QB3 / sparse-ho

Fast hyperparameter settings for non-smooth estimators:
http://qb3.github.io/sparse-ho
BSD 3-Clause "New" or "Revised" License
38 stars 15 forks source link

ENH - Replace ``sklearn`` by ``celer`` #136

Closed Badr-MOUFAD closed 2 years ago

Badr-MOUFAD commented 2 years ago

This aims to use celer solver instead of sklearn.

closes #135

mathurinm commented 2 years ago

Thanks Badr

@Klopfe @QB3 was there a particular reason for which we used sklearn's Lasso instead of celer's ?

Klopfe commented 2 years ago

Thanks Badr

@Klopfe @QB3 was there a particular reason for which we used sklearn's Lasso instead of celer's ?

hmm... Not that I know of...

mathurinm commented 2 years ago

small failure here : https://github.com/QB3/sparse-ho/runs/6920007820?check_suite_focus=true#step:5:184 celer Logisticregression does not support multiclass so you need to keep sklearn's

I would use sklearn LogisticRegression everywhere in fact because it is quite fast (relying on liblinear etc)