[MNT] Remove REDCOMETs from testing exclusion list (aeon-toolkit#2630)

* remove excluded estimators

* redcomets fix

Author: MatthewMiddlehurst

aeon/classification/dictionary_based/_redcomets.py

@@ -255,8 +255,7 @@ def _build_univariate_ensemble(self, X, y):
         sfa_clfs = []
         for sfa in sfa_transforms:
             sfa.fit(X_smote, y_smote)
-            sfa_dics = sfa.transform_words(X_smote)
-            X_sfa = sfa_dics[:, 0, :]
+            X_sfa = sfa.transform_words(X_smote)[0]
 
             rf = RandomForestClassifier(
                 n_estimators=self.n_trees,
@@ -417,8 +416,7 @@ def _predict_proba_unvivariate(self, X) -> np.ndarray:
         pred_mat = np.zeros((X.shape[0], self.n_classes_))
 
         for sfa, (rf, weight) in zip(self.sfa_transforms, self.sfa_clfs):
-            sfa_dics = sfa.transform_words(X)
-            X_sfa = sfa_dics[:, 0, :]
+            X_sfa = sfa.transform_words(X)[0]
 
             rf_pred_mat = rf.predict_proba(X_sfa)
 

aeon/testing/testing_config.py

@@ -24,7 +24,6 @@
 
 # exclude estimators here for short term fixes
 EXCLUDE_ESTIMATORS = [
-    "REDCOMETS",
     "HydraTransformer",  # returns a pytorch Tensor
 ]