add mixins (#3009)

Author: TonyBagnall

aeon/forecasting/machine_learning/_setar.py

@@ -5,7 +5,7 @@
 
 import numpy as np
 
-from aeon.forecasting.base import BaseForecaster
+from aeon.forecasting.base import BaseForecaster, IterativeForecastingMixin
 
 
 def _lagmat_1d(y: np.ndarray, maxlag: int) -> np.ndarray:
@@ -38,7 +38,7 @@ def _ols_fit(X: np.ndarray, y: np.ndarray):
     return intercept, coefs, sse
 
 
-class SETAR(BaseForecaster):
+class SETAR(BaseForecaster, IterativeForecastingMixin):
     """
     Self-Exciting Threshold AutoRegressive (SETAR) forecaster with 2 regimes.
 

aeon/forecasting/machine_learning/_setarforest.py

@@ -4,15 +4,15 @@
 
 import numpy as np
 
-from aeon.forecasting.base import BaseForecaster
+from aeon.forecasting.base import BaseForecaster, IterativeForecastingMixin
 
 from ._setartree import SETARTree
 
 __maintainer__ = ["TinaJin0228"]
 __all__ = ["SETARForest"]
 
 
-class SETARForest(BaseForecaster):
+class SETARForest(BaseForecaster, IterativeForecastingMixin):
     """
     SETAR-Forest: Bagging + random subspace ensemble of SETAR-Tree base learners.
 

aeon/forecasting/machine_learning/_setartree.py

@@ -5,13 +5,13 @@
 from scipy.stats import f
 from sklearn.linear_model import LinearRegression
 
-from aeon.forecasting.base import BaseForecaster
+from aeon.forecasting.base import BaseForecaster, IterativeForecastingMixin
 
 __maintainer__ = ["TinaJin0228"]
 __all__ = ["SETARTree"]
 
 
-class SETARTree(BaseForecaster):
+class SETARTree(BaseForecaster, IterativeForecastingMixin):
     """
     SETAR-Tree: A tree algorithm for global time series forecasting.
 

aeon/forecasting/machine_learning/tests/test_setartree.py

@@ -25,20 +25,6 @@ def test_linear_series():
     assert np.isclose(pred, 21.0, atol=0.01), f"Prediction {pred} not close to 21.0"
 
 
-def test_iterative_forecast_linear():
-    """Test iterative forecasting on a linear series (1-D array of length H)."""
-    y = np.arange(1, 11.0)
-    f = SETARTree(lag=2)
-    preds = f.iterative_forecast(y, 3)
-    expected = np.arange(11.0, 14.0)
-    assert preds.shape == (
-        3,
-    ), f"iterative_forecast should return shape (H,), got {preds.shape}"
-    assert np.allclose(
-        preds, expected, atol=0.01
-    ), f"Predictions {preds} not close to {expected}"
-
-
 def test_fixed_lag():
     """Test the fixed_lag parameter."""
     y = np.arange(1, 21.0)