skglm

skglm copied to clipboard

ENH/BUG? check_estimator has some very adversarial cases for SVC

1

To reproduce, run this in `test_estimators.py`: ```python estimator_name = "SVC" clf = clone(dict_estimators_ours[estimator_name]) clf.verbose = 2 clf.tol = 1e-6 # failure in float32 computation otherwise # if isinstance(clf, WeightedLasso): #...

mathurinm

[WIP] DOC - add example of scikit-learn ``GridSearchCV``

1

This verifies that we do support [scikit-learn ``GridSearchCV``](https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.GridSearchCV.html) and thereby adds an example to the documentation to illustrate that.

Badr-MOUFAD

Currently `penalty.value` on takes `w` in argument https://github.com/scikit-learn-contrib/skglm/blob/99d0a0636b678876e26e93f28bf1d87b5c0d0b97/skglm/penalties/separable.py#L77 Computation could be made faster by using `penalty.value(w, ws)` (one cannot simply call `penalty.value(w[ws])` because of the weighted l1 norm

QB3

BUG - Ill-conditioned residual matrix in AA can lead to wasted computations

1

As AA has been refactored in a class, we should make sure to catch a sneaky bug that can make AA fail silently. I've already stumbled upon cases where the...

PABannier

Probably calls for a non working set solver, or we can work something out with `is_penalized` (that could be called `yields_sparse_coefs` or something, ie if this is not true, we...

mathurinm

FEAT - Strata in COX model

3

## Description of the feature Can you support strata in COX model? It is included in lifetimes and statsmodels. **Additional context** https://lifelines.readthedocs.io/en/latest/fitters/regression/CoxPHFitter.html https://www.statsmodels.org/stable/generated/statsmodels.duration.hazard_regression.PHReg.predict.html#statsmodels.duration.hazard_regression.PHReg.predict

hermanhmchan

FEAT - Double quadratic datafit (quad-quad)

6

## Description of the feature In many applications there is a need to have an asymmetric loss that weights negative and positive residuals differently. A good function for this is...

tomaszkacprzak

Help with custom DataFit for Tweedie Regression

7

I am trying to create a custom datafit for Tweedie GLM. I followed the tutorial for Poisson but it errors out with a not so useful error statement. See below...

fsaforo1

## Context of the PR Follow up of #137, Make jit-compilation of datafits and penalties, and thereby datafit initialization, implicit within solvers. ## Contributions of the PR - Initialize datafit...

Badr-MOUFAD

TST add sparse group lasso testing against GSROPTIM

1

for the record, if one want s to compare against gsrpotim (by Eugene Ndiaye), here's a script showing that we get the same results: ```python from skglm.utils.data import make_correlated_data from...

mathurinm