edit what's new, and api

Author: floriankozikowski

doc/api.rst

@@ -30,6 +30,18 @@ Estimators
    WeightedLasso
 
 
+Covariance
+==========
+
+.. currentmodule:: skglm
+
+.. autosummary::
+   :toctree: generated/
+
+   GraphicalLasso
+   AdaptiveGraphicalLasso
+
+
 Penalties
 =========
 

doc/changes/0.5.rst

@@ -5,3 +5,6 @@ Version 0.5 (in progress)
 - Add support for fitting an intercept in :ref:`SqrtLasso <skglm.experimental.sqrt_lasso.SqrtLasso>` (PR: :gh:`298`)
 - Add experimental :ref:`QuantileHuber <skglm.experimental.quantile_huber.QuantileHuber>` and :ref:`SmoothQuantileRegressor <skglm.experimental.quantile_huber.SmoothQuantileRegressor>` for quantile regression, and an example script (PR: :gh:`312`).
 - Add :ref:`GeneralizedLinearEstimatorCV <skglm.cv.GeneralizedLinearEstimatorCV>` for cross-validation with automatic parameter selection for L1 and elastic-net penalties (PR: :gh:`299`)
+- Add :ref:`GraphicalLasso <skglm.covariance.GraphicalLasso>` for sparse inverse covariance estimation with both primal and dual algorithms
+- Add :ref:`AdaptiveGraphicalLasso <skglm.covariance.AdaptiveGraphicalLasso>` for non-convex penalty variations using iterative reweighting strategy
+- Add example demonstrating regularization paths for Graphical Lasso and its adaptive variations

skglm/__init__.py

@@ -5,3 +5,4 @@
     SparseLogisticRegression, GeneralizedLinearEstimator, CoxEstimator, GroupLasso,
 )
 from .cv import GeneralizedLinearEstimatorCV  # noqa F401
+from .covariance import GraphicalLasso, AdaptiveGraphicalLasso  # noqa F401