Hi Steven,
You can just cut the matrix up into a 5 column matrix and use apply as normal
m2<-matrix(as.vector(t(m)), ncol=5, byrow=TRUE)
result<-matrix(apply(m2, 1, mean), ncol=ncol(m)/ncol(m2), byrow=TRUE)
result
--
Patrick Rogers
Dept. of Political Science
University of California, San Diego
Hi All,
I'm trying to run a kernel principal component analysis on a corpus with 89769
text documents. I'm using the kpca command from the kernlab package. Here is
the code:
output<-kpca(text, kernel=worddot(type="spectrum", length=1))
The problem is that when I run the kpca, it bails
2 matches
Mail list logo