[Developers] Running ADMB with separate exe and data dirs

[Arni Magnusson]

I forgot to mention that the results from Experiments 1-4 are identical in 
Linux and Windows.

Also, when I wrote "what the user would expect" (see below), I really 
meant "what the user would find most practical". For analysis and 
archiving purposes, I would always want the data input and model output 
close to each other.

In practice, I think Experiment #2 is the relevant one. To allow ADMB 
users to keep the model in a 'exe' directory somewhere, and be able to use 
it in a 'data' directory.

Okay, I just tried a new experiment where I add the 'exe' (bin, really) 
directory to the PATH:

   Experiment #5, from within data (with 'exe' in the PATH)
   simple -ind simple.dat
   => all results in 'data', excellent

I think Experiment #2 should work the same way, and I wouldn't be 
surprised if the upcoming ADMB release includes this improvement.

Arni



On Fri, 5 Nov 2010, Ben Bolker wrote:

> I guess I would have expected all of the output to show up in whatever 
> directory the executable was *called from*, regardless of where the data 
> or executable files were.  Probably best not to change behavior that 
> users are relying on ... But it's good to know the reason in any case.
>
> Is the behavior different on Windows?  In the current version of 
> glmmADMB the copying of the executables to a temporary directory only 
> happens on Unix, not Windows -- on Windows it's run from the system 
> directory.
>
>
>
>>   Experiment #1, from within exe:
>>   simple -ind ~/data/simple.dat
>>   => all results created in 'exe', and no files are added to 'data'
>>
>>   Experiment #2, from within data:
>>   ~/exe/simple -ind simple.dat
>>   => useful output in 'exe' and garbage in 'data'
>>
>>   Experiment #3, from within data:
>>   ~/exe/simple -ind ~/data/simple.dat
>>   => useful output in 'exe' and garbage in 'data'
>>
>>   Experiment #4, from within third:
>>   ~/exe/simple -ind ~/data/simple.dat
>>   => useful output in 'exe' and garbage in 'third'
>>
>> I guess what the user would expect is that in all four cases, all 
>> output files should appear in the same directory as the input data 
>> file.
>