Flag to skip centering in PCA

[eric-czech]

Hey @TomAugspurger, how opposed would you be to introducing small deviations from the scikit-learn interface to PCA like adding a flag in dask-ml.PCA to avoid requiring this centering?

dask-ml/dask_ml/decomposition/pca.py

Lines 270 to 271 in b94c587

	self.mean_ = X.mean(0)
	X -= self.mean_

That would certainly be a bad default, but if I know that is going to happen downstream from a scaling operation in a pipeline that already does the centering, I would like to have the option to avoid the extra operation. Is that reasonable or are you trying maintain something closer to perfect consistency?

[TomAugspurger]

Seems reasonable to me!