bycycle-tools / bycycle

Cycle-by-cycle analysis of neural oscillations.

Home Page:https://bycycle-tools.github.io/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

DOC: Add a glossary

TomDonoghue opened this issue · comments

commented

In FOOOF and NDSP we have glossary pages on the documentation sites that explain the key terminology and words we use.

ByCycle doesn't yet have one - but I think we should add one. Once we settle on the new 1.0 API, let's revisit and collect / define the key term we need and use across the module. We can use this issue to collect key terms and definitions, etc.

commented

At some point in the past when I did some updates I started collected some terms for this, so dropping in here.

This list is incomplete, and any definition are copied randomly from places / may not be good.

sinusoidality
    xx
period
    A single cycle of a rhythm, defined as the time between two consecutive troughs (or peaks).
rise-decay symmetry
    The fraction of the period in the rise phase.
peak-trough symmetry
    The fraction of the period in the peak phase.
period
    The time between consecutive troughs (or peaks, if default is changed).
amplitude
    Average voltage changes of the rise and decay
amplitude consistency
    The amplitude consistency of a cycle is equal to the maximum relative difference between rises and 
    decay amplitudes across all pairs of adjacent rises and decays that include one of the flanks in the 
    cycle (3 pairs) (e.g. if a rise is 10mV and a decay is 7mV, then its amplitude consistency is 0.7)
period consistency
    Period consistency is equal to the maximum relative difference between all pairs of adjacent periods
    that include the cycle of interest (2 pairs: current + previous cycles and current + next cycles) (e.g. if the
    previous, current, and next cycles have periods 60ms, 100ms, and 120ms, respectively, then the period 
    consistency is min (60/100, 100/120) = 0.6.))
monotonicity
    The monotonicity is the fraction of samples that the instantaneous derivative (numpy.diff) is consistent with 
    the direction of the flank. (e.g. if in the rise, the instantaneous derivative is 90% positive, and in the decay, 
    the instantaneous derivative is 80% negative, then the monotonicity of the cycle would be 0.85 ((0.9+0.8)/2). 
    The rise and decay flanks of the cycle should be mostly monotonic. 

Completed in #70