20171130_tensorflow_tf.Variable

tf.Variable

转自：TensorFlow图变量tf.Variable的用法解析

1.在TensorFlow的世界里，变量的定义和初始化是分开的，所有关于图变量的赋值和计算都要通过tf.Session的run来进行。想要将所有图变量进行集体初始化时应该使用tf.global_variables_initializer。
2.
tf.Variable

tf.Variable.init(initial_value, trainable=True, collections=None, validate_shape=True, name=None)

In [1]: import tensorflow as tf
In [2]: a = tf.Variable(3,name='a')
In [3]: a2 = a.assign(5)
In [4]: sess = tf.Session()
In [5]: sess.run(a.initializer)   #必须先定义a的值，否则会报错
In [6]: sess.run(a)
Out[6]: 13
In [7]: sess.run(a2)
Out[7]: 24

#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Date    : 2017-11-30 14:57:27
# @Author  : Lebhoryi@gmail.com
# @Link    : https://morvanzhou.github.io/tutorials/machine-learning/tensorflow/2-2-example2/
# @Version : Tensorflow 例子2

import tensorflow as tf
import numpy as np

#creat data
x_data = np.random.rand(100).astype(np.float32)
y_data = x_data*0.1 + 0.3

### creat tensorflow strucure start ###
Weights = tf.Variable(tf.random_uniform([1],-1.0,1.0))
biases = tf.Variable(tf.zeros([1]))

y = Weights*x_data + biases

loss = tf.reduce_mean(tf.square(y-y_data))
optimizer = tf.train.GradientDescentOptimizer(0.5)
train = optimizer.minimize(loss)

init = tf.initialize_all_variables()
### creat tensorflow strucure end ###

sess = tf.Session()
sess.run(init)    #Important

for step in range(201):
    sess.run(train)
    if step % 10 == 0:
        print(step,sess.run(Weights),sess.run(biases))

#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Date    : 2017-11-30 19:51:50
# @Author  : Lebhoryi@gmail.com
# @Link    : https://morvanzhou.github.io/tutorials/machine-learning/tensorflow/2-4-variable/
# @Version : Variable 变量

import tensorflow as tf

state = tf.Variable(0,name='counter')
print(state.name)
one = tf.constant(1)

new_value = tf.add(state,one)    #add
update = tf.assign(state,new_value)    #赋值，new_value赋值state

init = tf.initialize_all_variables()   #must have if define variable

with tf.Session() as sess:
    sess.run(init)
    for _ in range(3):
        sess.run(update)
        print(sess.run(state))