MultiRFM (version 1.1.0)

selectFac.MultiRFM: Select the number of factors

Description

Select the number of factors that are shared among studies q and thos that are specific to individual studies(qs).More details are in Section 3.1 of the article.

Usage

selectFac.MultiRFM(
  XList,
  q_max = 15,
  qs_max = 4,
  method = c("SSVR", "CUP"),
  threshold = 1e-05,
  cup.upper = 0.95,
  epsELBO = 1e-05,
  maxIter = 30,
  verbose = TRUE,
  seed = 1
)

Value

return a list contains the following components:(1) q, the number of shared factors; (2) qs,the number of specified factors.

Arguments

XList

A length-M list, where each component represents a matrix and is the

q_max

an optional integer, specify the maximum number of study-shared factors; default as 15.

qs_max

an optional integer, specify the maximum number of study-specified factors; default as 4.

method

an optional character, contains the methods of "SSVR" and "CUP", where `SSVR` is the sequential singular value ratio method while `CUP` is the criterion based on cumulative proportion of explained variance.

threshold

the cutoff of the singular values, where the singular values less than this value will be removed.

cup.upper

upper limit of the cumulative proportion of explained variance.

epsELBO

an optional positive value, tolerance of relative variation rate of the evidence lower bound value, defualt as '1e-5'.

maxIter

the maximum iteration of the VEM algorithm. The default is 30.

verbose

a logical value, whether output the information in iteration.

seed

an optional integer, specify the random seed for reproducibility in initialization;default as 1.

Details

None

Examples

p <- 100
nvec <- c(150,200); qs <- c(2,2)
datList <- gendata_simu_multi(seed=1, nvec=nvec, p=p, q=3, qs=qs, rho=c(5,5),
        err.type='mvt', sigma2_eps = 1, nu=3)
XList <- datList$Xlist;
## Set maxIter=5 for demonstration while set it to 30 in the formal run.
hqlist <- selectFac.MultiRFM(XList, q_max=6, qs_max= rep(4,2), maxIter = 5) #
str(hqlist)