hfart

Functions for high-frequency analysis (HFA) of electophysiological signals for detection of short artifacts and quantitative assessment of the quality of signals.

This module written with Anastasia Kuzmina in 2014.

egegsignals.hfart.best_fragment(t, at, ln, percents=False, n=0)[source]

Selects the best signal’s fragment of a given length

Parameters:
  • t (numpy.ndarray) – Time sequence (sec)
  • at (numpy.ndarray) – Time sequence where artifacts are located (sec)
  • ln (float) – Length of a fragment (sec)
Returns:

tuple
egegsignals.hfart.hfa(t, x)[source]

HFA procedure

Parameters:
  • t (numpy.ndarray) – Time sequence (sec)
  • x (numpy.ndarray) – Sample sequence
Returns:

tuple
egegsignals.hfart.hfa_filter(t, x, l=60, cutoff=0.3)[source]

Filtrates signal for HFA using FIR filter

Parameters:
  • t (numpy.ndarray) – Time sequence (sec)
  • x (numpy.ndarray) – Sample sequence
  • l (float (sec)) – Length of operator
  • cutoff (float) – Bound (Hz)
Returns:

numpy.ndarray
egegsignals.hfart.longest_fragment(t, at, n=0)[source]

Selects longest fragment of signal with n artifacts

Parameters:
  • t (numpy.ndarray) – Time sequence (sec)
  • at (numpy.ndarray) – Time sequence where artifacts are located (sec)
  • n (numpy.ndarray) – Number of artifacts
Returns:

tuple
egegsignals.hfart.merge_artifacts(at1, at2)[source]

Merges artifacts locations.

Parameters:
  • at1 (numpy.ndarray) – Time sequence where artifacts from 1’st group are located (sec)
  • at2 (numpy.ndarray) – Time sequence where artifacts from 2’nd group are located (sec)
Returns:

numpy.ndarray
egegsignals.hfart.outliers(t, x)[source]

Finds outliers

Parameters:
  • t (numpy.ndarray) – Time sequence (sec)
  • x (numpy.ndarray) – Sample sequence
Returns:

numpy.ndarray
egegsignals.hfart.quality(t, at, n=0)[source]

Calculates the quality of signal

Parameters:
  • t (numpy.ndarray) – Time sequence (sec)
  • at (numpy.ndarray) – Time sequence where artifacts are located (sec)
  • n (integer) – Number of artifacts
Returns:

float
egegsignals.hfart.three_sigma(t, x, aver=600, step=30)[source]

Calculates the 3*sigma zone (normal distribution) with averaging on intervals.

Parameters:
  • t (numpy.ndarray) – Time sequence (sec)
  • x (numpy.ndarray) – Sample sequence
  • aver (float) – Length of averaging interval (sec)
  • step (float) – Step (sec)
Returns:

numpy.ndarray