| obliqueR {lazy.fa} | R Documentation |
Planar Oblique Rotation with grid search of optimal rho.
Description
Planar Oblique Rotation with grid search of optimal rho.
Usage
obliqueR(
func,
A,
...,
sim = 1,
method = "nlm",
SQUAREM = 3,
init = "pca",
minr = 0,
maxr = 0.7,
npoint = 0,
by = 0.1,
print = 0
)
Arguments
A |
The factor loadings matrix |
... |
additional parameters to func such as kappa or gamma parameter. |
sim |
= 0 to use successive method |
SQUAREM |
1 or 3 to use SQUAREM |
init |
= Initial rotation method: "pc", "varimax", or "promax" |
minr |
The lower bound of rho |
maxr |
The upper bound of rho |
npoint |
# of discrete points for rho in [minr,maxr] or 0 |
by |
The by parameter for rho=seq(minr,maxr,by) |
print |
= 1 to print the result |
funuc |
The criterion function to be minimized. |
Details
For each rho in [minr,maxr],
define the factor correlation matrix R as
R=(1-\rho)I - \rho 1 1'
and use planarR function to find the rotation matrix T, and accordingly
invtW=P %*% diag(sqrt(Lambda)) %*% Tk %*% diag(1/sqrt(Delta)) %*% t(Q)
so that B=A %*% invtW has a simple structure mesured by
function "func"..
The value of rho which minimizes the criterion function calculated
by "func" will be chosen as the best value of rho.
Examples
seed=1701
set.seed(seed)
n=20; ndim=5
A=gendatafa_A( n, ndim, pc=1 )$loadings
res=obliqueR( critOBR, A, sim=1, print=1, minr=0, maxr=0.8, by=0.1 )