[R] R2WinBugs problem

[Uwe Ligges]

If the exactly same code runs on one machine and not on another where R 
version, script, model  file, inits, data etc. are all identical, then I 
suspect you have different versions of R2WinBUGS installed.

Please check is really all files and all version numbers are the same.

If you still have the problem afterwards, please send the relevant files 
that make the code below reproducible for us.

Best wishes,
Uwe Ligges


On 09.09.2010 22:53, Mark Irwin wrote:
>
> I'm having difficulty running R2WinBugs on a setup that previously
> worked for me (Dell Laptop, Windows XP service pack 3, R 2.11.1, WinBugs
> 1.43) . When I issue the following command
>
> smds.sim <- bugs (data, inits, parameters, "SMDSbrandLoc2.bug",
> debug = T,
> n.thin = n.thin,
> n.chains = n.chains,
> n.burnin = n.burnin,
> n.iter = n.iter)
>
> I get the following error in R
>
> Error in while (.regexpr("\r", info) > 0) { : argument is of length zero
>
> Also in WinBugs I get the following messages
>
> dic.set()
> command #Bugs:dic.set cannot be executed (is greyed out)
> update(5)
> coda(*,C:/DOCUME~1/irwin/LOCALS~1/Temp/RtmpOwZetY/coda)
> command #Bugs:coda cannot be executed (is greyed out)
> stats(*)
> command #Bugs:stats cannot be executed (is greyed out)
> dic.stats()
>
> DIC
> history(*,C:/DOCUME~1/irwin/LOCALS~1/Temp/RtmpOwZetY/history.odc)
> command #Bugs:history cannot be executed (is greyed out)
>
> I am able to run the analysis on a Mac under Parallel Desktop (XP
> Service Pack 3, R 2.11.1, WinBugs 1.43) and on an Amazon Cloud Machine
> (R 2.11.1, WinBugs 1.43, not sure which version of Windows).
>
> Does anybody have any idea what might be causing that error? An even
> better, an idea how to fix it?
>
> Mark
>
>