jakevdp / wpca

Weighted Principal Component Analysis (PCA) in Python

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Possible double counting of weights

christiangil opened this issue · comments

cWc = np.dot(cW, cW.T)

This appears to be calculating something more akin to cWWc. Should it instead be something like cWc = np.dot(cW, self.components_.T)?

I performed a simple transform - inverse_transform test and it suggests current implementation is much closer to naive PCA.