my design matrix Φ has a column of one's to handle the intercept. so anytime I Z-score transform, I get NaNs after dividing by a zero variance. would be nice if transform had an option to handle this, ie. not transform when sigma=0.0.
eg. in scikit-learn's StandardScaler: "If a variance is zero, we can’t achieve unit variance, and the data is left as-is, giving a scaling factor of 1."
my design matrix
Φ
has a column of one's to handle the intercept. so anytime I Z-score transform, I getNaN
s after dividing by a zero variance. would be nice iftransform
had an option to handle this, ie. not transform when sigma=0.0.