Add fmt argument back in for backward compatibility

sktime/annotation/clasp.py

@@ -226,10 +226,11 @@ class ClaSPSegmentation(BaseSeriesAnnotator):
         "python_dependencies": "numba",
     }  # for unit test cases
 
-    def __init__(self, period_length=10, n_cps=1, exclusion_radius=0.05):
+    def __init__(self, period_length=10, n_cps=1, fmt="sparse", exclusion_radius=0.05):
         self.period_length = int(period_length)
         self.n_cps = n_cps
         self.exclusion_radius = exclusion_radius
+        self.fmt = fmt
         super().__init__()
 
     def _fit(self, X, Y=None):
@@ -261,7 +262,12 @@ def _predict(self, X):
         Y : pd.Series or an IntervalSeries
             Change points in sequence X.
         """
-        return self._predict_points(X)
+        change_points = self._predict_points(X)
+        if self.fmt == "dense":
+            return self.change_points_to_segments(
+                change_points, X.index.min(), X.index.max()
+            )
+        return change_points
 
     def _predict_points(self, X):
         """Predict changepoints on test/deployment data.

sktime/annotation/tests/test_clasp.py

@@ -46,13 +46,11 @@ def test_clasp_dense():
     ts, period_size, cps = load_gun_point_segmentation()
 
     # compute a ClaSP segmentation
-    clasp = ClaSPSegmentation(period_size, n_cps=1)
+    clasp = ClaSPSegmentation(period_size, n_cps=1, fmt="dense")
     clasp.fit(ts)
-    segmentation = clasp.transform(ts)
+    segmentation = clasp.predict(ts)
 
-    # Find the index of the first 1
-    cp_index = segmentation.index[segmentation.values == 1][0]
     _, profile = clasp.predict_scores(ts)
 
-    assert len(segmentation) == len(ts) and cp_index == 893
+    assert len(segmentation) == 2 and segmentation.index[0].right == 893
     assert np.argmax(profile) == 893