nispace.stats.coloc.elasticnet

nispace.stats.coloc.elasticnet(x, y, cv=None, seed=None, **kwargs)[source]

Elastic-net regularized regression of x on y via sklearn.linear_model.ElasticNetCV.

Parameters:

  • x (np.ndarray, shape (n_obs, n_predictors)) – Does not handle NaN – sklearn errors on NaN input; pre-mask.

  • y (np.ndarray, shape (n_obs,)) –

  • cv (int, cross-validation generator, or None) – Passed to ElasticNetCV for selecting alpha/l1_ratio.

  • seed (int, optional) – Passed as ElasticNetCV’s random_state.

  • **kwargs – Forwarded to ElasticNetCV.

Returns:

out – "alpha"/"l1ratio" (selected regularization strength/mix), "r2", "beta" (shape (n_predictors,)).

Return type:

dict

Notes

Used by core/colocalize.py’s regularized-regression colocalization case. Unlike the other coloc.py methods (which exclude NaN case-wise, i.e. per predictor combination), the regularized methods (elasticnet/lasso/ridge) exclude NaN list-wise across all predictors at once before calling this function.