paddle 基础函数 clip 对输入Tensor每个元素的数值进行裁剪

官方文档：https://www.paddlepaddle.org.cn/documentation/docs/zh/api_cn/layers_cn/clip_cn.html

 

示例：

import paddle.fluid as fluid
import numpy as np

input = fluid.data(
    name='data', shape=[-1, 5], dtype='float32')

reward = fluid.layers.clip(x=input, min=0.5, max=0.9)


place = fluid.CPUPlace()
exe = fluid.Executor(place)
exe.run(fluid.default_startup_program())

np_x = np.random.random(size=(3, 5)).astype('float32')
print("np value: \n", np_x)

output = exe.run(feed={"data": np_x}, fetch_list = [reward])
print()
print("output: \n", output)

结果：

np value: 
 [[0.4650584  0.8003481  0.9856342  0.86992013 0.19400549]
 [0.18352304 0.7047317  0.7592452  0.83213526 0.57000315]
 [0.53873336 0.84840363 0.7514601  0.13381955 0.03195545]]

output: 
 [array([[0.5       , 0.8003481 , 0.9       , 0.86992013, 0.5       ],
       [0.5       , 0.7047317 , 0.7592452 , 0.83213526, 0.57000315],
       [0.53873336, 0.84840363, 0.7514601 , 0.5       , 0.5       ]],
      dtype=float32)]