simu.capr: Simulate covariance matrices compatible with capr()
Description
Generates a simple synthetic dataset for CAP regression consisting of a
covariance cube, design matrix, and the latent orthogonal directions used to
build the covariance slices.
Usage
simu.capr(seed = 123L, n = 120L)
Value
A list with components:
S
Array of dimension \(p \times p \times n\) holding the simulated
covariance matrices.
X
Design matrix of size \(n \times 2\) with an intercept and a
Bernoulli covariate.
Q
Orthogonal matrix whose columns are the latent directions.
BetaMat
True coefficient matrix used to form the eigenvalues.
H
Average covariance matrix \(\frac{1}{n}\sum_i S_i\).
p, n
The dimension and sample size supplied to the generator.