Flag acquisition intervals whose rate or signal departs from the file

Flag acquisition intervals whose rate or signal departs from the file

Usage

detect_time_anomalies(
  time,
  mat,
  n_bins = 40L,
  mad_k = 5,
  min_events_per_bin = 50L
)

Arguments

time

numeric acquisition time per event, in the file's own units

mat

matrix of channel values on the analysis scale, one column per channel, same number of rows as time

n_bins

number of equal-width time intervals

mad_k

robust z above which a bin is flagged

min_events_per_bin

bins holding fewer events than this are not judged on their channel medians, because a median over a handful of events is noise; they are still judged on their rate, which is the point

Value

list(bins = data.frame, flagged = logical per event, reason)