psych (version 1.0-17)

polychor.matrix: Actually, what does this do? Convert a matrix of phi coefficients to a matrix of polycoric correlations

Description

Given a vector of a vector of frequencies, use John Fox's polycor function to convert these to polychoric correlations.

Not ready for public consumption.

Usage

polychor.matrix(x, y = NULL)

Arguments

x

a matrix of phi coefficients

y

perhaps

Value

• ~Describe the value returned If it is a LIST, use
• comp1
  Description of 'comp1'
• comp2
  Description of 'comp2'
• ...

Details

This is a stub of a function to convert a matrix of phi coefficients with an accompanying vector of frequencies into polychoric correlations. Clearly it is not there yet, but is included in the package as a stopgap.

Please do not use.

Examples

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as
function(x,y=NULL) {
	require(polycor)
 sizex <- dim(x)[2]
if (((is.data.frame(y))|(is.matrix(y))))  sizey<-dim(y)[2] 
   else  sizey <- dim(x)[2]
        
 result<-matrix(1,nrow=sizey,ncol=sizex)   #create the output array

 xnames<- names(x)
 colnames(result)<- names(x)

if (((is.data.frame(y))|(is.matrix(y))))  rownames(result) <- names(y)
    else  rownames(result) <- names(x)
 
    
if (!((is.data.frame(y))|(is.matrix(y)))) {     #default case returns a square matrix
   for (i in 2: sizex ) {
     for (j in 1:( i-1)) {
      result[j,i]<-polychor(table(x[,j],x[,i]) )
       result[i,j] <- result[j,i]
       }
      }
      }
     else {                   #if y is input, then return the rectangular array
        for (i in 1: sizex ) {
            for (j in 1:sizey) {
                 result[j,i]<-polychor(table(x[,i],y[,j]) )
                     }
                } }
   return (result) }