On Jul 13, 2011, at 6:09 AM, Mitra, Sumona wrote:
Dear all,
Thank you so much for your help. I was able to get a script that ran
and did what i wanted it to do. It was as follows:-
correlation <- function(a)
+ {
+ r <- matrix(nrow = dim(a)[1], ncol=dim(a)[1])
+ for(x in c(2:dim(a)[1]))
+ {
+ for(y in c(x:dim(a)[1]))
+ {
+ r[x,y] <- (cor.test(as.vector(as.matrix(a)[x,],
mode="double"), as.vector(as.matrix(a)[y,], mode="double"))$p.value)
+ r[y,x] <- r[x,y]
+ }
+ }
+ return(r)
+ }
Now I want to extend this to include all the possible correlations
that cor.test can do.
You are constructing a technique that in the wrong hands will do very
bad statistics.
One of the possible arguments of cor.test is method="method", where
method can be pearson, spearman or kendall. To include this I
modified my code in the following way:-
correlation <- function(a, b)
+ {
+ r <- matrix(nrow = dim(a)[1], ncol=dim(a)[1])
+ for(x in c(2:dim(a)[1]))
+ {
+ for(y in c(x:dim(a)[1]))
+ {
+ r[x,y] <- (cor.test(as.vector(as.matrix(a)[x,],
mode="double"), as.vector(as.matrix(a)[y,], mode="double"),
method="b")$p.value)
If you quote "b", then the interpreter will not substitute the value
of `b`.
+ r[y,x] <- r[x,y]
+ }
+ }
+ return(r)
+ }
This when executed though, gives the following error:-
correlation(a, spearman)
Error in match.arg(method) :
'arg' should be one of “pearson”, “kendall”, “spearman”
Please help!
Sumona
--
David Winsemius, MD
West Hartford, CT
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
