getModelStructure {bgmm}

Model structure

Description

This function creates an object which describes constraints over the model parameters.

Usage

getModelStructure(mean = "D", between = "D", within = "D", cov = "D")

Arguments

Each argument is a single character, by default equal "D" (Different). If argument is set to "E" (or "0" for the argument cov) then the given parameter is constrained. By default all arguments are set to "D".

mean

mean="E" forces equality of the means between the components,

between

between="E" forces equality of the covariance matrices among the components,

within

within="E" forces equality of variances within each covariance matrix i to some constant vi and equality of covariances to some constant wi

cov

cov="0" forces equality of covariances within each covariance matrix i to "0",

Value

List of four elements specifying the constraints on 1) relations between the component means, 2) relations between the covariance matrices of the model components, 3) relations within each covariance matrix and 4) the covariances within each matrix. By default, the function returns an unconstrained structure.

Author(s)

Ewa Szczurek

References

Przemyslaw Biecek, Ewa Szczurek, Martin Vingron, Jerzy Tiuryn (2012), The R Package bgmm: Mixture Modeling with Uncertain Knowledge, Journal of Statistical Software.

Examples

 getModelStructure()
 getModelStructure(mean="E")