D0CRT Accept Scikit-learn models directly instead of parameter dictionaries

examples/plot_dcrt_example.py

@@ -58,7 +58,9 @@
     y = np.maximum(0.0, y)
 
     ## dcrt Lasso ##
-    d0crt_lasso = D0CRT(estimator=LassoCV(random_state=42, n_jobs=1), screening=False)
+    d0crt_lasso = D0CRT(
+        estimator=LassoCV(random_state=42, n_jobs=1), screening_threshold=None
+    )
     d0crt_lasso.fit_importance(X, y)
     pvals_lasso = d0crt_lasso.pvalues_
     results_list.append(
@@ -73,7 +75,7 @@
     ## dcrt Random Forest ##
     d0crt_random_forest = D0CRT(
         estimator=RandomForestRegressor(n_estimators=100, random_state=42, n_jobs=1),
-        screening=False,
+        screening_threshold=None,
     )
     d0crt_random_forest.fit_importance(X, y)
     pvals_forest = d0crt_random_forest.pvalues_

examples/plot_model_agnostic_importance.py

@@ -35,7 +35,7 @@
 from sklearn.model_selection import KFold
 from sklearn.svm import SVC
 
-from hidimstat import LOCO, D0CRT
+from hidimstat import D0CRT, LOCO
 
 #############################################################################
 # Generate data where classes are not linearly separable
@@ -70,11 +70,11 @@
 # test (:math:`H_0: X_j \perp\!\!\!\perp y | X_{-j}`) for each variable. However,
 # this test is based on a linear model (LogisticRegression) and fails to reject the null
 # in the presence of non-linear relationships.
-d0crt_linear = D0CRT(estimator=clone(linear_model), screening=False)
+d0crt_linear = D0CRT(estimator=clone(linear_model), screening_threshold=None)
 d0crt_linear.fit_importance(X, y)
 pval_dcrt_linear = d0crt_linear.pvalues_
 
-d0crt_non_linear = D0CRT(estimator=clone(non_linear_model), screening=False)
+d0crt_non_linear = D0CRT(estimator=clone(non_linear_model), screening_threshold=None)
 d0crt_non_linear.fit_importance(X, y)
 pval_dcrt_non_linear = d0crt_non_linear.pvalues_
 
@@ -162,8 +162,9 @@
 # As expected, when using linear models (d0CRT and LOCO-linear) that are misspecified,
 # the varibles are not selected. This highlights the benefit of using model-agnostic
 # methods such as LOCO, which allows for the use of models that are expressive enough
-# to explain the data.
-
+# to explain the data. While d0CRT can use any estimator, its distillation step
+# restricts it from capturing variable interactions. A more advanced distillation
+# approach, such as d1CRT, may help overcome this limitation.
 
 #################################################################################
 # References