[Mesa-users] advice for removing unwanted outputs in favor of speed

Fri Mar 26 14:42:44 UTC 2021

Hi Amber,

If this is for your XRB project, please save the final profiles! Curious people want to know what the ashes are.  I’ve found they seem to converge after around ~15 bursts (except for a couple of odd ones in the A=16-29 range that I have yet to understand), so it’s ok if each run doesn’t have the same number of bursts. Also, assuming you’re considering the whole envelope that’s been enriched by bursts in your mass-fraction calculation, it doesn’t much matter if you happen to be grabbing the profile during or after the burst.

If this isn’t for your XRB project, then please ignore me.

-Zach

----
Zach Meisel
Associate Professor of Physics & Astronomy
Director, Edwards Accelerator Laboratory
Ohio University
Email: meisel at ohio.edu
Web: http://inpp.ohio.edu/~meisel

From: Rob Farmer via Mesa-users<mailto:mesa-users at lists.mesastar.org>
Sent: Friday, March 26, 2021 10:37 AM
To: Francis Timmes<mailto:fxt44 at mac.com>
Cc: amber lauer<mailto:amberlauer at gmail.com>; mesa-users<mailto:mesa-users at lists.mesastar.org>
Subject: Re: [Mesa-users] advice for removing unwanted outputs in favor of speed

Hi,

One strategy I use a lot is to not save any profiles (except maybe the final model). Then just rerun a handful of models that need lots of profile output. This is more to help me keep down the amount of a disk space I use when running many models. The other strategy is to think whether or not you can "summarise" the profile data into a few numbers (say the location where something interesting happens) in your run_star_extras, and then save it to the history output (again so you don't need to write full profiles which can take up alot of space).

I also find photos a hassle when running big grids (keeping track of which models should be ./rn and which ones should be ./re), so unless I expect my models to run for a long time (and thus might be killed by job schedulers on a cluster) I completely turn off photos. If the models need to be rerun I just start them from scratch.

I disagree slightly with Frank about the pgstar plots, i'd recommend not turning them off. Saving the plots is cheap and is what I use to make sure the models are proceeding on track. It saves me more time overall to have the pgstar plots saved and to be able to see whether I need to kill a model and run it again, than saving time from not writing pgstar plots but having to wait till the models finish before realising they did something wrong.

Rob

On Wed, 24 Mar 2021 at 21:35, Francis Timmes via Mesa-users <mesa-users at lists.mesastar.org<mailto:mesa-users at lists.mesastar.org>> wrote:
hi amber,

broadly, minimize i/o. if all plots can be done post-process, eliminate the pgstar writes.
eliminate or minimize what goes into a profile file and the writing of photos.

fxt

> On Mar 24, 2021, at 1:02 PM, amber lauer via Mesa-users <mesa-users at lists.mesastar.org<mailto:mesa-users at lists.mesastar.org>> wrote:
>
> I am getting ready to run a larger group of runs, and it occurs to me that I am very specific in my output needs and that I might speed up the process by removing everything but the outputs I need, which are really few. For instance, I only need about 5 columns in the history.list, and have removed most of the others. I have set my plots to only output every 500 or so, as well as terminal and profile. Are there other things I might not realize can be turned off? Just looking for suggestions.
> =================================
> Dr. Amber Lauer-coles
> Postdoctoral Researcher
> Triangle Universities Nuclear Lab
> Duke University
> amber.lauer at tunl.duke.edu<mailto:amber.lauer at tunl.duke.edu>
> _______________________________________________
> mesa-users at lists.mesastar.org<mailto:mesa-users at lists.mesastar.org>
> https://lists.mesastar.org/mailman/listinfo/mesa-users<https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.mesastar.org%2Fmailman%2Flistinfo%2Fmesa-users&data=04%7C01%7Cmeisel%40ohio.edu%7Cdbf69b7e447643967b0e08d8f062c5dd%7Cf3308007477c4a70888934611817c55a%7C0%7C0%7C637523662647510490%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=5nPhmgFug%2B842%2FbY8VzveMIOo4TFYkv0ZwzGDKHXpKQ%3D&reserved=0>
>

_______________________________________________
mesa-users at lists.mesastar.org<mailto:mesa-users at lists.mesastar.org>
https://lists.mesastar.org/mailman/listinfo/mesa-users<https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.mesastar.org%2Fmailman%2Flistinfo%2Fmesa-users&data=04%7C01%7Cmeisel%40ohio.edu%7Cdbf69b7e447643967b0e08d8f062c5dd%7Cf3308007477c4a70888934611817c55a%7C0%7C0%7C637523662647510490%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=5nPhmgFug%2B842%2FbY8VzveMIOo4TFYkv0ZwzGDKHXpKQ%3D&reserved=0>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20210326/cef35a8b/attachment.htm>