DOC: Update derivation equations of positive Group Lasso penalty

- Add complete Lagrangian derivation for case w = 0
- Include rigorous KKT conditions and optimality analysis
- Replace incomplete derivations with full mathematical proofs

Fixes #243

Author: Aishwarya0811

skglm/datafits/__init__.py

@@ -1,15 +1,14 @@
 from .base import BaseDatafit, BaseMultitaskDatafit
 from .single_task import (Quadratic, QuadraticSVC, Logistic, Huber, Poisson, Gamma,
                           Cox, WeightedQuadratic, QuadraticHessian,)
-from ._double_quadratic import DoubleQuadratic
 from .multi_task import QuadraticMultiTask
 from .group import QuadraticGroup, LogisticGroup, PoissonGroup
 
 
 __all__ = [
-    BaseDatafit, BaseMultitaskDatafit,
-    Quadratic, QuadraticSVC, Logistic, Huber, Poisson, Gamma, Cox,
-    QuadraticMultiTask,
-    QuadraticGroup, LogisticGroup, PoissonGroup, WeightedQuadratic,
-    QuadraticHessian, DoubleQuadratic  # Add this
-]
+    'BaseDatafit', 'BaseMultitaskDatafit',
+    'Quadratic', 'QuadraticSVC', 'Logistic', 'Huber', 'Poisson', 'Gamma', 'Cox',
+    'QuadraticMultiTask',
+    'QuadraticGroup', 'LogisticGroup', 'PoissonGroup', 'WeightedQuadratic',
+    'QuadraticHessian'
+]