StVal: Computing Starting Values for PGICA

In PGICA: Parallel Group ICA Algorithm

Description

The main function uses a iterative algorithm. This function calculates the starting values for the algorithm.

Usage

StVal(fileDir, m = 20, N.s = 150, V = 67749)

Arguments

fileDir	fileDir is a file directory containing the original .rda files.
m	m is the number of independent components you want.
N.s	N.s is the number of subjects.
V	V is the number of observed signals. In the fMRI case, it is the number of voxels.

Details

We used the PVD algorithm by C.M.Crainiceanu et.al to compute starting values for our algorithm.

Value

This function has no returned values, instead it will write the output into .rda files. The W0.rda file contains starting values for W matrix and the appi.rda files are the starting values for S.

Author(s)

Ani Eloyan, Shaojie Chen, Lei Huang, Huitong Qiu

References

Ciprian M.Crainiceanu, Brian S.Caffo, Sheng Luo, Vadim Zipunikov Population Value Decomposition, A Framework for the Analysis of Image Populations http://biostats.bepress.com/jhubiostat/paper220/

Examples

dir.create('./data')
data(PC,package="PGICA")
save(PC,file="./data/sample.rda")
StVal("./data/",m=20,N.s=1,V=30000)
# You will find files W0.rda and app1.rda under your current working directory

PGICA documentation built on May 29, 2017, 7:52 p.m.