tf.assign定义及 tf.assign_add定义

# tf.assign(
#     ref,  # ref: A mutable Tensor. Should be from a Variable node. May be uninitialized. 待赋值的变量
#     value,# value: A Tensor. Must have the same type as ref. The value to be assigned to the variable. 使用张量赋值
#     validate_shape=None,
#     use_locking=None,
#     name=None
#)

# return A Tensor that will hold the new value of 'ref' after the assignment has completed. 返回ref。
# https://www.tensorflow.org/api_docs/python/tf/assign
#assign  指派
#将 ref 赋值为 value，必须维度相同

# tf.assign_add(
#     ref,
#     value,
#     use_locking=None,
#     name=None
# )
#ref = ref + value  相加操作，赋值给ref
# return Same as "ref". Returned as a convenience for operations that want to use the new value after the variable has been updated.

import tensorflow as tf

x = tf.Variable(tf.constant(0.0), dtype=tf.float32)
with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    print("value x: ", sess.run(x))
    y = tf.assign(x, 20)
    print("value y； ", sess.run(y))
    print("tensor y: ", y)
    z = tf.assign_add(y, 11)
    print("value z: ",sess.run(z))

output:
    value x:  0.0
    value y；  20.0
    tensor y:  Tensor("Assign_6:0", shape=(), dtype=float32_ref)
    value z:  31.0