glmmGS.CovarianceModel: Construct covariance models for the 'glmmGS' function
In glmmGS: Gauss-Seidel Generalized Linear Mixed Model solver

Description Usage Arguments Value Author(s) See Also Examples

Construct a list defining a covariance model for the glmmGS function.

1	glmmGS.CovarianceModel(type, ...)

`type`	a string specifying the type of the covariance model. Current valid types are: `identity`, `precision`. The covariance components of the identity and precision models represent the global precision parameters.
`...`	optional arguments containing information about the precision or covariance structure. If the type is `identity`, no optional arguments are required; if the type is `precision`, the optional argument is a precision matrix of either R `matrix` type or `glmmGS.SparseMatrix` type.

A list of parameters specifying a covariance model for the glmmGS function.

Michele Morara, Louise Ryan, Subharup Guha, Christopher Paciorek

glmmGS, glmmGS.SparseMatrix

# Set number of columns
	ncols <- 100;

# Identity covariance model (no need to specify matrix dimensions)
 	# Define dense-precision model
 	I <- glmmGS.CovarianceModel("identity");

# Dense-precision model

	# Dense precision matrix
	R <- as.matrix(rnorm(ncols * ncols), nrow = ncols, ncol = ncols);
	R <- R %*% t(R);
	
 	# Define dense-precision model
 	T.dense <- glmmGS.CovarianceModel("precision", R);

# Sparse-precision model

	# Sparse precision matrix
	R <- diag(rep(1, ncols));
	for (i in 2:ncols)
	{
		R[i - 1, i] <- 0.5;
		R[i, i - 1] <- 0.5;
	}
	
	# Create sparse matrix from R
	R.sparse = glmmGS.SparseMatrix(R);
	
 	# Define sparse-precision model
 	T.sparse <- glmmGS.CovarianceModel("precision", R.sparse);