extract_glm_betas_in_mask: Function to extract subject-level coefficients from an FSL...

In UNCDEPENdLab/fmri.pipeline: Analysis of fMRI Data on HPC clusters

Function to extract subject-level coefficients from an FSL group analysis

Description

Function to extract subject-level coefficients from an FSL group analysis

Usage

extract_glm_betas_in_mask(
  gpa,
  mask_files,
  what = c("cope", "zstat"),
  out_dir = getwd(),
  extract_l1 = "all",
  extract_l2 = "all",
  extract_l3 = "all",
  ncores = NULL,
  scheduler = "local",
  aggregate = TRUE,
  aggFUN = mean,
  return_data = TRUE,
  write_data = TRUE
)

Arguments

gpa	a glm_pipeline_arguments object for which results of GLM analyses are already available
mask_files	vector of filenames for NIfTI-based atlases. The function will loop over each and extract coefficients from each unique value in each mask.
what	which statistics to extract from each parcel. Default is 'cope' (aka 'beta') and 'zstat'.
out_dir	the directory to which statistics are written as .csv.gz files.
extract_l1	a character vector of l1 models from which to extract run-level (level 1) statistics. If "none", l1 statistic extraction will be skipped. If "all", then all l1 models will be extracted.
extract_l2	a character vector of l2 models from which to extract subject-level (level 2) statistics. If "none", l2 statistic extraction will be skipped. If "all", then all l2 models will be extracted (you will still be prompted for which l1 models you want to extract from each l2 model).
extract_l3	a character vector of l3 models from which to extract group-level (level 3) statistics. If "none", l3 statistic extraction will be skipped. If "all", then all l3 models will be extracted (you will still be prompted for which l1 and l2 models you want to extract from each l3 model).
ncores	The number of cores to use for extracting statistics from lower-level imgages. If NULL, lookup value in gpa$parallel$extract_glm_betas_ncores. If this value is > 1 and scheduler = "local", then mclapply will be used locally to extract.
scheduler	The scheduler to use for extracting statistics. If "local", use lapply/mclapply within the current compute session. If 'slurm', use doFuture with multiple slurm jobs to extract.
aggregate	whether to take the average (or other central tendency measure) of voxels within a given mask value. This only pertains to integer-valued masks, not continuous ones.
aggFUN	the function used to aggregate statistics for voxels within a given mask value. Default is mean.
return_data	if TRUE, then extracted statistics will be returned as a list of data.frames with elements l1, l2, and l3.
write_data	if TRUE, then extracted statistics will be written as .csv.gz files to out_dir.

UNCDEPENdLab/fmri.pipeline documentation built on April 3, 2025, 3:21 p.m.