[ADMB Users] survey of memory settings in ADMB
Ian Taylor
ian.taylor at noaa.gov
Fri Mar 23 13:13:07 PDT 2012
Thanks for all these Arni,
I'm also looking at the examples models included with ADMB and indeed the
admb-re examples have many more settings than the non-re examples.
I'm also assuming that very few ADMB users have less than 2GB available,
but will try not to require more by default.
Someday we'll follow Dave's suggestion to make the memory allocation
dynamic, but for now some new defaults will hopefully work alright for
average models.
-Ian
On Fri, Mar 23, 2012 at 12:56 PM, Arni Magnusson <arnima at hafro.is> wrote:
> More examples I have come across:
>
> ADCAM (stock assmt model by Bjornsson)
>
> TOP_OF_MAIN_SECTION
> gradient_structure::set_**CMPDIF_BUFFER_SIZE(10000000);
> gradient_structure::set_**GRADSTACK_BUFFER_SIZE(1000000)**;
> gradient_structure::set_MAX_**NVAR_OFFSET(1500);
> gradient_structure::set_NUM_**DEPENDENT_VARIABLES(1500);
> arrmblsize = 50000000;
>
>
> AMAK (stock assmt model by Ianelli)
>
> TOP_OF_MAIN_SECTION
> gradient_structure::set_MAX_**NVAR_OFFSET(1000);
> gradient_structure::set_NUM_**DEPENDENT_VARIABLES(1000);
> gradient_structure::set_**GRADSTACK_BUFFER_SIZE(1000000)**;
> gradient_structure::set_**CMPDIF_BUFFER_SIZE(10000000);
> arrmblsize=500000000;
>
>
> Stock Synthesis (stock assmt model by Methot)
>
> TOP_OF_MAIN_SECTION
> arrmblsize = 50000000;
> gradient_structure::set_**GRADSTACK_BUFFER_SIZE(**20000000000);
> gradient_structure::set_**CMPDIF_BUFFER_SIZE(170000000);
> gradient_structure::set_MAX_**NVAR_OFFSET(5000);
> gradient_structure::set_NUM_**DEPENDENT_VARIABLES(10000);
>
>
> Pacific Ocean Perch (stock assmt by Stewart)
>
> TOP_OF_MAIN_SECTION
> gradient_structure::set_MAX_**NVAR_OFFSET(1000);
> gradient_structure::set_NUM_**DEPENDENT_VARIABLES(800);
> gradient_structure::set_**GRADSTACK_BUFFER_SIZE(100000);
> gradient_structure::set_**CMPDIF_BUFFER_SIZE(1000000);
> arrmblsize=900000;
>
>
> Red Snapper (stock assmt by Prager)
>
> TOP_OF_MAIN_SECTION
> arrmblsize=20000000;
> gradient_structure::set_MAX_**NVAR_OFFSET(1600);
> gradient_structure::set_**GRADSTACK_BUFFER_SIZE(2000000)**;
> gradient_structure::set_**CMPDIF_BUFFER_SIZE(2000000);
> gradient_structure::set_NUM_**DEPENDENT_VARIABLES(500);
>
>
> S-Atlantic Vermillion Snapper (stock assmt by Williams)
>
> TOP_OF_MAIN_SECTION
> arrmblsize=2000000;
> gradient_structure::set_MAX_**NVAR_OFFSET(300);
> gradient_structure::set_**GRADSTACK_BUFFER_SIZE(**20000000);
> gradient_structure::set_**CMPDIF_BUFFER_SIZE(100000000);
> gradient_structure::set_NUM_**DEPENDENT_VARIABLES(300);
>
>
> Walleye (stock assmt by Bence)
>
> TOP_OF_MAIN_SECTION
> gradient_structure::set_**GRADSTACK_BUFFER_SIZE(200000);
> gradient_structure::set_**CMPDIF_BUFFER_SIZE(1000000);
> gradient_structure::set_MAX_**NVAR_OFFSET(400);
>
>
> Many of these settings might be arbitrary and even unnecessary, but may
> still give us some ideas for new defaults.
>
>
>
> Arni
>
>
>
>
> On Thu, 22 Mar 2012, Ian Taylor wrote:
>
> Hi ADMB Users,
>>
>> The default memory settings in ADMB are out of date and I've been tasked
>> with coming up with replacements that will work for reasonable sized models
>> (big models will still require manual settings). It occurs to me that it
>> makes sense to survey what common requirements of ADMB users might be.
>>
>> Have you been required to change the memory settings for your model using
>> either commands like the following in the TOP_OF_MAIN_SECTION,
>>
>> arrmblsize=500000;
>> gradient_structure::set_**GRADSTACK_BUFFER_SIZE(200000);
>> gradient_structure::set_**CMPDIF_BUFFER_SIZE(2100000);
>> gradient_structure::set_MAX_**NVAR_OFFSET(500);
>>
>> or through command line arguments like -ams, -gbs, -cbs, or -mno?
>>
>> If so, and if you've put thought into the choices (rather than blindly
>> picking values), could you send me your settings in a reply to this email?
>>
>> Thanks,
>>
>> -Ian
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.admb-project.org/pipermail/users/attachments/20120323/36053862/attachment.html>
More information about the Users
mailing list