matrix-constructors: Constructors for structured correlation and covariance matrix

In emitanaka/simulate: Parametric simulation framework to generate complex multivariate and multilevel data

Constructors for structured correlation and covariance matrix

Description

Correlation and covariance matrices are assumed to be symmetrics, semi-positive definite. In the case of correlation matrices, the diagonal elements should be 1 and the off-diagonal elements between -1 and 1 inclusive.

Usage

matrix_id(n, vars = 1, names = NULL, inverse_return = FALSE)

matrix_diag(n, vars = 1, names = NULL, inverse_return = FALSE)

matrix_cs(n, rho = 0, vars = 1, names = NULL, inverse_return = FALSE)

matrix_ar1(n, rho = 0, vars = 1, names = NULL, inverse_return = FALSE)

matrix_tp(n, rho = 0, vars = 1, names = NULL, inverse_return = FALSE)

matrix_ad1(n, u, vars = 1, names = NULL, inverse_return = FALSE)

Arguments

n	The dimension of the matrix.
vars	A vector of diagonal variances or a single variance.
names	Optional character vector of the same size as n.
inverse_return	Whether to return the inverse or not.
rho	The

Functions

• matrix_id: Identity matrix.

• matrix_diag: A synonym for matrix_id.

• matrix_cs: Compound symmetric matrix.

• matrix_ar1: First order autoregressive order matrix.

• matrix_tp: Toeplitz.

• matrix_ad1: First order ante-dependence order matrix.

emitanaka/simulate documentation built on July 1, 2022, 9:32 a.m.