fmri_lm_control() creates an fmri_lm_config object collecting all
options for robust and autoregressive modelling. It validates inputs and
applies defaults so downstream functions receive a single structured list.
Details
robust_options may contain:
type(FALSE, "huber", "bisquare")k_huberc_tukeymax_iterscale_scope("run", "global")reestimate_phi(logical)
ar_options may contain:
struct("iid", "ar1", "ar2", "arp")p(order for "arp")iter_gls(integer number of GLS iterations)global(logical, use global phi)voxelwise(logical)exact_first(logical)