metafor matrix error

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
On Behalf Of Branwen Owen
Sent: Tuesday, May 21, 2013 23:12
To: r-help at r-project.org
Subject: [R] metafor matrix error

hi
     I've been using this code to set a reference level for uni and
multivariate analysis
          rmix$interviewmethodcode<-
relevel(mix$interviewmethodcode,ref="SAQ")
          summary(rma.1<-
rma(yi,vi,mods=~interviewmethodcode,data=rmix,method="SJ",knha=F,weighted=
F,intercept=T))
    giving this output:
          Model Results:
                              estimate      se     zval    pval    ci.lb
ci.ub
          intrcpt                     0.1437  0.0176   8.1485  <.0001
0.1091  0.1783  ***
          interviewmethodcodeACASI    0.0022  0.0412   0.0545  0.9566  -
0.0784  0.0829
          interviewmethodcodeFTFI    -0.0926  0.0656  -1.4112  0.1582  -
0.2213  0.0360

However, when i use the same code with a different variable (below) I get
the error message -  "Error in qr.solve(wX, diag(k)) : singular matrix 'a'
in solve".
         rmix$continent<-relevel(mix$continent,ref="North America")
         summary(rma.1<-
rma(yi,vi,mods=~continent,data=rmix,method="SJ",knha=F,weighted=F,intercep
t=T))

Can anyone help to see where the problem is?
    thanks
    Branwen

--
View this message in context: http://r.789695.n4.nabble.com/metafor-
matrix-error-tp4667653.html
Sent from the R help mailing list archive at Nabble.com.
	[[alternative HTML version deleted]]