Best practices when migrating from sklearn to torchnmf #21

yoyolicoris · 2022-01-05T03:11:38Z

yoyolicoris
Jan 5, 2022
Maintainer

fit

model = NMF(V.shape, rank=rank)
model.fit(V)

fit_transform

model = NMF(V.shape, rank=rank)
model.fit(V)
transformed_data = model.H.detach()      # the roles of H and W are swapped, which is different from sklearn

transform

model = NMF(V.shape, rank=rank)
model.fit(V)
model_new = NMF(W=model.W.detach(), H=(V_new.shape[0], model.H.shape[1]), trainable_W=False)
model_new.fit(V_new)

inverse_transform

V_hat = model(H)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

Best practices when migrating from sklearn to torchnmf #21

Uh oh!

{{title}}

Uh oh!

Replies: 0 comments

Select a reply

Uh oh!

Uh oh!

Best practices when migrating from sklearn to torchnmf #21

Uh oh!

yoyolicoris Jan 5, 2022 Maintainer

fit

fit_transform

transform

inverse_transform

Replies: 0 comments

yoyolicoris
Jan 5, 2022
Maintainer