RMark 2.1.3

[jlaake]

2.1.2 never made it to CRAN and instead I posted 2.1.3 with the accumulation code and added a hessian=TRUE argument to specify use of hessian for v-c matrix rather than second partial matrix in MARK. It is currently on the Austria CRAN site and may take a day or two to reach all of the mirror sites.

Let me know if there are any problems. The archived packages back to 2.0.7 are on my github site
https://github.com/jlaake/RMark

[lrae]

Hi Jeff, I'm a novice user having some issues using RMark 2.1.3 (with R v. 2.13.1 in Win7), though not sure the package is the problem since I recently updated to MARK v. 6.1 as well. When I attempt to run a model (e.g. "mark(dipper)") program MARK launches (which is something that never occurred before the updates), and while I don't get any error messages in R, I don't get much of anything -- R simply continues to process to seemingly no end.

Thinking there may be a problem with MarkPath I've run the following,

Code: Select all

> os=R.Version()$os
> if(os=="mingw32")
+ {
+ if(!exists("MarkPath"))
+ {
+ MarkPath=Sys.which("mark.exe")
+ if(MarkPath=="")
+ if(file.exists("c:/Program Files/Mark/mark.exe"))
+ MarkPath="c:/Program Files/Mark/mark.exe"
+ else
+ stop("mark.exe cannot be found. Add to system path or specify MarkPath object (e.g., MarkPath='C:/Program Files (x86)/Mark'")
+ }else
+ {
+ if(substr(MarkPath,nchar(MarkPath),nchar(MarkPath))%in%c("\\","/"))
+ MarkPath=paste(MarkPath,"mark.exe",sep="")
+ else
+ MarkPath=paste(MarkPath,"mark.exe",sep="/")
+ }}
> os
[1] "mingw32"
> MarkPath
[1] "c:/Program Files/Mark/mark.exe"


Incidentally, it's only after running that snippet that I can get to the stage in debug results giving the following error message (not all debug output is shown here):

Code: Select all

> data(dipper)
> debug(run.mark.model)
> mark(dipper)
....
Browse[2]>
Error in run.mark.model(model, invisible = invisible, adjust = adjust,  :
  mark.exe, mark32.exe or mark64.exe cannot be found. Add to system path or specify MarkPath object (e.g., MarkPath='C:/Programme/Mark'


********Following model failed to run : Phi(~1)p(~1) **********
....


Any ideas what could be going on?

Thanks for reading, would appreciate any thoughts!

[lrae]

Got this sorted out thanks to an e-mail response from Jeff. It was a matter of straightening out .exe files by simply uninstalling and reinstalling new MARK, no problems with RMark 2.1.3 to speak of!

[Hughiedoogal1]

Hi Jeff and Lrae,

I am also experiencing that exact issue. When I try to run a model (ie: "mark(dipper)"), nothing happens.
Program MARK launches but R processes endlessly. Would you guys mind posting how to fix this? I tried re-installing MARK but this hasn't solved my issue.

Thanks,
Dave

[jlaake]

The solution as posted was to uninstall and re-install MARK. Did you uninstall before re-installing?

--jeff

[Hughiedoogal1]

Thanks for your reply. Prior to posting I uninstalled the program before re-installing. I didn't realize that one file remained in the MARK folder after uninstalling. This needed to be deleted. problem solved. Stupid me!
Thanks again