Skip to contents

Configuration for fmri_lm fitting

Source: R/fmri_lm_config.R
fmri_lm_control.Rd

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.

Usage

fmri_lm_control(
  robust_options = list(),
  ar_options = list(),
  volume_weights_options = list(),
  soft_subspace_options = list()
)

Arguments

robust_options

list of robust fitting options. See Details.

ar_options

list of autoregressive modelling options. See Details.

volume_weights_options

list of volume weighting options. See Details. For simple cases, use the volume_weights parameter in fmri_lm() instead.

soft_subspace_options

list of soft subspace projection options. See Details. For simple cases, use the nuisance_projection parameter in fmri_lm() instead.

Value

An object of class fmri_lm_config.

Details

For common use cases, fmri_lm() provides convenience parameters that are easier to use than these detailed option lists:

  • volume_weights = TRUE enables volume weighting with defaults

  • volume_weights = "tukey" enables with Tukey method

  • nuisance_projection = N enables soft projection with matrix N

  • nuisance_projection = "mask.nii" enables with mask file

Use the *_options lists below only when you need fine-grained control.

robust_options may contain:

  • type (FALSE, "huber", "bisquare")

  • k_huber

  • c_tukey

  • max_iter

  • scale_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)

volume_weights_options may contain:

  • enabled (logical, whether to compute and apply volume weights)

  • method ("inverse_squared", "soft_threshold", "tukey")

  • threshold (numeric, DVARS threshold for weighting)

  • weights (optional pre-computed weight vector)

soft_subspace_options may contain:

  • enabled (logical, whether to apply soft subspace projection)

  • nuisance_mask (path to NIfTI mask or logical vector)

  • nuisance_matrix (pre-computed nuisance timeseries matrix)

  • lambda (numeric, "auto", or "gcv")

  • warn_redundant (logical, warn if baseline has nuisance terms)