Compute an HRF smoothing kernel

This function computes a temporal similarity matrix from a series of hemodynamic response functions.

Usage

hrf_smoothing_kernel(
  len,
  TR = 2,
  form = onset ~ trialwise(),
  buffer_scans = 3L,
  normalise = TRUE,
  method = c("gram", "cosine")
)

Arguments

len

The number of scans.

TR

The repetition time (default is 2 seconds).

form

the trialwise formula expression, see examples.

buffer_scans

The number of scans to buffer before and after the event.

normalise

Whether to normalise the kernel.

method

The method to use for computing the kernel.

Value

a smoothing matrix

Examples

form <- onset ~ trialwise(basis="gaussian")
sk <- hrf_smoothing_kernel(100, TR=1.5, form)