tensorflow常见函数——clip_by_value、numpy.random.RandomState、argmax

常见简单函数用法

- tf.clip_by_value()

tf.clip_by_value(V, min, max)
功能：截取在V，使V里面的各个元素在min和max之间

具体代码用法

import tensorflow as tf
v = tf.constant([[1.0, 2.0, 4.0],[4.0, 5.0, 6.0]])
result = tf.clip_by_value(v,2.5,4.5)
with tf.Session() as sess:
    print(sess.run(result))

# 输出结果如下
'''
[[ 2.5  2.5  4. ]
 [ 4.   4.5  4.5]]
'''

- numpy.random.RandomState()

numpy.random.RandomState(None|int)
功能：产生随机数种子

具体代码用法

import numpy
for i in [1,2,3,4]:
    a = numpy.random.RandomState(None)
    b = a.rand(1,2)
    print(i)
    print(b)

for i in [1,2,3,4]:
    a = numpy.random.RandomState(1)
    b = a.rand(1,2)
    print(i)
    print(b)

# 结果输出如下
'''
1 
[[ 0.63678388 0.53997544]] 
2 
[[ 0.88420701 0.11569489]] 
3 
[[ 0.55099434 0.9790908 ]] 
4 
[[ 0.6769419 0.42401973]]

1 
[[ 0.417022 0.72032449]] 
2 
[[ 0.417022 0.72032449]] 
3 
[[ 0.417022 0.72032449]] 
4 
[[ 0.417022 0.72032449]]
'''

- tf.argmax()

tf.argmax(input,dimension=None,name=None)
功能：返回沿dimension最大值的索引

具体代码用法

import tensorflow as tf
val = tf.constant([[1,2,3],[4,5,6]])
valShape = tf.shape(val)
t = tf.argmax(val,1)
with tf.Session() as sess:
    result, shape = sess.run([t,valShape])
    print(result)
    print("shape = %s" % shape)

#输出结果如下
'''
[2 2]
shape = [2 3]
'''

参考

https://blog.csdn.net/qq_41694195/article/details/79573494
https://blog.csdn.net/william_hehe/article/details/78635815
https://blog.csdn.net/m0_37041325/article/details/77159660

 

记录日期

2018/9/11 19:50 第一次