Page 1 of 1
ivmlwin macro problem with "obey vmake.iv"
Posted: Tue Jun 18, 2013 11:57 pm
by drddtorres
When I run MLwiN's IGLS estimation procedure with the ivmlwin macros in the pre and post paths, I can get estimates of coefficients but the run doesn't complete because for some reason the "obey vmake.iv" command in the post.txt file isn't executing. Does anyone know how to fix this?
Thanks,
Diego
Re: ivmlwin macro problem with "obey vmake.iv"
Posted: Wed Jun 19, 2013 8:35 am
by ChrisCharlton
It's probably worth checking that the file is in the directory that you have defined as the "fpath" in Options->Directories. If it is then you should also check that Windows hasn't renamed it when you extracted the file (it sometimes adds a .txt extension). If you have "hide file extensions" turned on in Windows then you will need to turn this off (you can do this in explorer through the Folder Options->View settings. If neither of these work could you post the exact error message that MLwiN is giving?
Re: ivmlwin macro problem with "obey vmake.iv"
Posted: Wed Jun 19, 2013 5:45 pm
by drddtorres
Chris,
You were correct about windows adding .txt to the end of the ivmlwin files I downloaded. I simply changed those extension to .iv and reran my simply model. I do, however, get the following error:
error while obeying batch file vmake.iv at line number 14:
omega b208 'temp3'
level id columns must be the same length.
When I view the Names window, I notice the n for the individual level id is no longer the same as the other two ids. What might be the cause of this?
Diego
Re: ivmlwin macro problem with "obey vmake.iv"
Posted: Wed Jun 19, 2013 5:54 pm
by ChrisCharlton
This column overwriting is the result of the bug in MLwiN v2.27 that I referenced here:
http://www.cmm.bristol.ac.uk/forum/view ... =752#p1726. To use these macros you will either have to use a previous version (e.g. v2.26) or wait for the next update.
Re: ivmlwin macro problem with "obey vmake.iv"
Posted: Wed Jun 19, 2013 6:17 pm
by drddtorres
Ah, Chris,
You're beautiful. Haha. Thanks so much for all your help. I'm getting closer and closer to being able to run my model without issue.
Diego