tensorflow boolean_mask

发表于 | 更新于 | 分类于 | 评论数：

tf.boolean_mask

简单解释

用一个mask数组和输入的tensor做与操作，忽略为0的值。

api

定义在tensorflow/python/ops/array_ops.py

1
2
3
4
5
6
tf.boolean_mask(
    tensor, # 要处理的tensor
    mask, # 掩码，也需要是一个tensor
    name='boolean_mask', # 这个op的名字
    axis=None #
)

代码示例

代码地址

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
import tensorflow as tf

sess = tf.Session()
a = tf.Variable([1, 2, 3])
b = tf.Variable([2, 1.0, 4.0])
c = tf.Variable([2, 1.0, 0.0])
d = tf.Variable([2, 0.0, 4.0])
e = tf.Variable([0, 1.0, 4.0])
f = tf.Variable([0, 1.0, 0.0])
g = tf.Variable([0, 0.0, 0.0])

sess.run(tf.global_variables_initializer())
print("a: ", sess.run(a))
print("b: ", sess.run(b))
print("c: ", sess.run(c))
print("d: ", sess.run(d))
print("e: ", sess.run(e))
print("f: ", sess.run(f))
print("g: ", sess.run(g))
# c = tf.maximum(a, b)
a1 = tf.boolean_mask(a, b)
a2 = tf.boolean_mask(a, c)
a3 = tf.boolean_mask(a, d)
a4 = tf.boolean_mask(a, e)
a5 = tf.boolean_mask(a, f)
a6 = tf.boolean_mask(a, g)

print("tf.boolean(a, b):\n  ", sess.run(a1))
print("tf.boolean(a, c):\n  ", sess.run(a2))
print("tf.boolean(a, d):\n  ", sess.run(a3))
print("tf.boolean(a, e):\n  ", sess.run(a4))
print("tf.boolean(a, f):\n  ", sess.run(a5))
print("tf.boolean(a, g):\n  ", sess.run(a6))

输出如下：

a: [1 2 3]
b: [2. 1. 4.]
c: [2. 1. 0.]
d: [2. 0. 4.]
e: [0. 1. 4.]
f: [0. 1. 0.]
g: [0. 0. 0.]
tf.boolean(a, b):
[1 2 3]
tf.boolean(a, c):
[1 2]
tf.boolean(a, d):
[1 3]
tf.boolean(a, e):
[2 3]
tf.boolean(a, f):
[2]
tf.boolean(a, g):
[]

参考文献

1.http://landcareweb.com/questions/27920/zai-tensorflowzhong-ru-he-cong-pythonde-zhang-liang-zhong-huo-qu-fei-ling-zhi-ji-qi-suo-yin
2.https://www.tensorflow.org/api_docs/python/tf/boolean_mask