Description Usage Arguments Value Author(s) References Examples
This function conducts rotations of loading matrix based on cluster structure of variables. It can be used for two ways. One is to find simple and well-grouped structure simultaneously, and the other is to find simple structure based on the prior information on cluster structure. The objective function is optimized using the Gradient Projection (GP) algorithm and k-means algorithm alternately.
1 2 3 4 |
A |
The loading matrix for rotation. |
cluster |
The vector of cluster parameters which indicate a cluster where each variable is assigned. |
normalize |
If |
N.random |
The number of sets of random initial values of rotation matrix |
fixed |
If |
N.cluster |
The number of clusters. If this is null, the number of factors is used for the number of clusters. |
ini.cluster |
The vector of initial cluster parameters. |
maxit |
The upper limit of iteration of the algorithm. |
alpha |
The initial value of |
method |
For "oblimin", |
geomin.par |
The parameter for geomin rotation. The default is 0.01. |
oblimin.index |
The parameter for oblimin rotation. The default is 0, which imply the Quartimin rotation. |
fit.cr |
The criterion for convergence of algorithm, abs(criteria.new - criteria.old). |
A |
Rotated loading matrix. |
T |
Estimated rotation matrix. |
E |
Estimated correlation matrix among factors. |
conv |
If |
cluster |
Estimated cluster parameter. |
cr |
The value of objective function finally attained. |
Michio Yamamoto
michio.koko@gmail.com
Yamamoto, M. and Jennrich, R.I. (2012). A cluster-based factor rotation. submitted.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 | ##loading matrix for rotation (Perfect simple structure)
##Find optimal T and U (cluster parameter) simultaneously
A <- matrix(c(.8, 0, 0,
.7, 0, 0,
0,.8, 0,
0,.7, 0,
0, 0,.8,
0, 0,.7), 6, 3, byrow=TRUE)
(out <- obliclus(A, N.random=1)) ##Using many random starts is recommended.
##loading matrix for rotation (More complex structure)
##Find optimal T based on the information on cluster structure
A <- matrix(c( 1, 0, 0,
0, 1, 0,
0, 0, 1,
0.9, 0.6,-0.3,
-0.3, 0.9, 0.6,
0.6,-0.3, 0.9) ,6,3,byrow=TRUE)
cluster <- c(1,2,3,4,4,4)
(out <- obliclus(A, cluster=cluster, fixed=TRUE, N.cluster=4))
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.