hconfigs: Enumeration of all possible vectors of association status.

In repfdr: Replicability Analysis for Multiple Studies of High Dimension

Description

The function generates a matrix with all possible vectors of association status (in rows), given the number of studies and number of possible association status states in each study (2 or 3).

Usage

hconfigs(n.studies, n.association.status = 3, studies.names = NULL)

Arguments

n.studies	Number of studies in the analysis.
n.association.status	either 2 for no-association\association or 3 for no-association\negative-association\positive-association.
studies.names	Optional study names to display.

Details

This matrix should be used when selecting the rows indices for the association status vectors that are in the non-null set, specified by the used in non.null.rows in the function repfdr.

Value

Matrix with rows indicating all the possible vectors of association status.

See Also

repfdr

Examples

(H <- hconfigs(n.studies = 3))
# in replication analysis the non-null vectors are:
H[apply(H,1,function(y){ sum(y==1)>1 | sum(y==-1)>1 }),]
# in meta-analysis there is only one null vector (c(0,0,0)):
H[rowSums(abs(H))!=0,]

hconfigs(n.studies = 3, n.association.status= 2)

repfdr documentation built on May 1, 2019, 9:20 p.m.