time_window_isc – webgazeR v0.8.2

Time-windowed Intersubject Correlation (ISC) per subject

Description

Computes ISC for each subject by correlating that subject’s time series against the mean of all other subjects within sliding windows, then Fisher z-transforms correlations, averages them per subject, and converts back to r.

Usage

time_window_isc(
  data,
  window_size = 10,
  step = 1,
  min_overlap = 3,
  method = c("pearson", "spearman"),
  return_per_window = FALSE
)

Arguments

data A numeric matrix or data.frame with rows = time points, columns = participants.

window_size Integer, window length in time points (default 10).

step Integer, step size between consecutive windows (default 1).

min_overlap Integer, minimum number of non-NA paired points within a window required to compute a correlation (default 3).

method Correlation method passed to [stats::cor()], usually "pearson" (default), or "spearman".

return_per_window Logical; if TRUE, also return a data.frame of per-window Fisher-z correlations per subject (default FALSE).

Value

If ‘return_per_window = FALSE’, a named numeric vector of ISC values (length = n subjects). If ‘TRUE’, a list with: - ‘isc’: named numeric vector of per-subject ISC (as above) - ‘per_window’: data.frame with columns ‘window_start’, ‘window_end’, ‘subject’, ‘z’