[R] Significance Test for Correlation Coefficients

Matt McCall mmccall at mail.nih.gov
Mon Jul 26 22:30:30 CEST 2004


Hello all,
I have a set of  microarray timeseries (approx. 18000 genes and 7 time 
points per gene) data that I have correlated with a template (a vector 
of length 7).  I have also correlated random data from a normal 
distribution with the same mean and standard deviation as my data set 
with the template, as well as, correlating a random permutation of my 
data set with the template.  So I now have 3 vectors of length 18000 
contains correlation coefficients.  I was wondering if there was a test 
in R to determine if the correlation coefficients are significantly 
different.  In particular, I want to compare the correlation 
coefficients from my data set with each of the other two.  Thank you 
for any help you can give me.
Sincerely,
Matt




More information about the R-help mailing list