Pull from origin

.github/workflows/pr_precommit.yml

@@ -1,9 +1,6 @@
 name: PR pre-commit
 
 on:
-  push:
-    branches:
-      - main
   pull_request_target:
     branches:
       - main
@@ -27,7 +24,7 @@ jobs:
       - name: Checkout
         uses: actions/checkout@v4
         with:
-          ref: ${{ github.head_ref }}
+          ref: ${{ github.event.pull_request.head.sha }}
           token: ${{ steps.app-token.outputs.token }}
 
       - name: Setup Python 3.10

aeon/anomaly_detection/_merlin.py

@@ -18,7 +18,7 @@ class MERLIN(BaseAnomalyDetector):
     """MERLIN anomaly detector.
 
     MERLIN is a discord discovery algorithm that uses a sliding window to find the
-    most anomalous subsequence in a time series. The algorithm is based on the
+    most anomalous subsequence in a time series [1]_. The algorithm is based on the
     Euclidean distance between subsequences of the time series.
 
     .. list-table:: Capabilities
@@ -43,6 +43,13 @@ class MERLIN(BaseAnomalyDetector):
         length. If no anomaly is found, the algorithm will move to the next length
         and reset ``r``.
 
+    References
+    ----------
+    .. [1] Nakamura, M. Imamura, R. Mercer and E. Keogh, "MERLIN: Parameter-Free
+           Discovery of Arbitrary Length Anomalies in Massive Time Series
+           Archives," 2020 IEEE International Conference on Data Mining (ICDM),
+           Sorrento, Italy, 2020, pp. 1190-1195.
+
     Examples
     --------
     >>> import numpy as np

aeon/datasets/_data_writers.py

@@ -15,8 +15,6 @@ def write_to_tsfile(
     Write metadata and data stored in aeon compatible data set to file.
     A description of the ts format is in examples/load_data.ipynb.
 
-    Note that this file is structured to still support the
-
     Parameters
     ----------
     X : np.ndarray (n_cases, n_channels, n_timepoints) or list of np.ndarray[

aeon/forecasting/compose/_pipeline.py

@@ -761,8 +761,7 @@ class TransformedTargetForecaster(_Pipeline):
     ...     ("detrender", Detrender()),
     ...     ("forecaster", NaiveForecaster(strategy="drift")),
     ... ])
-    >>> pipe.fit(y)
-    TransformedTargetForecaster(...)
+    >>> t = pipe.fit(y)
     >>> y_pred = pipe.predict(fh=[1,2,3])
 
         Example 2: without strings
@@ -1558,7 +1557,7 @@ class Permute(_DelegatedForecaster, BaseForecaster, _HeterogenousMetaEstimator):
     >>> from aeon.forecasting.base import ForecastingHorizon
     >>> from aeon.forecasting.compose import ForecastingPipeline, Permute
     >>> from aeon.forecasting.naive import NaiveForecaster
-    >>> from aeon.transformations.boxcox import BoxCoxTransformer
+    >>> from aeon.transformations.series._boxcox import BoxCoxTransformer
     >>> from aeon.transformations.exponent import ExponentTransformer
 
     Simple example: permute sequence of estimator in forecasting pipeline

aeon/forecasting/compose/tests/test_bagging.py

@@ -7,8 +7,8 @@
 from aeon.forecasting.compose import BaggingForecaster
 from aeon.forecasting.compose._bagging import _calculate_data_quantiles
 from aeon.forecasting.naive import NaiveForecaster
+from aeon.transformations._legacy._boxcox import _LogTransformer
 from aeon.transformations.bootstrap import STLBootstrapTransformer
-from aeon.transformations.boxcox import LogTransformer
 from aeon.utils.validation._dependencies import _check_soft_dependencies
 
 y = load_airline()
@@ -18,7 +18,7 @@
     not _check_soft_dependencies("statsmodels", severity="none"),
     reason="skip test if required soft dependency for BaggingForecaster not available",
 )
-@pytest.mark.parametrize("transformer", [LogTransformer, NaiveForecaster])
+@pytest.mark.parametrize("transformer", [_LogTransformer, NaiveForecaster])
 def test_bagging_forecaster_transformer_type_error(transformer):
     """Test that the right exception is raised for invalid transformer."""
     with pytest.raises(TypeError) as ex:
@@ -37,7 +37,7 @@ def test_bagging_forecaster_transformer_type_error(transformer):
     not _check_soft_dependencies("statsmodels", severity="none"),
     reason="skip test if required soft dependency not available",
 )
-@pytest.mark.parametrize("forecaster", [LogTransformer])
+@pytest.mark.parametrize("forecaster", [_LogTransformer])
 def test_bagging_forecaster_forecaster_type_error(forecaster):
     """Test that the right exception is raised for invalid forecaster."""
     with pytest.raises(TypeError) as ex:

aeon/forecasting/compose/tests/test_pipeline.py

@@ -94,7 +94,7 @@ def test_nesting_pipelines():
     from aeon.testing.utils.scenarios_forecasting import (
         ForecasterFitPredictUnivariateWithX,
     )
-    from aeon.transformations.boxcox import LogTransformer
+    from aeon.transformations._legacy._boxcox import _LogTransformer as LogTransformer
     from aeon.transformations.compose import OptionalPassthrough
     from aeon.transformations.detrend import Detrender
 

aeon/segmentation/_clasp.py

@@ -11,7 +11,7 @@
 import pandas as pd
 
 from aeon.segmentation.base import BaseSegmenter
-from aeon.transformations.clasp import ClaSPTransformer
+from aeon.transformations.series import ClaSPTransformer
 
 
 def find_dominant_window_sizes(X, offset=0.05):
@@ -66,7 +66,7 @@ def _is_trivial_match(candidate, change_points, n_timepoints, exclusion_radius=0
         List of change points chosen so far
     n_timepoints : int
         Total length
-    exclusion_radius : int
+    exclusion_radius : float
         Exclusion Radius for change points to be non-trivial matches
 
     Returns
@@ -97,7 +97,7 @@ def _segmentation(X, clasp, n_change_points=None, exclusion_radius=0.05):
         the transformer
     n_change_points : int
         the number of change points to find
-    exclusion_radius :
+    exclusion_radius : float
         the exclusion zone
 
     Returns
@@ -183,6 +183,8 @@ class ClaSPSegmenter(BaseSegmenter):
         The number of change points to search.
     exclusion_radius : int
         Exclusion Radius for change points to be non-trivial matches.
+    n_jobs : int, default=1
+        Number of jobs to be used.
 
     References
     ----------
@@ -206,10 +208,11 @@ class ClaSPSegmenter(BaseSegmenter):
 
     _tags = {"fit_is_empty": True}  # 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, exclusion_radius=0.05, n_jobs=1):
         self.period_length = int(period_length)
         self.n_cps = n_cps
         self.exclusion_radius = exclusion_radius
+        self.n_jobs = n_jobs
         super().__init__(axis=1, n_segments=n_cps + 1)
 
     def _predict(self, X: np.ndarray):
@@ -263,7 +266,9 @@ def get_fitted_params(self):
 
     def _run_clasp(self, X):
         clasp_transformer = ClaSPTransformer(
-            window_length=self.period_length, exclusion_radius=self.exclusion_radius
+            window_length=self.period_length,
+            exclusion_radius=self.exclusion_radius,
+            n_jobs=self.n_jobs,
         ).fit(X)
 
         self.found_cps, self.profiles, self.scores = _segmentation(

aeon/testing/mock_estimators/tests/test_mock_forecasters.py

@@ -22,7 +22,7 @@
 )
 from aeon.testing.utils.deep_equals import deep_equals
 from aeon.transformations.base import BaseTransformer
-from aeon.transformations.boxcox import BoxCoxTransformer
+from aeon.transformations.series._boxcox import BoxCoxTransformer
 
 y_series = load_airline().iloc[:-5]
 y_frame = y_series.to_frame()

aeon/testing/test_config.py

@@ -21,10 +21,10 @@
     import aeon.testing.utils._cicd_numba_caching  # noqa: F401
 
 EXCLUDE_ESTIMATORS = [
-    "BoxCoxTransformer",
+    "_BoxCoxTransformer",
     "TimeBinAggregate",
     "PartialAutoCorrelationTransformer",
-    "LogTransformer",
+    "_LogTransformer",
     "ReverseAugmenter",
     "RandomSamplesAugmenter",
     "SqrtTransformer",

aeon/transformations/boxcox.py → aeon/transformations/_legacy/_boxcox.py

@@ -1,10 +1,9 @@
 """Implemenents Box-Cox and Log Transformations."""
 
 __maintainer__ = []
-__all__ = ["BoxCoxTransformer", "LogTransformer"]
+__all__ = ["_BoxCoxTransformer", "_LogTransformer"]
 
 import numpy as np
-from deprecated.sphinx import deprecated
 from scipy import optimize, special, stats
 from scipy.special import boxcox, inv_boxcox
 from scipy.stats import boxcox_llf, distributions, variation
@@ -41,21 +40,14 @@ def _calc_uniform_order_statistic_medians(n):
     return v
 
 
-# TODO: remove in v0.10.0
-@deprecated(
-    version="0.9.0",
-    reason="BoxCoxTransformer will be removed in version 0.10.0 and replaced with a "
-    "BaseSeriesTransformer version in the transformations.series module.",
-    category=FutureWarning,
-)
-class BoxCoxTransformer(BaseTransformer):
+class _BoxCoxTransformer(BaseTransformer):
     r"""Box-Cox power transform.
 
     Box-Cox transformation is a power transformation that is used to
     make data more normally distributed and stabilize its variance based
     on the hyperparameter lambda. [1]_
 
-    The BoxCoxTransformer solves for the lambda parameter used in the Box-Cox
+    The _BoxCoxTransformer solves for the lambda parameter used in the Box-Cox
     transformation given `method`, the optimization approach, and input
     data provided to `fit`. The use of Guerrero's method for solving for lambda
     requires the seasonal periodicity, `sp` be provided. [2]_
@@ -88,7 +80,7 @@ class BoxCoxTransformer(BaseTransformer):
 
     See Also
     --------
-    LogTransformer :
+    _LogTransformer :
         Transformer input data using natural log. Can help normalize data and
         compress variance of the series.
     aeon.transformations.exponent.ExponentTransformer :
@@ -113,14 +105,6 @@ class BoxCoxTransformer(BaseTransformer):
        Journal ofthe Royal Statistical Society, Series B, 26, 211-252.
     .. [2] V.M. Guerrero, "Time-series analysis supported by Power
        Transformations ", Journal of Forecasting, vol. 12, pp. 37-48, 1993.
-
-    Examples
-    --------
-    >>> from aeon.transformations.boxcox import BoxCoxTransformer
-    >>> from aeon.datasets import load_airline
-    >>> y = load_airline()
-    >>> transformer = BoxCoxTransformer()
-    >>> y_hat = transformer.fit_transform(y)
     """
 
     _tags = {
@@ -214,14 +198,7 @@ def _inverse_transform(self, X, y=None):
         return Xt
 
 
-# TODO: remove in v0.10.0
-@deprecated(
-    version="0.9.0",
-    reason="LogTransformer will be removed in version 0.10.0 and replaced with a "
-    "BaseSeriesTransformer version in the transformations.series module.",
-    category=FutureWarning,
-)
-class LogTransformer(BaseTransformer):
+class _LogTransformer(BaseTransformer):
     """Natural logarithm transformation.
 
     The Natural logarithm transformation can be used to make the data more normally
@@ -238,7 +215,7 @@ class LogTransformer(BaseTransformer):
 
     See Also
     --------
-    BoxCoxTransformer :
+    _BoxCoxTransformer :
         Applies Box-Cox power transformation. Can help normalize data and
         compress variance of the series.
     aeon.transformations.exponent.ExponentTransformer :
@@ -251,14 +228,6 @@ class LogTransformer(BaseTransformer):
     Notes
     -----
     The log transformation is applied as :math:`ln(y)`.
-
-    Examples
-    --------
-    >>> from aeon.transformations.boxcox import LogTransformer
-    >>> from aeon.datasets import load_airline
-    >>> y = load_airline()
-    >>> transformer = LogTransformer()
-    >>> y_hat = transformer.fit_transform(y)
     """
 
     _tags = {

aeon/transformations/adapt.py

@@ -223,7 +223,7 @@ def get_test_params(cls, parameter_set="default"):
 # TODO: remove in v0.11.0
 @deprecated(
     version="0.10.0",
-    reason="PandasTransformAdaptorwill be removed in version 0.11.0.",
+    reason="PandasTransformAdaptor will be removed in version 0.11.0.",
     category=FutureWarning,
 )
 class PandasTransformAdaptor(BaseTransformer):

aeon/transformations/bootstrap/_mbb.py

@@ -7,12 +7,19 @@
 
 import numpy as np
 import pandas as pd
+from deprecated.sphinx import deprecated
 from sklearn.utils import check_random_state
 
+from aeon.transformations._legacy._boxcox import _BoxCoxTransformer
 from aeon.transformations.base import BaseTransformer
-from aeon.transformations.boxcox import BoxCoxTransformer
 
 
+# TODO: remove in v0.11.0
+@deprecated(
+    version="0.10.0",
+    reason="STLBootstrapTransformer will be removed in version 0.11.0.",
+    category=FutureWarning,
+)
 class STLBootstrapTransformer(BaseTransformer):
     """Creates a population of similar time series.
 
@@ -265,7 +272,7 @@ def _fit(self, X, y=None):
         )
 
         # fit boxcox to get lambda and transform X
-        self.box_cox_transformer_ = BoxCoxTransformer(
+        self.box_cox_transformer_ = _BoxCoxTransformer(
             sp=self.sp, bounds=self.lambda_bounds, method=self.lambda_method
         )
         self.box_cox_transformer_.fit(X)
@@ -399,6 +406,12 @@ def get_test_params(cls, parameter_set="default"):
         return params
 
 
+# TODO: remove in v0.11.0
+@deprecated(
+    version="0.10.0",
+    reason="MovingBlockBootstrapTransformer will be removed in version 0.11.0.",
+    category=FutureWarning,
+)
 class MovingBlockBootstrapTransformer(BaseTransformer):
     """Moving Block Bootstrapping method for synthetic time series generation.