resurfemg.helper_functions.helper_functions module

Copyright 2022 Netherlands eScience Center and University of Twente Licensed under the Apache License, version 2.0. See LICENSE for details. This file contains general functions to support the functions in this repository

class resurfemg.helper_functions.helper_functions.Range(start, end)

Bases: RangeBase

Utility class for working with ranges (intervals). :ivar start: Start of the range :type start: ~number.Number :ivar end: End of the range :type end: ~number.Number

intersects(other)

Returns True if this range intersects other range. :param other: Another range to compare this one to :type other: ~resurfemg.helper_functions.Range :returns: True if this range intersects another range :rtype: bool

precedes(other)

Returns True if this range precedes other range. :param other: Another range to compare this one to :type other: ~resurfemg.helper_functions.Range :returns: True if this range strictly precedes another range :rtype: bool

to_slice()

Converts this range to a slice. :returns: A slice with its start set to this range’s start and end set to this range’s end :rtype: slice

resurfemg.helper_functions.helper_functions.count_decision_array(decision_array)

This is a function that, practically speaking, counts events on a time series array that has been reduced down to a binary (0, 1) output. It counts changes then divides by two. :param decision_array: Array. :type decisions_array: ~numpy.ndarray :returns: Number of events :rtype: float

resurfemg.helper_functions.helper_functions.delay_embedding(data, emb_dim, lag=1)

The following code is adapted from openly licensed code written by Christopher Schölzel in his package nolds (NOnLinear measures for Dynamical Systems). It performs a time-delay embedding of a time series :param data: array-like :type data: array :param emb_dim: the embedded dimension :type emb_dim: int :param lag: the lag between elements in the embedded vectors :type lag: int :returns: matrix_vectors :rtype: ~nd.array

resurfemg.helper_functions.helper_functions.distance_matrix(array_a, array_b)
Parameters:

  • array_a (array or list) – an array of same size as other parameter array

  • array_b (array or list) – an array of same size as other parameter array

Returns:

distances

Return type:

pd.DataFrame

resurfemg.helper_functions.helper_functions.intersections(left, right)

This function works over two arrays, left and right, and allows a picking based on intersections. It only takes ranges on the left that intersect ranges on the right. :param left: List of ranges :type left: List[Range] :param right: List of ranges :type right: List[Range] :returns: Ranges from the left that intersect ranges from the right. :rtype: List[Range]

resurfemg.helper_functions.helper_functions.merge(left, right)

Mergey function

resurfemg.helper_functions.helper_functions.preprocess(file_directory, our_chosen_leads, algorithm, processed, force=False)

This function is written to be called by the cli module. The cli module supports command line pre-processing. This function is currently written to accomodate Poly5 files types. It can be refactored later.

Parameters:

  • file_directory (str) – the directory with EMG files

  • processed (str) – the output directory

  • our_chosen_leads (list) – the leads selected for the pipeline to run over

resurfemg.helper_functions.helper_functions.ranges_of(array)

This function is made to work with Range class objects, such that is selects ranges and returns tuples of boundaries. :param my_own_array: array :type my_own_array: ~numpy.ndarray :return: range_return :rtype: tuple

resurfemg.helper_functions.helper_functions.raw_overlap_percent(signal1, signal2)

This function takes two binary 0 or 1 signal arrays and gives the percentage of overlap. :param signal1: Binary signal 1 :type signal1: ~numpy.ndarray :param rsignal2: Binary signal 2 :type rsignal2: ~numpy.ndarray :returns: Raw overlap percent :rtype: float

resurfemg.helper_functions.helper_functions.relative_levenshtein(signal1, signal2)

Here we take two arrays, and create an edit distance based on Levelshtien edit distance The distance is then normalized between 0 and one regardless of signal length

resurfemg.helper_functions.helper_functions.save_preprocessed(array, out_fname, force)

This function is written to be called by the cli module. It stores arrays in a directory.

resurfemg.helper_functions.helper_functions.scale_arrays(array, maximumn, minimumn)

This function will scale all arrays along the vertical axis to have an absolute maximum value of the maximum parameter :param array: Original signal array with any number iflayers :type array: ~numpy.ndarray :param maximumn: the absolute maximum below which the new array exists :type maximumn: float :param minimumn: the absolute maximum below which the new array exists :type minimumn: float :returns: reformed, a new array with absolute max of maximum :rtype: ~numpy.ndarray

resurfemg.helper_functions.helper_functions.slices_jump_slider(array_sample, slice_len, jump)

This function produces continuous sequential slices over an array of a certain length spaced out by a ‘jump’. The function yields, does not return these slices. :param array_sample: array containing the signal :type array_sample: ~numpy.ndarray :param slice_len: the length of window on the array :type slice_len: int :param jump: the amount by which the window is moved at iteration :type jump: int :returns: Actually yields, no return :rtype: ~numpy.ndarray

resurfemg.helper_functions.helper_functions.slices_slider(array_sample, slice_len)

This function produces continuous sequential slices over an array of a certain length. The inputs are the following - array_sample, the signal and slice_len - the window which you wish to slide with. The function yields, does not return these slices. :param array_sample: array containing the signal :type array_sample: ~numpy.ndarray :param slice_len: the length of window on the array :type slice_len: int :returns: Actually yields, no return :rtype: ~numpy.ndarray

resurfemg.helper_functions.helper_functions.zero_one_for_jumps_base(array, cut_off)

This function takes an array and makes it binary (0, 1) based on a cut-off value. :param array: An array :type array: ~numpy.ndarray :param cut_off: The number defining a cut-off line for binarization :type cut_off: float :returns: Binarized list that can be turned into array :rtype: list