calcStat: Pairwise weights

In gRapHD: Efficient Selection of Undirected Graphical Models for High-Dimensional Datasets

Description

Calculates pairwise statistics (-2*log-LR, AIC, or BIC) for each variable pair (edge) in the dataset.

Usage

calcStat(dataset,homog=TRUE,forbEdges=NULL,stat="LR")

Arguments

dataset	matrix or data frame (nrow(dataset) observations and ncol(dataset) variables).
homog	TRUE for homogeneous covariance structure, FALSE for heterogeneous. This is only meaningful with mixed models. Default is homogeneous (TRUE).
forbEdges	list with edges that should not be considered. Matrix with 2 columns, each row representing one edge, and each column one of the vertices in the edge. Default is NULL.
stat	measure to be minimized: LR (-2*log-likelihood), AIC, or BIC. Default is LR. It can also be a user defined function with format: FUN(newEdge,numCat,dataset); where numCat is a vector with number of levels for each variable (0 if continuous); newEdge is a vector with length two; and dataset is a matrix (n by p).

Details

Calculates pairwise statistics (-2*log-LR, AIC, or BIC) for all possible edges, returning the values sorted in descending order.

Value

A matrix with p(p-1)/2 lines and 4 columns, where each line refers to a possible edge, and the columns are: vertex 1, vertex 2, value of the statistic, and number of estimated parameters (degrees of freedom) associated with the edge.

Author(s)

Gabriel Coelho Goncalves de Abreu ([e-mail])
Rodrigo Labouriau ([e-mail])
David Edwards ([e-mail])

Examples

set.seed(7,kind="Mersenne-Twister")
dataset <- matrix(rnorm(1000),nrow=100,ncol=10)
m <- calcStat(dataset,stat="BIC")

data(dsCont)
# m1 <- calcStat(dataset,homog=TRUE,forbEdges=NULL,stat="LR")
#          1. in this case, there is no use for homog
#          2. no forbidden edges
#          3. the measure used is the LR (the result is a tree)
v <- calcStat(dsCont,homog=TRUE,forbEdges=NULL,stat="LR")

# result
head(v)
# column 1: first vertex of the edge
# column 2: second vertex of the edge
# column 3: in this case, -LR
# column 4: number of parameters for the edge
#         [,1] [,2]     [,3] [,4]
#    [1,]   17   27 393.0072    1
#    [2,]   21   27 343.5780    1
#    [3,]   22   25 306.0097    1
#    [4,]   17   21 302.9414    1
#    [5,]   27   32 300.0275    1
#    [6,]   21   32 289.4179    1

gRapHD documentation built on Feb. 9, 2018, 6:05 a.m.