Re: [R] survey: estimating a covariance matrix

2008-01-28 Thread Daniel Oberski
Thomas and David, Thanks for your answers. The svyvar() function does the trick. It does not provide a vcov() method for the variance-covariance matrix of these estimates themselves. But I guess I can bootstrap them using library(boot)... Thanks again, daniel On Jan 25, 2008 6:13 PM, Thomas Lu

Re: [R] survey: estimating a covariance matrix

2008-01-25 Thread Thomas Lumley
On Wed, 23 Jan 2008, Daniel Oberski wrote: > Hello > > Does anybody happen to know if it is possible to use the survey package to > estimate a covariance matrix from a complex survey? Yes. svyvar() in the survey package estimates a population covariance matrix. It doesn't give standard errors f

Re: [R] survey: estimating a covariance matrix

2008-01-23 Thread David Winsemius
David Winsemius <[EMAIL PROTECTED]> wrote in news:[EMAIL PROTECTED]: > "Daniel Oberski" <[EMAIL PROTECTED]> wrote in > news:[EMAIL PROTECTED]: > >> Hello >> >> Does anybody happen to know if it is possible to use the survey >> package to estimate a covariance matrix from a complex survey? >>

Re: [R] survey: estimating a covariance matrix

2008-01-23 Thread David Winsemius
"Daniel Oberski" <[EMAIL PROTECTED]> wrote in news:[EMAIL PROTECTED]: > Hello > > Does anybody happen to know if it is possible to use the survey > package to estimate a covariance matrix from a complex survey? > > I have design weights and clusters (no strata), and want to get a > covariance m

[R] survey: estimating a covariance matrix

2008-01-23 Thread Daniel Oberski
Hello Does anybody happen to know if it is possible to use the survey package to estimate a covariance matrix from a complex survey? I have design weights and clusters (no strata), and want to get a covariance matrix with preferably the effective sample size or else an estimate of the variance-co