def z_score(x, axis=None):
if not isinstance(x, R.Tensor):
x = R.Tensor(x)
if axis is not None:
mean = R.mean(x, axis=axis)
std = R.std(x, axis=axis)
else:
mean = R.mean(x)
std = R.std(x)
return R.div(R.sub(x, mean), std)
def start_info(self, X):
"""
Calculate mean and standard deviation
"""
for feature in zip(*X):
yield {'std': R.std(feature), 'mean': R.mean(features)}
def standardize(x):
"""
Standardize an array
"""
if not isinstance(x, R.Tensor):
x = R.Tensor(x)
mean = R.mean(x)
std = R.std(x)
return R.div(R.sub(x, mean), std)
def stat_info(self, X):
"""
Calculate mean and standard deviation
"""
for feature in zip(*X):
feature = R.Tensor(list(feature), name = 'feature')
std = R.std(feature)
mean = R.mean(feature)
yield {
'std': std,
'mean': mean
}