Feature Package Overview#

The eegdash.features namespace re-exports feature extractors, decorators, and dataset utilities from the underlying submodules so callers can import the most common helpers from a single place. To avoid duplicated documentation in the API reference, the classes themselves are documented in their defining modules (see the links below). This page focuses on the high-level orchestration helpers that only live in the package __init__.

High-level discovery helpers#

eegdash.features.get_all_features() → list[tuple[str, Callable]][source]#

Get a list of all available feature functions.

Scans the eegdash.features.feature_bank module for functions that have been decorated to have a feature_kind attribute.

Returns:

A list of (name, function) tuples for all discovered features.

Return type:

list[tuple[str, callable]]

eegdash.features.get_feature_kind(feature: Callable) → MultivariateFeature[source]#

Get the ‘kind’ of a feature function.

The feature kind (e.g., univariate, bivariate) is typically attached by a decorator.

Parameters:

feature (callable) – The feature function to inspect.

Returns:

An instance of the feature kind (e.g., UnivariateFeature()).

Return type:

MultivariateFeature

eegdash.features.get_feature_predecessors(feature_or_extractor: Callable | None) → list[source]#

Get the dependency hierarchy for a feature or feature extractor.

This function recursively traverses the parent_extractor_type attribute of a feature or extractor to build a list representing its dependency lineage.

Parameters:

feature_or_extractor (callable) – The feature function or FeatureExtractor class to inspect.

Returns:

A nested list representing the dependency tree. For a simple linear chain, this will be a flat list from the specific feature up to the base FeatureExtractor. For multiple dependencies, it will contain tuples of sub-dependencies.

Return type:

list

eegdash.features.get_all_feature_extractors() → list[tuple[str, Callable]][source]#

Get a list of all available feature extractor callables.

A feature extractor is any callable in the feature bank that participates in the feature graph, meaning it declares a parent_extractor_type via FeaturePredecessor. This includes both preprocessors and the final feature functions.

Returns:

A list of (name, callable) tuples for all discovered feature extractors.

Return type:

list[tuple[str, callable]]

eegdash.features.get_all_feature_kinds() → list[tuple[str, type[MultivariateFeature]]][source]#

Get a list of all available feature ‘kind’ classes.

Scans the eegdash.features.extractors module for all classes that subclass MultivariateFeature.

Returns:

A list of (name, class) tuples for all discovered feature kinds.

Return type:

list[tuple[str, type[eegdash.features.extractors.MultivariateFeature]]]

Dataset and extraction utilities#

eegdash.features.extract_features(concat_dataset: BaseConcatDataset, features: FeatureExtractor | Dict[str, Callable] | List[Callable], *, batch_size: int = 512, n_jobs: int = 1) → FeaturesConcatDataset[source]#

Extract features from a concatenated dataset of windows.

This function applies a feature extractor to each WindowsDataset within a BaseConcatDataset in parallel and returns a FeaturesConcatDataset with the results.

Parameters:

  • concat_dataset (BaseConcatDataset) – A concatenated dataset of WindowsDataset or EEGWindowsDataset instances.

  • features (FeatureExtractor or dict or list) – The feature extractor(s) to apply. Can be a FeatureExtractor instance, a dictionary of named feature functions, or a list of feature functions.

  • batch_size (int, default 512) – The size of batches to use for feature extraction.

  • n_jobs (int, default 1) – The number of parallel jobs to use for extracting features from the datasets.

Returns:

A new concatenated dataset containing the extracted features.

Return type:

FeaturesConcatDataset

eegdash.features.fit_feature_extractors(concat_dataset: BaseConcatDataset, features: FeatureExtractor | Dict[str, Callable] | List[Callable], batch_size: int = 8192) → FeatureExtractor[source]#

Fit trainable feature extractors on a dataset.

If the provided feature extractor (or any of its sub-extractors) is trainable (i.e., subclasses TrainableFeature), this function iterates through the dataset to fit it.

Parameters:

  • concat_dataset (BaseConcatDataset) – The dataset to use for fitting the feature extractors.

  • features (FeatureExtractor or dict or list) – The feature extractor(s) to fit.

  • batch_size (int, default 8192) – The batch size to use when iterating through the dataset for fitting.

Returns:

The fitted feature extractor.

Return type:

FeatureExtractor

eegdash.features.load_features_concat_dataset(path: str | Path, ids_to_load: list[int] | None = None, n_jobs: int = 1) → eegdash.features.datasets.FeaturesConcatDataset[source]#

Load a stored FeaturesConcatDataset from a directory.

This function reconstructs a FeaturesConcatDataset by loading individual FeaturesDataset instances from subdirectories within the given path. It uses joblib for parallel loading.

Parameters:

  • path (str or pathlib.Path) – The path to the directory where the dataset was saved. This directory should contain subdirectories (e.g., “0”, “1”, “2”, …) for each individual dataset.

  • ids_to_load (list of int, optional) – A list of specific dataset IDs (subdirectory names) to load. If None, all subdirectories in the path will be loaded.

  • n_jobs (int, default 1) – The number of jobs to use for parallel loading. -1 means using all processors.

Returns:

A concatenated dataset containing the loaded FeaturesDataset instances.

Return type:

eegdash.features.datasets.FeaturesConcatDataset

See also#