RecursiveNested: RecursiveNested - a recursive function that will calculate...

In DCAN-Labs/MarginalModelCIFTI: MarginalModelCifti constructs and tests marginal models for surface and volume data.

Description

This function is used in the fastSWEcalculation.R function to quickly calculate SwE for a nested variable.

Usage

RecursiveNested(
  iter,
  nests,
  nested,
  datavec,
  Sinit,
  residarray,
  Breadvec,
  npred,
  adjustment,
  X,
  Sb
)

Arguments

iter	The current iteration, when called it should be set to 1.
nests	The range of blocks within the nested design should be ordered in same order as the set of whole numbers (e.g. 1,2,3,4...)
nested	An integer vector denoting the blocks ordered by the input data
datavec	An integer denoting the number of vertices/voxels to be tested
Sinit	The initial covariance structure used by the sandwich estimator, default is a 1xPxdatavec numeric array set to zeros
residarray	The residuals from the initial fit of the model to the voxel/vertex-wide dataset, should be a numeric array of num_casesXnum_voxel/vertices
Breadvec	The "bread" of the sandwich estimator, calculated from the predictor variables P, should be a numeric array of Pxnum_cases
npred	The number of predictors, P. Should be an integer.
S	The covariance structure estimated from the sandwich estimator, updated upon each recursive estimation.

Examples

SwE <- RecursiveNested(iter=1,nests=1:Nschool,nested=Ischool,datavec=Nelm,Sinit=S0,residarray=res,Breadvec=BreadX,npred=P,adjustment=adjustment,hat_adjust=hat_adjust,S)

DCAN-Labs/MarginalModelCIFTI documentation built on Nov. 30, 2021, 3:40 p.m.