Update tests for BBI(2)

hidimstat/test/test_BBI.py

@@ -89,9 +89,9 @@ def _generate_data(n_samples=100, n_features=10, prob_type="regression",
     return X, y, grps, list_nominal
 
 
-def test_BBI_DNN():
+def test_BBI_condDNN():
 
-    X, y, grps, list_nominal = _generate_data()
+    X, y, _, list_nominal = _generate_data()
     # Compute importance with residuals
     bbi_res = BlockBasedImportance(
         estimator=None,
@@ -128,85 +128,26 @@ def test_BBI_DNN():
     results_samp = bbi_samp.compute_importance()
     assert len(results_samp["pval"]) == X.shape[1]
 
-# def test_BBI_samplingRF():
-
-#     X, y, grps, list_nominal = _generate_data()
-#     # Permutation Method
-#     bbi_perm = BlockBasedImportance(
-#         estimator=None,
-#         importance_estimator="Mod_RF",
-#         do_hyper=True,
-#         dict_hyper=None,
-#         conditional=False,
-#         group_stacking=False,
-#         prob_type="regression",
-#         k_fold=2,
-#         list_nominal=list_nominal,
-#         n_jobs=10,
-#         verbose=0,
-#         n_perm=100,
-#         )
-#     bbi_perm.fit(X, y)
-#     results_perm = bbi_perm.compute_importance()
-#     assert len(results_perm["pval"]) == X.shape[1]
-    # # Conditional Method
-    # bbi_cond = BlockBasedImportance(
-    #     estimator='RF',
-    #     importance_estimator="Mod_RF",
-    #     do_hyper=True,
-    #     dict_hyper=None,
-    #     conditional=True,
-    #     group_stacking=False,
-    #     prob_type="regression",
-    #     k_fold=k_fold,
-    #     list_nominal=list_nominal,
-    #     n_jobs=10,
-    #     verbose=0,
-    #     n_perm=100,
-    #     )
-    # bbi_cond.fit(X, y)
-    # results_cond = bbi_cond.compute_importance()
-    # pvals_cond = -np.log10(results_cond["pval"] + 1e-5)
-    # assert len(pvals_cond) == X.shape[1]
-
-
-# def test_BBI_residuals():
-#     # Permutation Method
-#     bbi_perm = BlockBasedImportance(
-#         estimator='RF',
-#         importance_estimator=None,
-#         do_hyper=True,
-#         dict_hyper=None,
-#         conditional=False,
-#         group_stacking=False,
-#         prob_type="regression",
-#         k_fold=k_fold,
-#         list_nominal=list_nominal,
-#         n_jobs=10,
-#         verbose=0,
-#         n_perm=100,
-#         )
-#     bbi_perm.fit(X, y)
-#     results_perm = bbi_perm.compute_importance()
-#     pvals_perm = -np.log10(results_perm["pval"] + 1e-10)
-#     assert len(pvals_perm) == X.shape[1]
-
-#     # Conditional Method
-#     bbi_cond = BlockBasedImportance(
-#         estimator='RF',
-#         importance_estimator=None,
-#         do_hyper=True,
-#         dict_hyper=None,
-#         conditional=True,
-#         group_stacking=False,
-#         prob_type="regression",
-#         k_fold=k_fold,
-#         list_nominal=list_nominal,
-#         n_jobs=10,
-#         verbose=0,
-#         n_perm=100,
-#         )
-#     bbi_cond.fit(X, y)
-#     results_cond = bbi_cond.compute_importance()
-#     pvals_cond = -np.log10(results_cond["pval"] + 1e-5)
-#     assert len(pvals_cond) == X.shape[1]
+
+def test_BBI_permDNN():
+
+    X, y, _, list_nominal = _generate_data()
+    # Compute importance with sampling RF
+    bbi_perm = BlockBasedImportance(
+        estimator=None,
+        importance_estimator="Mod_RF",
+        do_hyper=True,
+        dict_hyper=None,
+        conditional=False,
+        group_stacking=False,
+        prob_type="regression",
+        k_fold=2,
+        list_nominal=list_nominal,
+        n_jobs=10,
+        verbose=0,
+        n_perm=100,
+        )
+    bbi_perm.fit(X, y)
+    results_perm = bbi_perm.compute_importance()
+    assert len(results_perm["pval"]) == X.shape[1]
+