Learn R Programming

muStat (version 1.7.0)

mu.AND: Pairwise AND

Description

mu.AND aggregates vectors of pairwise orderings created by mu.GE in a hierarchical fashion.

Usage

mu.AND(GE, frml=NULL)

Arguments

GE
matrix each column of which corresponds to one variable anddescribes pairwise orderings
frml
formula describing the hierarchical structure of the variables

Value

mu.AND returns a vector of the same length as the columns of GE, with information about pairwise orderings aggregated according to frml.

Algorithm

mu.AND <- function(GE, frml=NULL) {
  <\dots>
  if (is.null(frml)) {
    <\dots>
    GE  <- sq.array(GE)
    AND <- GE[,,1]^0
    nNA <- AND[,1]*0
    for (i in 1:dim(GE)[3]) {
      nNA <- nNA + diag(GEi <- GE[,,i])
      AND <- AND * (GEi + (1-GEi)*(1-t(GEi))) }
    return(as.numeric(AND * ((c(nNA)%o%c(nNA))>0))) } tkn <- unlist(strsplit(frml,""))
  nok <- attr(regexpr("[0-9,()]+",frml),"match.length")
  <\dots>
  tmp <- matrix(0, dim(GE)[1]+1, <\dots>)
  FstFree <- function(tmp) match(TRUE, tmp[1,]==0, nomatch=0) level <- i <- 0
  while ((i <- i+1) <= 0="" nok)="" {="" switch(="" tkn[i],="" "("="level" <-="" level="" +="" 1,="" ","="next," ")"="{" tmp[1,="" use="" (tmp[1,]="=level)]" tmp[,="" fstfree(tmp)]="" c(level="" level-1,="" mu.and(tmp[-1,="" use]))},="" num="" as.numeric(substring(="" frml,i,i<-i-1+regexpr("[,)]",substring(frml,i+1)))="" <\dots="">
        tmp[,FstFree(tmp)] <- c(level, GE[, num]) } ) }
  return(tmp[-1,1])
}

Examples

Run this code
mu.AND(mu.GE(matrix(1:60, , 3)))
a.1 <- 1:10
a.2 <- 3:12
b <- c(1:5,2:6)
mu.AND(mu.GE(cbind(a.1,a.2,b)), frml="((1,2),3)")

Run the code above in your browser using DataLab