Merge pull request #71 from kndbvortex/main

adding train_x variable in SETSTCF for feat mode

Author: JHoelli

TSInterpret/InterpretabilityModels/counterfactual/SETSCF.py

@@ -88,6 +88,7 @@ def __init__(
             self.ts_len = train_x.shape[1]
         elif mode == "feat":
             change = False
+            self.train_x = np.array(train_x)
             self.ts_len = train_x.shape[2]
         self.train_x_n = from_3d_numpy_to_nested(self.train_x)
         if backend == "PYT":