Theano 条件语句
theano
中提供了两种条件语句,ifelse
和 switch
,两者都是用于在符号变量上使用条件语句:
ifelse(condition, var1, var2)
- 如果
condition
为true
,返回var1
,否则返回var2
- 如果
switch(tensor, var1, var2)
- Elementwise
ifelse
操作,更一般化
- Elementwise
switch
会计算两个输出,而ifelse
只会根据给定的条件,计算相应的输出。
ifelse
需要从 theano.ifelse
中导入,而 switch
在 theano.tensor
模块中。
In [1]:
import theano, time
import theano.tensor as T
import numpy as np
from theano.ifelse import ifelse
Using gpu device 1: Tesla K10.G2.8GB (CNMeM is disabled)
假设我们有两个标量参数:\(a, b\),和两个矩阵 \(\mathbf{x, y}\),定义函数为:
\[ \mathbf z = f(a, b,\mathbf{x, y}) = \left\{ \begin{aligned} \mathbf x & ,\ a <= b\\="" \mathbf="" y="" &="" ,\="" a=""> b \end{aligned} \right. \]
定义变量:
In [2]:
a, b = T.scalars('a', 'b')
x, y = T.matrices('x', 'y')
用 ifelse
构造,小于等于用 T.lt()
,大于等于用 T.gt()
:
In [3]:
z_ifelse = ifelse(T.lt(a, b), x, y)
f_ifelse = theano.function([a, b, x, y], z_ifelse)
用 switch
构造:
In [4]:
z_switch = T.switch(T.lt(a, b), x, y)
f_switch = theano.function([a, b, x, y], z_switch)
测试数据:
In [5]:
val1 = 0.
val2 = 1.
big_mat1 = np.ones((10000, 1000), dtype=theano.config.floatX)
big_mat2 = np.ones((10000, 1000), dtype=theano.config.floatX)
比较两者的运行速度:
In [6]:
n_times = 10
tic = time.clock()
for i in xrange(n_times):
f_switch(val1, val2, big_mat1, big_mat2)
print 'time spent evaluating both values %f sec' % (time.clock() - tic)
tic = time.clock()
for i in xrange(n_times):
f_ifelse(val1, val2, big_mat1, big_mat2)
print 'time spent evaluating one value %f sec' % (time.clock() - tic)
time spent evaluating both values 0.638598 sec
time spent evaluating one value 0.461249 sec