Phi coefficient matrix (package psych)

I want to calculate phi coefficient for every pair of the columns. Is there
a way to generate a matrix like a correlation matrix? I know cor function
in the case below gives same answer as phi coefficient.

x <- sample(c(0,1), 10, replace=TRUE)
y <- sample(c(0,1), 10, replace=TRUE)
z <- sample(c(0,1), 10, replace=TRUE)
df <- data.frame(x,y,z)
cor(df)
library(psych)
phi(df)

Thank you,
Kumar Mainali
Postdoctoral Associate
Department of Biology
University of Maryland
?

        [[alternative HTML version deleted]]

HI Kumar,
A simple way is:

phimat<-function(x) {
 xcol<-dim(x)[2]
 newx<-matrix(NA,nrow=xcol,ncol=xcol)
 for(i in 1:xcol) {
  for(j in 1:xcol) newx[i,j]<-phi(table(x[,i],x[,j]))
 }
 rownames(newx)<-colnames(newx)<-colnames(x)
 return(newx)
}
phimat(df)

Jim


On Wed, Apr 22, 2015 at 6:34 AM, Kumar Mainali <kpmainali at gmail.com>
wrote:

I want to calculate phi coefficient for every pair of the columns. Is

there

a way to generate a matrix like a correlation matrix? I know cor function
in the case below gives same answer as phi coefficient.

x <- sample(c(0,1), 10, replace=TRUE)
y <- sample(c(0,1), 10, replace=TRUE)
z <- sample(c(0,1), 10, replace=TRUE)
df <- data.frame(x,y,z)
cor(df)
library(psych)
phi(df)

Thank you,
Kumar Mainali
Postdoctoral Associate
Department of Biology
University of Maryland
?

        [[alternative HTML version deleted]]