mathurinm

I think the error messages help a lot. Suggestion: add datafit/penalty and sovler from ``` AttributeError: Poisson is not compatible with AndersonCD. It must implement `get_lipschitz` and `gradient_scalar` Missing `get_lipschitz`....

> * I’m +1 with having `_solve` method and adding `run_checks` argument to `solve` method. > I feel it give us more freedom to standardize the behavior and make less...

Hi @v4gadkari, this has been inactive for a while so I'll close. Feel free to reopen if you want to continue working on this.

Hi Viren, No need to apologize. If you find time at some point we'll be happy to work with you.

We have them for some penalties : https://github.com/scikit-learn-contrib/skglm/blob/main/skglm/penalties/separable.py#L206 It regularly pops up because people are not aware of this and use an alpha_max that way too small. I don't know...

Thanks for the pointer @PascalCarrivain, I was not aware of this feature. It seems to help a lot (not for the first compilation, but for the subsequent calls), on a...

@PascalCarrivain do you know if this can help the first compilation too ?

Hi @hoodaty, to ease reviewing we try to address only one issue per PR ; your latest commit is for #231 rather than for #223, so can you revert its...

Yes much better like this, thank you! Can you add "Closes #231" in the first message of this PR, so that the corresponding is matched and closed automatically when this...

@hoodaty do you have any update on this?