Computer-specific model selection results

posts related to the RMark library, which may not be of general interest to users of 'classic' MARK

Computer-specific model selection results

Postby wjensen1 » Mon Aug 21, 2017 5:20 pm

Most strange -- RMark code apparently working fine EXCEPT for the AIC table results for this particular data set -- results (table / model ranks) are consistent among my laptop and my grad student’s two computers – HOWEVER, my office computer returns a different table (different model ranks) – despite having used the SAME script, SAME data set, from SAME thumb drive! Very strange! Is this cosmic interference from the eclipse?! Any advice appreciated.
wjensen1
 
Posts: 6
Joined: Tue Jul 25, 2017 12:39 pm

Re: Computer-specific model selection results

Postby jlaake » Tue Aug 22, 2017 4:07 pm

Do all have the same version of MARK and RMark? Did you use same version of R - 64 vs 32? If you use R64, it uses Mark64 and if you use R32 it uses Mark32. You didn't give much detail as to the differences but I'd start with the above for starters.
jlaake
 
Posts: 1417
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA

Re: Computer-specific model selection results

Postby wjensen1 » Wed Aug 23, 2017 12:09 pm

Thanks much for the reply. Both my grad student's and my version's of MARK are 64, R is 64, and to double check with "R.version$system" at the Console prompt we get:

[1] "x86_64, mingw32"

Not sure what "mingw32" (not 32-bit?) means but both of our computers return this output.

Upon re-installing the RMark package we get:

There is a binary version available but the source version is later:
binary source needs_compilation
RMark 2.2.2 2.2.4 FALSE

installing the source package ‘RMark’

trying URL 'https://cran.rstudio.com/src/contrib/RMark_2.2.4.tar.gz'
Content type 'application/x-gzip' length 915673 bytes (894 KB)
downloaded 894 KB

* installing *source* package 'RMark' ...
** package 'RMark' successfully unpacked and MD5 sums checked
** R
** data
** inst
** preparing package for lazy loading
** help
*** installing help indices
** building package indices
** testing if installed package can be loaded
*** arch - i386
*** arch - x64
* DONE (RMark)

...HOWEVER, my grad student did NOT get the output behind the asterisks above. Her output only reads:

package 'RMark' was successfully unpacked and MD5 sums checked

...but we both appear to be using RMark 2.2.4.

So appears that all are 64-bit. Both of our Operating Systems are 64-bit as well. The AICc ranks from the two machines are different as are the individual model estimates.
wjensen1
 
Posts: 6
Joined: Tue Jul 25, 2017 12:39 pm

Re: Computer-specific model selection results

Postby jlaake » Wed Aug 23, 2017 3:16 pm

Do u all have the same version of Mark? If the answer is yes, I'm not sure what to tell you. Are all models different or just some? You'll have to do some sleuthing. It is not likely this is an RMARK problem.
jlaake
 
Posts: 1417
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA

Re: Computer-specific model selection results

Postby wjensen1 » Thu Aug 24, 2017 6:18 pm

Same version of Mark. All model estimates differ. Not explained by Windows 7 vs. 10 operating systems. I'm at a loss. Will try re-loading Mark and R.
wjensen1
 
Posts: 6
Joined: Tue Jul 25, 2017 12:39 pm

Re: Computer-specific model selection results

Postby bacollier » Fri Aug 25, 2017 10:27 am

wjensen1,
If you don't have any luck with the reinstall, send your RMark code/data offlist to me (bret@lsu.edu) and I will run in through the various set ups Windows/Linux/R/Mark I have here and see if I can help you identify the issue.

If you can sent the MARK output files for the run on your computer and your grad students that would help as well.

\bret
bacollier
 
Posts: 230
Joined: Fri Nov 26, 2004 10:33 am
Location: Louisiana State University

Re: Computer-specific model selection results

Postby wjensen1 » Fri Aug 25, 2017 4:51 pm

Thanks Jeff and Bret! Reloaded Mark and R and now have results that are consistent with my our other machines. Not sure what bug made the mess.
wjensen1
 
Posts: 6
Joined: Tue Jul 25, 2017 12:39 pm


Return to RMark

Who is online

Users browsing this forum: No registered users and 17 guests