mubada: Multiple Subjects Barycentric Discriminant Analysis

View source: R/mubada.R

mubadaR Documentation

Multiple Subjects Barycentric Discriminant Analysis

Description

Multiple Subjects Barycentric Discriminant Analysis

Usage

mubada(
  Y,
  Xlist,
  ncomp = 2,
  preproc = center(),
  normalization = c("MFA", "RV", "None", "RV-MFA", "custom"),
  A = NULL,
  M = NULL,
  ...
)

Arguments

Y

dependent factor variable. If All 'X' matrices have same number of rows, 'Y' can be a single factor vector. If there are a different number of rows (e.g. different numbers of replications per subject), Y can be a list of factors.

Xlist

a list of X matrices, one per subject, or it is a list of projector objects.

ncomp

number of common components to estimate.

normalization

the type of normalization.

A

a vector or symmetric matrix of custom column constraints

M

a vector or symmetric matrix of custom row constraints

...

args to send to genpca

center

whether to center the variables.

scale

whether to scale the variables by 1/sd.

References

Abdi, H., Williams, L. J., & Bera, M. (2017). Barycentric discriminant analysis. Encyclopedia of Social Network Analysis and Mining, 1-20.

Abdi, H., Williams, L. J., Connolly, A. C., Gobbini, M. I., Dunlop, J. P., & Haxby, J. V. (2012). Multiple Subject Barycentric Discriminant Analysis (MUSUBADA): how to assign scans to categories without using spatial normalization. Computational and Mathematical Methods in Medicine, 2012.

Examples


Y <- replicate(3, factor(sample(letters[1:3], 20, replace=TRUE)), simplify=FALSE)
Xlist <- replicate(3, matrix(rnorm(20*100),20,100), simplify=FALSE)
mures <- mubada(Y, Xlist)

bbuchsbaum/neuropls documentation built on April 17, 2022, 8:46 a.m.