RMark/Linux Marknnn.out Extraction (bug?)

[bacollier]

Jeff/Evan,
I am posting this here in case someone else has this problem and happens to search for it. Working on some simple occupancy models using RMark on Linux I noticed that I cannot extract the Marknnn.out file as I usually would.

For example, set up a function "run.Bret.Models{RMark code here} then after the run I call

Bret.Models<-run.Bret.Models #works fine for model sets

I can do the extraction of the various results using

Bret.Models$model.1$results$beta or,

Bret.Models$model.1$results$real

These work fine.

But, if I do: Bret.Models$model.1, or Bret.Models[1], usually it would extract the .txt file (in notepad or gsnote or something). However, if I do this on my Linux machine, I get a line stating: "Received SIGHUP or SIGTERM"

I am assuming that there is some link that Jeff has written that ties the model names to the Marknnn.out files, I tried to dig it out, but Jeff's code was to efficient and I could not find it (sorry Jeff, be less efficient next time )

I am using R 2.10.1, RMark 1.9.6(built under R 2.9.2) with 32-bit MARK on a Linux Ubuntu (Karmic) with ESS-Emacs. I am root user and I do own all the folders I am working in.

Bret

[bacollier]

Sorry for the reply/flooding of inboxes, but I forgot to add that I can use

system("gedit marknnn.out") and that will pull the files up just fine but this is obviously cumbersome to try and remember what ties to what.

Bret

[jlaake]

Bret-

There are instructions for setting up RMark under Linux but you have to read the RMark webpage carefully

http://www.phidot.org/software/mark/rmark/linux/

From the above link you'll see that it says that the default viewer is pico with Linux and that you can set it by setting the variable MarkViewer like
MarkViewer="gedit"

See if that does the trick for you. Alterntatively, you can modify the line in print.mark that sets the viewer and rebuild the package. While many of my functions are admittedly dense, that one is not but you may not have realized where the offending code was. When you type the name of a mark object, R sees it has class "mark" and then looks for a function called print.mark. This is the case with all generic functions in R like print, summary, anova etc that use S3 classes. So when you type summary(model) and model is a mark object, it calls summary.mark. You aren't the only R user that has been fooled by this.

Glad it is working for you in Linux. --jeff

[bacollier]

Ahhhh, (he posts back sheepishly, thinking about all Evan's comments regarding RTFM before you post). I was rooting around in the RMark pdf, I should have checked the Linux website (RTFLW perhaps should be todays mantra). Works fine now Jeff, thanks.

Bret