[R] Re: none

Douglas Bates bates at stat.wisc.edu
Wed Dec 22 21:59:09 CET 1999

"Troels Ring" <tring at mail1.stofanet.dk> writes:

> I've forgotten how to simulate data with a known correlation,
> e.g. only two columns. I'd be most pleased to be told.

Assuming that you want data from a multivariate normal distribution
with a given correlation, install the MASS package and use mvrnorm.
 > library(MASS)
 > example(mvrnorm)

 mvrnrm> Sigma <- matrix(c(10, 3, 3, 2), 2, 2)

 mvrnrm> Sigma
      [,1] [,2]
 [1,]   10    3
 [2,]    3    2

 mvrnrm> var(mvrnorm(n = 1000, rep(0, 2), Sigma))
	  [,1]     [,2]
 [1,] 9.870587 2.764234
 [2,] 2.764234 1.840870

Note that as of version 0.90.1 you can get the MASS, nnet, spatial,
and class libraries in a single bundle with
 install.packages("VR")   # packages from the Venables and Ripley bundle
You do need to set the CRAN option to a legitimate CRAN archive site,
have access to the Internet, and have either the wget or lynx programs
on your search path.

The default for options("CRAN") is cran.r-project.org.  You may want
to use cran.dk.r-project.org, the Danish mirror, instead.
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list