[MNT] speed up various non-suite tests, part 2 (#5071)

This PR speeds up various non-suite tests:

* `test_gscv_hierarchical`, by shortening the hierarchical data used in
the test
* tests for classifiers in `test_sklearn_compatibility`, by using a
smaller data set
* `test_docs_tsfresh_extractor`, by using a smaller data set

sktime/classification/tests/test_sklearn_compatability.py

@@ -17,8 +17,6 @@
     GridSearchCV,
     GroupKFold,
     GroupShuffleSplit,
-    HalvingGridSearchCV,
-    HalvingRandomSearchCV,
     KFold,
     LeaveOneOut,
     LeavePGroupsOut,
@@ -31,6 +29,10 @@
     TimeSeriesSplit,
     cross_val_score,
 )
+
+# removed due to too small data:
+# HalvingGridSearchCV,
+# HalvingRandomSearchCV,
 from sklearn.pipeline import Pipeline
 
 from sktime.classification.interval_based import CanonicalIntervalForest
@@ -39,29 +41,29 @@
 from sktime.utils.validation._dependencies import _check_soft_dependencies
 
 DATA_ARGS = [
-    {"return_numpy": True, "n_columns": 2},
-    {"return_numpy": False, "n_columns": 2},
+    {"return_numpy": True, "n_columns": 2, "n_instances": 7, "n_timepoints": 12},
+    {"return_numpy": False, "n_columns": 2, "n_instances": 7, "n_timepoints": 12},
 ]
 
 # StratifiedGroupKFold(n_splits=2), removed because it is not available in sklearn 0.24
 CROSS_VALIDATION_METHODS = [
     KFold(n_splits=2),
     RepeatedKFold(n_splits=2, n_repeats=2),
     LeaveOneOut(),
-    LeavePOut(p=5),
+    LeavePOut(p=2),
     ShuffleSplit(n_splits=2, test_size=0.25),
     StratifiedKFold(n_splits=2),
     StratifiedShuffleSplit(n_splits=2, test_size=0.25),
     GroupKFold(n_splits=2),
-    LeavePGroupsOut(n_groups=5),
+    LeavePGroupsOut(n_groups=2),
     GroupShuffleSplit(n_splits=2, test_size=0.25),
     TimeSeriesSplit(n_splits=2),
 ]
 PARAMETER_TUNING_METHODS = [
     GridSearchCV,
     RandomizedSearchCV,
-    HalvingGridSearchCV,
-    HalvingRandomSearchCV,
+    # HalvingGridSearchCV,
+    # HalvingRandomSearchCV,
 ]
 
 if _check_soft_dependencies("numba", severity="none"):
@@ -107,7 +109,7 @@ def test_sklearn_cross_validation(data_args):
 def test_sklearn_cross_validation_iterators(data_args, cross_validation_method):
     """Test if sklearn cross-validation iterators can handle sktime panel data."""
     fit_args = make_classification_problem(**data_args)
-    groups = [1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8, 9, 9, 10, 10]
+    groups = [1, 1, 2, 2, 3, 3, 4]
 
     for train, test in cross_validation_method.split(*fit_args, groups=groups):
         assert isinstance(train, np.ndarray) and isinstance(test, np.ndarray)

sktime/forecasting/model_selection/tests/test_tune.py

@@ -77,14 +77,14 @@ def _create_hierarchical_data():
     y = _make_hierarchical(
         random_state=TEST_RANDOM_SEEDS[0],
         hierarchy_levels=(2, 2),
-        min_timepoints=20,
-        max_timepoints=20,
+        min_timepoints=15,
+        max_timepoints=15,
     )
     X = _make_hierarchical(
         random_state=TEST_RANDOM_SEEDS[1],
         hierarchy_levels=(2, 2),
-        min_timepoints=20,
-        max_timepoints=20,
+        min_timepoints=15,
+        max_timepoints=15,
     )
     return y, X
 
@@ -103,7 +103,7 @@ def _create_hierarchical_data():
 }
 CVs = [
     *[SingleWindowSplitter(fh=fh) for fh in TEST_OOS_FHS],
-    SlidingWindowSplitter(fh=1, initial_window=15),
+    SlidingWindowSplitter(fh=1, initial_window=12, step_length=3),
 ]
 ERROR_SCORES = [np.nan, "raise", 1000]
 

sktime/transformations/panel/tests/test_tsfresh.py

@@ -39,12 +39,11 @@ def test_tsfresh_extractor(default_fc_parameters):
 )
 def test_docs_tsfresh_extractor():
     """Test whether doc example runs through."""
-    X, y = load_arrow_head(return_X_y=True)
-    X_train, X_test, y_train, y_test = train_test_split(X, y)
+    X, _ = load_arrow_head(return_X_y=True)[:3]
     ts_eff = TSFreshFeatureExtractor(
         default_fc_parameters="efficient", disable_progressbar=True
     )
-    ts_eff.fit_transform(X_train)
+    ts_eff.fit_transform(X)
     features_to_calc = [
         "dim_0__quantile__q_0.6",
         "dim_0__longest_strike_above_mean",
@@ -53,7 +52,7 @@ def test_docs_tsfresh_extractor():
     ts_custom = TSFreshFeatureExtractor(
         kind_to_fc_parameters=features_to_calc, disable_progressbar=True
     )
-    ts_custom.fit_transform(X_train)
+    ts_custom.fit_transform(X)
 
 
 @pytest.mark.skipif(