eegdash.features.base_utils#

Basic Feature Extraction Utilities

This module defines basic utilities for feature extraction.

Functions

channel_names_to_indices(channels, ch_names)

Converts a list of channel names to channel indices in another list.

get_underlying_func(func)

Retrieve the original Python function from a potential wrapper.

Classes

BivariateIterator(pairs[, directed])

Pairs iterator for iterating pairs of channels.
class eegdash.features.base_utils.BivariateIterator(pairs: Iterable[Tuple[int, int]] | int, directed=False)[source]

Bases: object

Pairs iterator for iterating pairs of channels.

Parameters:

  • pairs (Iterable[tuple[int, int]] | int) – If an iterable of tuples is given, it represents the channel index pairs to iterate If an integer n is given, iterate through all unique pairs out of n channels.

  • directed (bool) – If an integer was given in pairs, this parameter controls whether all directed pairs should be iterated. Otherwise this parameter is ignored. Default is False.

get_pair_iterators() → tuple[ndarray, ndarray][source]

Get indices for pairs of channels.

Computes the upper triangle indices of an (n, n) matrix, excluding the diagonal.

Returns:

The row and column indices for the unique pairs.

Return type:

tuple of ndarray

eegdash.features.base_utils.channel_names_to_indices(channels: List[str], ch_names: List[str]) → List[int][source]

Converts a list of channel names to channel indices in another list.

Parameters:

  • channels (List[str]) – A list of channel names.

  • ch_names (List[str]) – A list of existing channel names to take indices from.

Returns:

A list of channel indices.

Return type:

List[int]

Raises:

ValueError – If the channel name was not found in the existing channels list.

eegdash.features.base_utils.get_underlying_func(func: Callable) → Callable[source]

Retrieve the original Python function from a potential wrapper.

Parameters:

func (callable) – The function to unwrap. Typically a raw function, a functools.partial object, or a Numba Dispatcher.

Returns:

The underlying Python function.

Return type:

callable

Notes

This utility specifically handles: * functools.partial: Returns the .func attribute. * numba.Dispatcher: Returns the .py_func attribute.