[mesa-users] Accessing profile data from run_stars_extra.f
paxton at kitp.ucsb.edu
Fri Sep 12 15:08:22 EDT 2014
I agree that we need to extend star_lib. For profile values, I think it would be best to make it include "other" columns as well as the built-in ones. Let's split it into 2 routines: star_get_profile_id(star_id, name) which calls get_axis_id in pgstar_support, and star_get_profile_val(star_id, profile_id, k) which returns the value of the profile at cell k by calling get_val in pgstar_support. How does that sound?
Also, it would be great if you'd do a sweep through star/private/star_utils and add interfaces to star_lib to any of those routines where it makes sense.
And while you are at it, we should have similar interface routines for history items for the current step -- you might base that on get1_hist_value in history.f.
On Sep 12, 2014, at 11:55 AM, Robert FARMER wrote:
> So i'm seeing if anyone else has ran into this/found a solution before i start hacking mesa to do it.
> I want to access certain values that normally get stored in a profile file from run_stars_extra.f.
> So for my use case on a cluster we make too much information if we write out a profile file at the temporal resolution we need. So the plan is too work out what we actually want from the profile file, which is some_interesting_quantity(interesting_cell) ie we don't need the whole profile just a profile variable at 1 cell, at which point we can then insert that into the history file as an extra column and thus not write out the whole profile file to get 1 number.
> Thats all fine and works, and so far i've only needed variables that are accessible from the s pointer, but now i need some quantities that mesa per-processes before outputting to the profile file. In this case i want L_rad, which has a function get_L_rad in star/private/star_utils.f. I could hack run_stars to include that file and call get_L_rad myself, but them i'm mixing the public/private code up.
> It looks like all i need to do is create a public interface to getval_for_profiles.f? Any other ideas before i start?
> Want excitement?
> Manually upgrade your production database.
> When you want reliability, choose Perforce
> Perforce version control. Predictably reliable.
> mesa-users mailing list
> mesa-users at lists.sourceforge.net
More information about the Mesa-users