What’s new#
Feature: something that you couldn’t do before.
Efficiency: an existing feature now may not require as much computation or memory.
Enhancement: a miscellaneous minor improvement.
Fix: something that previously didn’t work as documentated
API: you will need to change your code to have the same effect in the future; or a feature will be removed in the future.
Dependencies#
Wildboar 1.3 requires Python 3.12+, numpy 1.24.1+, scipy 1.10.0+ and scikit-learn 1.8+.
Version 1.3.0#
In development
New and changed models#
Wildboar 1.3 introduces several new models.
explain.counterfactual.NativeGuideCounterfactuala baseline counterfactual explainer as proposed by Delaney et al. (2021).explain.counterfactual.PrototypeCounterfactuala novel prototype guided counterfactual explainer as proposed by Samsten (2026).explain.counterfactual.NiceCounterfactuala nearest-neighbor based counterfactual explainer.CastorRegressorfor regression using competing dilated shapelets.NearestNeighborsfor unsupervised nearest-neighbor search with time series distance metrics.Add a new module
wildboar.segmentfor time series segmentation. The module containsFlussSegmenterand the helper functiondetect_changepoints.Adds a new module
wildboar.dimension_selectionto sample a subset of the most important dimensions when considering multivariate time series. The new module contains six new selector algorithms:DistanceVarianceThreshold: remove dimensions where the pairwise distances has a variance below the threshold.SequentialDimensionSelector: remove dimensions sequentially by adding (or removing) dimensions greedily.SelectDimensionPercentile: only retain the specified fraction of dimensions with the highest score.SelectDimensionTopK: only retain the specified top k dimensions with the highest score.SelectDimensionSignificance: only retain the dimensions with a p-value below the specified alpha level.ECSSelector: Select time series dimensions based on the sum of distances between pairs of classes.
Add a new transform
QuantTransformas proposed by Dempster et.al (2024).Add a new transform
FftTransformfor discrete Fourier transform.Add support for scikit-learn style preprocessing transformers:
Standardize,MinMaxScale,MaxAbsScale,Truncate,Interpolate.Add a new module
lbfor distance lower bounds. The module contains four lower bounds.DtwKeoghLowerBound: lower bounds dynamic time warping.DtwKimLowerBound: lower bounds dynamic time warping.SaxLowerBound: lower bounds z-normalized Euclidean distance.PaaLowerBound: lower bounds z-normalized Euclidean distance.
Feature Native guide counterfactuals.
Feature Add support for searching multiple nearest neighbors in
NiceCounterfactualusing n_neighbors.API Remove the deprecated scoring and valid_scoring parameters in
counterfactuals. Use proximity instead of scoring and the returned validity mask to select successful counterfactuals.API Remove support for passing train_x and train_y through method_args in
counterfactuals. Use the train_x and train_y keyword parameters instead.API Remove the deprecated
target="auto"option in counterfactual explainers. Usetarget="predict"instead.
Enhancement Add support for interval depth, coverage probability and variability in
IntervalImportance.Fix Correct the per-timestep averaging in
IntervalImportance.Enhancement Add support for configurable frequency-spectrum analysis in
FrequencyImportance.Enhancement Add support for coverage probability, variability and top-k shapelet matching in
ShapeletImportance. Matching can be controlled using k_best, score_func and alpha.
Fix Increase the timeout upon first use. We also issue a better error message to alert the user to use
refresh_repositoriesto refresh the repositories.Fix Avoid division by zero in
standardize.Feature Add support for interpolating missing values. Support is implemented for both
load_dataset(withpreprocess="interpolate") and as a scikit-learn compatible transformerInterpolateFeature Add support for scikit-learn compatible transformers for standardize, minmax_scale, maxabs_scale and truncate.
Fix Correctly apply callable filters in
load_datasets.
API Rename matrix_profile to paired_matrix_profile and issue a deprecation warning in matrix_profile. The new function reverses the meaning of X and Y, i.e., annotate every subsequence in X with the closest match in Y (instead of the reverse).
Feature A new function
matrix_profilefor computing the matrix profile for every subsequence in all time series. By default it will raise a deprecation warning and delegate to paired_matrix_profile (until 1.4), after which the kind=”default” will be the default value. To keep the current behaviour set kind=”paired” and swap the order of X and Y or usepaired_matrix_profile.Feature Add support for array-like thresholds in
subsequence_match.Feature Add a new estimator
NearestNeighborsfor nearest-neighbor search.API Change the default value of dim in distance functions from
"warn"to"mean". By default, distances are averaged over the dimensions of multivariate time series.API Rename the boolean metric parameter in
MDSto metric_mds. Rename dissimilarity and dissimilarity_params to metric and metric_params, respectively.API Change the default value of normalized_stress in
MDSto"auto".
Enhancement Add parameter validation to all models in the
wildboar.linear_modelmodule.Feature Add
CastorRegressorfor regression using competing dilated shapelets.
Feature Add
FlussSegmenterfor matrix profile based time series segmentation.Feature Add
detect_changepointsfor changepoint detection using PELT or binary segmentation.API Deprecate
segmentwhich will be removed in 1.4. UseFlussSegmenterinstead.
Fix Correctly use n_jobs in
ElasticEnsembleClassifier.
Fix Reset the split criterion between candidate shapelets when fitting shapelet trees. Previously, the first shapelet was selected irrespective of the number of candidates, effectively treating n_shapelets as 1. The fix also affects shapelet forests.
Feature Add support for dictionary-based metric specifications in
ProximityTreeClassifier.API Remove the deprecated
metric="default"option inProximityTreeClassifier.Feature Add a new hyper-parameter impurity_equality_tolerance which controls how we treat impurities as equal. If the impurity of two shapelets are the same we consider the separation gap. By default the distance separation gap is disabled (
impurity_equality_tolerance=None) but it can be enabled by setting a (small) non negative float.Feature Add support for plotting decision trees using the
plot_treefunction.Feature Add support for different strategies when constructing shapelet trees. When strategy=”best”, we use the matrix profile to find the best shapelets per sample in the sizes determined by the shapelet_size parameter. We can tune the trade-off between accuracy and computational cost by setting the sample_size parameter. The tree defaults to strategy=”random” to retain backward compatibility. The default value will change to strategy=”best” in 1.4 and we issue a deprecation warning.
Feature Add support for specifying a coverage probability instead of a shapelet size range using coverage_probability and variability.
Enhancement Ensure that we consider the number of dimensions when computing the default value for n_shapelets.
API Remove the deprecated
intervals="sample"option in interval-based transformations. To sub-sample intervals, useintervals="fixed"and set sample_size to a float.API Remove the deprecated
n_intervals="log"option in interval-based transformations. Usen_intervals="log2"instead.API Deprecate
RandomShapeletTransformwhich will be removed in 1.4. UseShapeletTransformwith strategy=”random” to keep the current behavior after 1.4.Feature Add a new class
ShapeletTransformthat accept an additional parameter strategy which can be set to “random” or “best”. If set to “best” we use the matrix profile to find the best shapelets per sample to use in the transformation. The shapelet size is determined by the shapelet_size parameter.Feature Add a new option to
DerivativeTransformmethodto select the way in which the difference is computed. Supported options include: backward, central, and slope.API Deprecate the
estimateparameter inSAX. The same behavior can be had usingscale=True(default).Enhancement Add support for dyadic intervals in
IntervalTransform.Enhancement Add support for quantile summarizer in
IntervalTransform.Enhancement Add support for summarizer parameters in
IntervalTransform.Feature Add support for specifying a coverage probability instead of a minimum and maximum interval size in
IntervalTransformwhen using intervals=”random”.Feature Add support for the phase spectrum in
FftTransform.API Change the default value of n_jobs in fixed shapelet transformations from -1 to 1.
Fix Fix memory leaks in interval and attribute generation.
API Remove the deprecated iseos function. Use
is_end_of_seriesinstead.