从锅炉工到AI专家(7)

#!/usr/bin/env python
# -*- coding=UTF-8 -*-

# 注意以上部分官方原本是没有的，其中直接执行是为了使用更方便，
# 不然需要用这种方式执行： python mnist_softmax.py
# 字体编码是因为需要添加中文注释，否则执行时会报错

# ==============================================================================
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================

"""A very simple MNIST classifier.

See extensive documentation at
https://www.tensorflow.org/get_started/mnist/beginners
"""
#下面三行是为了利用python3中的一些先进特征而添加，
#实际在本代码中并没有使用这些特征，因此技术上说是可以屏蔽的，
#但不管是否使用，每次引入避免因为习惯引起的低级错误也是推荐的，
#因为python大多当做脚运行，很可能执行到的时候才会报错。
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

#引入命令行参数解析库
import argparse
import sys

#所有的示例，如果使用官方推荐的方式安装tensorflow包，
#都应当已经已经安装在本地，比如mac上路径是：
#/usr/local/lib/python2.7/site-packages/tensorflow/examples/tutorials/mnist
#所以下面这个是直接从系统包中引用数据准备脚本
from tensorflow.examples.tutorials.mnist import input_data
import tensorflow as tf

#这个变量将用来保存命令行参数，首先清空
#在这里声明是为了成为全局变量，可以直接在函数中调用
FLAGS = None

#主程序，其实python作为重要的脚本语言，
#并不是必须定义主函数，但显然这种方式让程序更规范
def main(_):
  # Import data
  #下面的算法部分基本不需要重复解释，可以看以前版本的注释
  #FLAGS.data_dir是从命令行传递来的参数，
  #代表数据目录
  mnist = input_data.read_data_sets(FLAGS.data_dir)

  # Create the model
  x = tf.placeholder(tf.float32, [None, 784])
  W = tf.Variable(tf.zeros([784, 10]))
  b = tf.Variable(tf.zeros([10]))
  y = tf.matmul(x, W) + b

  # Define loss and optimizer
  y_ = tf.placeholder(tf.int64, [None])

  # The raw formulation of cross-entropy,
  #
  #   tf.reduce_mean(-tf.reduce_sum(y_ * tf.log(tf.nn.softmax(y)),
  #                                 reduction_indices=[1]))
  #
  # can be numerically unstable.
  #
  # So here we use tf.losses.sparse_softmax_cross_entropy on the raw
  # outputs of 'y', and then average across the batch.
  #交叉熵的计算，系统已经有内置的函数，
  #不需要自己计算了，对比原来的源码可以看一下，
  #上面英文也写出了自己计算的缺陷
  #注意参数:labels表示认为标注的正确值，logits是模型计算出的值
  cross_entropy = tf.losses.sparse_softmax_cross_entropy(labels=y_, logits=y)
  train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)

  sess = tf.InteractiveSession()
  tf.global_variables_initializer().run()
  # Train
  for _ in range(1000):
    batch_xs, batch_ys = mnist.train.next_batch(100)
    sess.run(train_step, feed_dict={x: batch_xs, y_: batch_ys})

  # Test trained model
  #注意这里的比较，y_没有跟以前版本一样做argmax，
  #因为y_的值mnist.test.labes已经是数字本身，
  #而不是原来代表10个分类的一维数组，
  #这个是由上面read_data_sets函数中one_hot=True参数决定的，
  #没有这个参数代表直接读数值数据，实际上在下载的数据包中就是数值
  correct_prediction = tf.equal(tf.argmax(y, 1), y_)
  accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
  print(sess.run(accuracy, feed_dict={x: mnist.test.images,
                                      y_: mnist.test.labels}))

if __name__ == '__main__':
  #使用引入的argparse库解析命令行参数
  parser = argparse.ArgumentParser()
  #本脚本只支持一个参数，用于指定测试数据集文件夹
  #默认的文件夹路径做了修改，指向本地已有的数据集，
  #免得每次启动时用参数重新指定
  parser.add_argument('--data_dir', type=str, default='MNIST_data',
                      help='Directory for storing input data')
  #解析的参数存入FLAGS,注意这是全局变量
  FLAGS, unparsed = parser.parse_known_args()
  #tf的架构定义，
  #包装了main主函数，熟悉c语言之类的用户看起来肯定更亲切
  tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)

	filesaver = tf.train.Saver()

	datafile='data/softmax_data.tfdata'
	filesaver.save(sess,datafile)

	datafile='data/softmax_data.tfdata'
	filesaver.restore(sess,datafile)

#!/usr/bin/env python
# -*- coding=UTF-8 -*-

# ==============================================================================
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================

"""A very simple MNIST classifier.

See extensive documentation at
https://www.tensorflow.org/get_started/mnist/beginners
"""

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import argparse
import sys,os

from tensorflow.examples.tutorials.mnist import input_data
import tensorflow as tf

FLAGS = None

 #数据文件保存的位置
datafile='data/softmax_data.tfdata'
 #增加一个函数用于判断数据文件是否已经存在
def datafile_exist():
	#tf写出的数据文件实际是分了几部分，
	#我们判断其中的索引文件是否存在
	#假设索引文件存在，就有完整的数据文件
    return os.path.exists(datafile+".index")

def main(_):    
    # Import data
    mnist = input_data.read_data_sets(FLAGS.data_dir)

    # Create the model
    x = tf.placeholder(tf.float32, [None, 784])
    W = tf.Variable(tf.zeros([784, 10]))
    b = tf.Variable(tf.zeros([10]))
    y = tf.matmul(x, W) + b

    # Define loss and optimizer
    y_ = tf.placeholder(tf.int64, [None])

    cross_entropy = tf.losses.sparse_softmax_cross_entropy(labels=y_, logits=y)
    train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)

    filesaver = tf.train.Saver()
    
    sess = tf.InteractiveSession()
    tf.global_variables_initializer().run()
    
    if FLAGS.train or (not datafile_exist()):
		#如果命令行指定训练模式，或者数据文件根本不存在则执行训练流程
        # Train
        for _ in range(1000):
            batch_xs, batch_ys = mnist.train.next_batch(100)
            sess.run(train_step, feed_dict={x: batch_xs, y_: batch_ys})
        print("Training finished, data write to file.")
		#训练结束，写出数据
        filesaver.save(sess,datafile)

        # Test trained model
        correct_prediction = tf.equal(tf.argmax(y, 1), y_)
        accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
        print(sess.run(accuracy, feed_dict={x: mnist.test.images,
                                      y_: mnist.test.labels}))
    if (not FLAGS.train) and datafile_exist():
		#如果已经存在数据文件，并且没有要求强行重新训练，则恢复文件
        print("Restore data...")
		#恢复数据
        filesaver.restore(sess,datafile)
        # Test trained model
        correct_prediction = tf.equal(tf.argmax(y, 1), y_)
        accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
        print(sess.run(accuracy, feed_dict={x: mnist.test.images,
                                      y_: mnist.test.labels}))
        
if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument('--data_dir', type=str, default='MNIST_data',
                      help='Directory for storing input data')
	#添加了一个参数，用于强制程序运行在训练数据状态
	#使用时，如果添加-t或者--train参数执行，则运行在训练模式
	#如果没有参数，则判断数据文件是否存在，已经存在直接进入测试模式
	#没有存在，则先训练，写出数据集，再恢复数据集做一次测试
    parser.add_argument('-t','--train', action='store_true',default=False,
                      help='Force do train')
    FLAGS, unparsed = parser.parse_known_args()
    tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)

config = tf.ConfigProto()
jit_level = tf.OptimizerOptions.ON_1
config.graph_options.optimizer_options.global_jit_level = jit_level
with tf.Session(config=config) as sess:
   ...

 #在Session建立后执行：
    train_writer = tf.summary.FileWriter('logs/train', sess.graph)
 #在Session关闭前执行：
    train_writer.close()

accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
 #第一个参数是指定显示时图表的名字，第二个参数是监控的值
tf.summary.scalar('accuracy', accuracy)

cross_entropy = tf.losses.sparse_softmax_cross_entropy(labels=y_, logits=y)
 #第一个参数是指定显示图表的名字，第二个参数是监控的值
tf.summary.histogram('cost', cross_entropy)

with tf.name_scope("pic_samples"):
    x = tf.placeholder(tf.float32, [None, 784])
with tf.name_scope("weight"):
    W = tf.Variable(tf.zeros([784, 10]))

x = tf.placeholder(tf.float32, [None, 784])
image_shaped_input = tf.reshape(x, [-1, 28, 28, 1])
tf.summary.image('input_images', image_shaped_input, 10)
...
W = tf.Variable(tf.zeros([784, 10]))
Wt=tf.transpose(W)	#因为权重值跟图片数据定义不同，要先转置一下，再转成图片
image_shaped_weight = tf.reshape(Wt, [-1, 28, 28, 1])
tf.summary.image('weight_images', image_shaped_weight, 10)

#!/usr/bin/env python
# -*- coding=UTF-8 -*-

# ==============================================================================
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================

"""A very simple MNIST classifier.

See extensive documentation at
https://www.tensorflow.org/get_started/mnist/beginners
"""

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import argparse
import sys,os

from tensorflow.examples.tutorials.mnist import input_data
import tensorflow as tf

FLAGS = None

datafile='data/softmax_data.tfdata'
def datafile_exist():
    return os.path.exists(datafile+".index")

def main(_):    
    # Import data
    mnist = input_data.read_data_sets(FLAGS.data_dir)

    # Create the model
    with tf.name_scope("pic_samples"):
        x = tf.placeholder(tf.float32, [None, 784])
        image_shaped_input = tf.reshape(x, [-1, 28, 28, 1])
        tf.summary.image('input_images', image_shaped_input, 10)
    with tf.name_scope("weight"):
        W = tf.Variable(tf.zeros([784, 10]))
        Wt=tf.transpose(W)
        image_shaped_weight = tf.reshape(Wt, [-1, 28, 28, 1])
        tf.summary.image('weight_images', image_shaped_weight, 10)
    b = tf.Variable(tf.zeros([10]))
    tf.summary.histogram('bias', b)
    y = tf.matmul(x, W) + b

    # Define loss and optimizer
    y_ = tf.placeholder(tf.int64, [None])

    cross_entropy = tf.losses.sparse_softmax_cross_entropy(labels=y_, logits=y)
    tf.summary.histogram('cost', cross_entropy)
    train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)

    filesaver = tf.train.Saver()
    
    
    sess = tf.InteractiveSession()
    tf.global_variables_initializer().run()
    
    correct_prediction = tf.equal(tf.argmax(y, 1), y_)
    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
    tf.summary.scalar('accuracy', accuracy)

    merged = tf.summary.merge_all()
    train_writer = tf.summary.FileWriter('logs/train', sess.graph)

    if FLAGS.train or (not datafile_exist()):
        # Train
        for i in range(1000):
            batch_xs, batch_ys = mnist.train.next_batch(100)
            sess.run(train_step, feed_dict={x: batch_xs, y_: batch_ys})
            summary,_ = sess.run([merged,train_step], feed_dict={x: batch_xs, y_: batch_ys})
            train_writer.add_summary(summary, i)
            # Test trained model
            if (i % 100 == 0):
                summary1,ac = sess.run([merged,accuracy], feed_dict={x: mnist.test.images,
                                              y_: mnist.test.labels})
                train_writer.add_summary(summary1,i)
        print("Training finished, data write to file.")
        
        filesaver.save(sess,datafile)
        print(ac)   

        train_writer.close()
    if (not FLAGS.train) and datafile_exist():
        print("Restore data...")
        filesaver.restore(sess,datafile)
        # Test trained model
        correct_prediction = tf.equal(tf.argmax(y, 1), y_)
        accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
        
        print(sess.run(accuracy, feed_dict={x: mnist.test.images,
                                      y_: mnist.test.labels}))

        
if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument('--data_dir', type=str, default='MNIST_data',
                      help='Directory for storing input data')
    parser.add_argument('-t','--train', action='store_true',default=False,
                      help='Force do train')
    FLAGS, unparsed = parser.parse_known_args()
    tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)