figsize=(10,2))
        for i in range(examples_to_show):
            a[0][i].imshow(np.reshape(mnist.test.images[i],(28,28)))
            a[1][i].imshow(np.reshape(reconstructions[i],(28,28)))
        f.show()
        plt.draw()
        plt.waitforbuttonpress()