I use this:
def data_norm(x):
mu = x.mean(1)
sigma = x.std(1)
return (x-mu)/sigma
but got error, how to deal with it
Do you mean
import numpy as np
def data_norm(x):
mu = x.mean(axis=1, keepdims=True)
sigma = x.std(axis=1, keepdims=True)
return (x-mu)/sigma
Got it! I has been confused by ‘keepdims’
In my understanding, keepdims
means that len(output.shape)
equals to len(input_shape)
.
Thx XD
I need to learn more about it