CODEWITHSUNDEEP
python-3.xmachine-learningscikit-learndata-sciencegridsearchcv
Anant Sinha
Anant Sinha

Reputation: 169

StandardScalar instead of normalize in parameter_grid

I'm trying to pass normalize in parameter search to pass it to GridSearchCV. I'm getting a warning that normalize is depreciated and that I should use StandardScalar instead. I can't just add StandardScalar to the pipeline because that would apply it all the time when I want it to be applied once and not applied the second time.

This is my code:

        'params': {
        'normalize': [True, False]
        }

How can I use StandardScalar here instead of normalize?

Upvotes: 0

Views: 109

Answers (1)

lejlot
lejlot

Reputation: 66775

Just incorporate StandardScaler into your pipeline, and control its parameters

class sklearn.preprocessing.StandardScaler(*, copy=True, 
                                           with_mean=True, with_std=True)

By setting with_mean=False and with_std=False you will get no normalisation, and setting both to true - you get whitening.

Quote from documentation:

The standard score of a sample x is calculated as:

z = (x - u) / s

where u is the mean of the training samples or zero if with_mean=False, and s is the standard deviation of the training samples or one if with_std=False.

Upvotes: 2

Related Questions