General updates

.github/workflows/lint.yml

@@ -25,7 +25,7 @@ jobs:
       - name: black
         uses: psf/black@stable
         with:
-          options: "--check --verbose"
+          options: "--check --verbose --diff --color"
           src: "./dtaianomaly"
 
       # Install dtaianomaly (not required for black)

.github/workflows/unit-tests.yml

@@ -31,7 +31,7 @@ jobs:
     - name: Install dtaianomaly
       run: |
         python -m pip install --upgrade pip
-        pip install .[tests]
+        pip install .[tests,tqdm]
         pip list
 
     - name: Test with pytest

docs/additional_information/changelog.rst

@@ -10,9 +10,22 @@ Added
 ^^^^^
 - Implemented ``KShapeAnomalyDetector`` anomaly detector.
 - Added arXiv citation to the documentation.
+- Added support for TOML configuration files in the ``Workflow``.
+- Added option to fit semi-supervised methods on test data in ``Workflow``.
+- Option to show progress bar when running a ``Workflow``.
+- Added optional feature names and time steps to ``DataSet``.
+- Added option for relative bounds when automatically computing the window size.
+- Added option to pass kwargs to the ``Workflow``.
+
 
 Changed
 ^^^^^^^
+- ``BestThresholdMetric`` now accepts an optional list of thresholds to use.
+- ``BestThresholdMetric`` stores all used thresholds and their respective scores.
+- ``BaseDetector`` by default checks the input variables, so this should no longer
+  be done in the implemented detectors.
+- Removed ``Evaluation.run()`` method, since it is not used.
+
 
 Fixed
 ^^^^^
@@ -21,6 +34,9 @@ Fixed
   'pyximport' within tslearn was not found, while this is not necessary for our
   codebase. Therefore, we have addid this dependency to the mock imports, which fixed
   the issue.
+- Ensured that ``interpret_additional_information()`` dynamically checks the possible
+  parameters of a ``Workflow``.
+- Parameter ``y`` in ``visualizations.plot_with_zoom()`` is now optional.
 
 [0.3.0] - 2025-01-31
 --------------------

docs/api/data.rst

@@ -9,20 +9,24 @@ Data module
 .. autoclass:: dtaianomaly.data.DataSet
     :members:
 
-Synthetic data
---------------
+Demonstration time series
+-------------------------
 
 .. autofunction:: dtaianomaly.data.demonstration_time_series
 
+.. autoclass:: dtaianomaly.data.DemonstrationTimeSeriesLoader
+
 .. image:: /../notebooks/Demonstration-time-series.svg
    :align: center
    :width: 100%
 
-.. autofunction:: dtaianomaly.data.make_sine_wave
-
 
 Loading data
 ------------
 
-.. autoclass:: dtaianomaly.data.UCRLoader
+.. autoclass:: dtaianomaly.data.PathDataLoader
+    :members:
+
 .. autofunction:: dtaianomaly.data.from_directory
+
+.. autoclass:: dtaianomaly.data.UCRLoader

docs/api/evaluation.rst

@@ -18,3 +18,4 @@ Evaluation module
 .. autoclass:: dtaianomaly.evaluation.AreaUnderPR
 .. autoclass:: dtaianomaly.evaluation.AreaUnderROC
 .. autoclass:: dtaianomaly.evaluation.BestThresholdMetric
+   :members: _compute

docs/api/visualization.rst

@@ -85,7 +85,7 @@ Visualization module
    >>> from dtaianomaly.visualization import plot_time_series_anomalies
    >>> from dtaianomaly.anomaly_detection import IsolationForest
    >>> from dtaianomaly.thresholding import FixedCutoff
-   >>> X, _ = demonstration_time_series()
+   >>> X, y = demonstration_time_series()
    >>> y_pred = IsolationForest(window_size=100).fit(X).predict_proba(X)
    >>> y_pred_binary = FixedCutoff(cutoff=0.9).threshold(y_pred)
    >>> fig = plot_time_series_anomalies(X, y, y_pred_binary, figsize=(10, 3))
@@ -100,5 +100,5 @@ Visualization module
    >>> from dtaianomaly.data import demonstration_time_series
    >>> from dtaianomaly.visualization import plot_with_zoom
    >>> X, y = demonstration_time_series()
-   >>> fig = plot_with_zoom(X, y, start_zoom=700, end_zoom=1200, figsize=(10, 3))
+   >>> fig = plot_with_zoom(X, y=y, start_zoom=700, end_zoom=1200, figsize=(10, 3))
    >>> fig.suptitle("Example of 'plot_with_zoom'")  # doctest: +SKIP

docs/getting_started/examples/anomaly_detection.rst

@@ -1,5 +1,20 @@
 :orphan:
 
+.. testsetup::
+
+   X, y = None, None
+   X_, y_ = None, None
+   preprocessor = None
+   detector = None
+   y_pred = None
+   pipeline = None
+   thresholding = None
+   y_pred_binary = None
+   precision = None
+   recall = None
+   f_1 = None
+   auc_roc, auc_pr = None, None
+
 Anomaly detection
 =================
 
@@ -26,12 +41,12 @@ demonstration time series. This time series can easily be loaded using the
 using the :py:func:`~dtaianomaly.visualization.plot_time_series_colored_by_score`
 method.
 
-.. code-block:: python
+.. doctest::
 
-    from dtaianomaly.data import demonstration_time_series
-    from dtaianomaly.visualization import plot_time_series_colored_by_score
-    X, y = demonstration_time_series()
-    plot_time_series_colored_by_score(X, y, figsize=(10, 2))
+    >>> from dtaianomaly.data import demonstration_time_series
+    >>> from dtaianomaly.visualization import plot_time_series_colored_by_score
+    >>> X, y = demonstration_time_series()
+    >>> plot_time_series_colored_by_score(X, y, figsize=(10, 2))  # doctest: +SKIP
 
 .. image:: /../notebooks/Demonstration-time-series.svg
    :align: center
@@ -44,19 +59,19 @@ Before detecting anomalies, we can preprocess the time series. In this case,
 we apply :py:class:`~dtaianomaly.preprocessing.MovingAverage` to remove some
 of the noise from the time series.
 
-.. code-block:: python
+.. doctest::
 
-    from dtaianomaly.preprocessing import MovingAverage
-    preprocessor = MovingAverage(window_size=10)
+    >>> from dtaianomaly.preprocessing import MovingAverage
+    >>> preprocessor = MovingAverage(window_size=10)
 
 In general, `any anomaly detector <https://dtaianomaly.readthedocs.io/en/stable/api/anomaly_detection.html>`_
 in ``dtaianomaly`` can be used to detect anomalies in this time series. Here, we use the
 :py:class:`~dtaianomaly.anomaly_detection.MatrixProfileDetector`
 
-.. code-block:: python
+.. doctest::
 
-    from dtaianomaly.anomaly_detection import MatrixProfileDetector
-    detector = MatrixProfileDetector(window_size=100)
+    >>> from dtaianomaly.anomaly_detection import MatrixProfileDetector
+    >>> detector = MatrixProfileDetector(window_size=100)
 
 
 Now that the components have been initialized, we can preprocess the time series and
@@ -66,10 +81,10 @@ does not process the ground truth, other preprocessors may change the ground tru
 For example, :py:class:`~dtaianomaly.preprocessing.SamplingRateUnderSampler` samples both
 the time series ``X`` and labels ``y``.
 
-.. code-block:: python
+.. doctest::
 
-    X_, y_ = preprocessor.fit_transform(X)
-    y_pred = detector.fit(X_).predict_proba(X_)
+    >>> X_, y_ = preprocessor.fit_transform(X)
+    >>> y_pred = detector.fit(X_).predict_proba(X_)
 
 Now we can plot the data along with the anomaly scores, and see that the predictions
 nicely align with the anomaly!
@@ -89,14 +104,14 @@ will automatically process the data before detecting anomalies. Note that it is
 possible to pass a list of preprocessors to apply multiple preprocessing steps before
 detecting anomalies.
 
-.. code-block:: python
+.. doctest::
 
-    from dtaianomaly.pipeline import Pipeline
-    pipeline = Pipeline(
-        preprocessor=preprocessor,
-        detector=detector
-    )
-    y_pred = pipeline.fit(X).predict_proba(X)
+    >>> from dtaianomaly.pipeline import Pipeline
+    >>> pipeline = Pipeline(
+    ...     preprocessor=preprocessor,
+    ...     detector=detector
+    ... )
+    >>> y_pred = pipeline.fit(X).predict_proba(X)
 
 Quantitative evaluation
 -----------------------
@@ -111,14 +126,14 @@ to 0 ("normal"). At this threshold, we see that all anomalous observations are d
 (recall=1.0), at the cost of some false positives near the borders of the ground truth
 anomaly (precision<1).
 
-.. code-block:: python
+.. doctest::
 
-    from dtaianomaly.thresholding import FixedCutoff
-    from dtaianomaly.evaluation import Precision, Recall
-    thresholding = FixedCutoff(0.85)
-    y_pred_binary = thresholding.threshold(y_pred)
-    precision = Precision().compute(y, y_pred_binary)
-    recall = Recall().compute(y, y_pred_binary)
+    >>> from dtaianomaly.thresholding import FixedCutoff
+    >>> from dtaianomaly.evaluation import Precision, Recall
+    >>> thresholding = FixedCutoff(0.85)
+    >>> y_pred_binary = thresholding.threshold(y_pred)
+    >>> precision = Precision().compute(y, y_pred_binary)
+    >>> recall = Recall().compute(y, y_pred_binary)
 
 
 Alternatively to manually applying a threshold to convert the continuous scores to
@@ -127,21 +142,35 @@ which will automatically apply a specified thresholding strategy before using a
 evaluation metric. Below, we use the same thresholding as above, but compute the
 :py:class:`~dtaianomaly.evaluation.FBeta` score with :math:`\\beta = 1`.
 
-.. code-block:: python
+.. doctest::
 
-    from dtaianomaly.evaluation import ThresholdMetric, FBeta
-    f_1 = ThresholdMetric(thresholding, FBeta(1.0)).compute(y, y_pred)
+    >>> from dtaianomaly.evaluation import ThresholdMetric, FBeta
+    >>> f_1 = ThresholdMetric(thresholding, FBeta(1.0)).compute(y, y_pred)
 
 Lastly, we also compute the :py:class:`~dtaianomaly.evaluation.AreaUnderROC` and
 :py:class:`~dtaianomaly.evaluation.AreaUnderPR`. Because these metrics create a
 curve for all possible thresholds, we can simply pass the predicted, continuous
 anomaly scores, as shown below.
 
-.. code-block:: python
-
-    from dtaianomaly.evaluation import AreaUnderROC, AreaUnderPR
-    auc_roc = AreaUnderROC().compute(y, y_pred)
-    auc_pr = AreaUnderPR().compute(y, y_pred)
+.. doctest::
+
+    >>> from dtaianomaly.evaluation import AreaUnderROC, AreaUnderPR
+    >>> auc_roc = AreaUnderROC().compute(y, y_pred)
+    >>> auc_pr = AreaUnderPR().compute(y, y_pred)
+
+.. doctest::
+   :hide:
+
+   >>> round(precision, 2)
+   0.64
+   >>> round(recall, 2)
+   1.0
+   >>> round(f_1, 2)
+   0.78
+   >>> round(auc_roc, 2)
+   0.99
+   >>> round(auc_pr, 2)
+   0.68
 
 The table below shows the computed performance metrics for this example.