Fix blending decoupled cv

paris-saclay-cds / ramp-workflow

Toolkit for building predictive workflows on top of pydata (pandas, scikit-learn, pytorch, keras, etc.).

https://paris-saclay-cds.github.io/ramp-docs/

BSD 3-Clause "New" or "Revised" License

68 stars 43 forks source link

Fix blending decoupled cv #282

Closed kegl closed 2 years ago

kegl commented 3 years ago

Porting the blending code to decoupled cv.

kegl commented 3 years ago

Yes, now that we can have nontrivial CVs, it would be good to have one among the starting kits (where folds are not pairs of indices but more complex objects).

kegl commented 2 years ago

It might be good to add a corresponding test, maybe not in this PR?

We are planning to have a transfer learning workflow/score/Predictions, Predictions will have a set of source and target sets, at that point we will have a unit test for the nontrivial CV.