R help - Nov 2012 - Choleski decomposition

m <- matrix(nrow=5, ncol=5)
	m <- ifelse(row(m)==col(m), 1, 0.2)
	c <- chol(m)	# Choleski decomposition
	u <- matrix(rnorm(2000*5), ncol=5)
	uc <- u %*% c
	cr <- pnorm(uc)
	cr <- qbinom(cr,1,0.5)
	cor(cr) 
I expected that the cor(cr) to be 0.2 as i set in m, but the result is
around 0.1.
Why is that? Thanks



--
View this message in context:
http://r.789695.n4.nabble.com/Choleski-decomposition-tp4651475.html
Sent from the R help mailing list archive at Nabble.com.

On Nov 30, 2012, at 10:34 AM, jaybell wrote:
> 	m <- matrix(nrow=5, ncol=5)
> 	m <- ifelse(row(m)==col(m), 1, 0.2)
> 	c <- chol(m)	# Choleski decomposition
> 	u <- matrix(rnorm(2000*5), ncol=5)
> 	uc <- u %*% c
> 	cr <- pnorm(uc)
> 	cr <- qbinom(cr,1,0.5)
> 	cor(cr) 
> I expected that the cor(cr) to be 0.2 as i set in m, but the result is
> around 0.1.
> Why is that? Thanks
Cross-posting to rhelp and Stackoverflow is deprecated (at least on Rhelp). If
you didn't get effective advice after a reasonable interval (measured in
days) it would make sense to look elsewhere. But this is essentially a
statistics questions rather than a coding question and you were advised on SO to
post the question on a website that is designed for statistics questions. Rhelp
is not that site.

-- 

David Winsemius, MD
Alameda, CA, USA