[mesa-users] mesa 4930 (ratios in chi2)

[João Faria]

Hi Bill

Great news on the new release! Looking forward to try it.

I did already take a quick look on the ratios in astero because I was 
trying to do the same thing myself. As it turns out, my implementation 
is remarkably similiar to yours (realizing this got me a little too 
self-confident, just now :)

I also considered the r02 ratio besides the r010, and probably we could 
discuss if the three point r01 and r10 versions are more appropriate for 
stars with less observed frequencies.
But what's been bugging me in including them in the chi2 is the 
covariance matrix issue. If we ask as input the sigmas for each ratio to 
weight in on the chi2, we are not considering correlations between 
adjacent points (which exist in the r010). My feeling is that this 
doesn't make a difference but I can't back that up with mathematical 
proof...

Anyway, here's my go at it, if it can still be useful.
Thanks for all the great work!

João Faria

On Wed 24 Apr 2013 05:22:26 AM WEST, Bill Paxton wrote:
>
> Hi,
>
>
> Last release introduced some large changes to diffusion, and now this 
> one has some bug fixes for that.
> Rich Townsend noticed that the "bump" in the RGB was being erased by 
> diffusion; now it is back.
> So if you are a user of diffusion, don't skip this revision.
>
>
> It also has changes to defaults discussed in the series of emails 
> initiated by Tug.
>
> prune_bad_cz_min_Hp_height now defaults to 0
> remove_mixing_glitches is still .true.,
> but all the cases are disabled except for okay_to_remove_mixing_singleton.
>
>
> Thanks to help from Victor Silva Aguirre, we now have an option to use 
> the Langer et al 1985 formula for
> semiconvective temperature gradient in addition to using their mixing 
> scheme. We still have the option
> to use the radiative temperature gradient of course. The new control 
> is "semiconvection_option".
>
> semiconvection_option = 'Langer_85 mixing; gradT = gradr'
> ! 'Langer_85 mixing; gradT = gradr' -- uses Langer scheme for mixing 
> but sets gradT = gradr
> ! 'Langer_85' -- this calculates special gradT as well as doing mixing.
>
>
> If you use the "approx21" net, you know it includes a fake electron 
> capture reaction from ni56 to cr56 that serves
> to mimic late stage neutronization. The ni56 electron capture now uses 
> the weaklib rate rather than the old Mazurek rate.
>
>
> I fixed a bug in my implementation of the Traxler et al option for 
> thermohaline so it finally works.
>
>
> We now have "hooks" for creating data for other pulsation codes 
> (other_get_pulsation_info and other_write_pulsation_info).
>
>
> I'm working with Dennis Stello to add an option in star/astero for 
> using ratios in the chi^2. It is still under development,
> but you are welcome to give it a try. There's even a new ratios plot 
> much like the existing echelle plot.
>
>
> I added separate control parameters to specify cache locations so that 
> they don't need to be in the mesa/data directory.
> They will still default to that location, but now we have the option 
> to have many users share the same main mesa
> source directory while having individual cache directories. The 
> controls are in star_job_defaults
>
> ! cache directories
>
> ! mesa uses caches to improve performance.
> ! the default location for these is in the mesa/data directory,
> ! but in some situations it is useful to keep the caches
> ! separately so, for example, multiple users can share the code
> ! and each can have a separate set of caches.
> ! '' means use default location for cache.
>
> eosDT_cache_dir = ''
> eosPT_cache_dir = ''
> ionization_cache_dir = ''
> kap_cache_dir = ''
> rates_cache_dir = ''
> weaklib_cache_dir = ''
>
> ! if you specify cache directories, use a separate one for each.
> ! e.g., something like this
> !eosDT_cache_dir = '/Users/bpaxton/mesa_caches/eosDT_cache'
> !eosPT_cache_dir = '/Users/bpaxton/mesa_caches/eosPT_cache'
> !ionization_cache_dir = '/Users/bpaxton/mesa_caches/ionization_cache'
> !kap_cache_dir = '/Users/bpaxton/mesa_caches/kap_cache'
> !rates_cache_dir = '/Users/bpaxton/mesa_caches/rates_cache'
> !weaklib_cache_dir = '/Users/bpaxton/mesa_caches/weaklib_cache'
> ! you much create the directories yourself -- mesa doesn't do it 
> automatically.
>
>
>
> Finally, if you are using ifort, you might want to change to FCopt = 
> -O1 instead of FCopt = -O2.
> I'll let Rich Townsend fill in the details on this one.
>
>
> Cheers,
> Bill
>
>
>
>
>
>
> ------------------------------------------------------------------------------
> Try New Relic Now & We'll Send You this Cool Shirt
> New Relic is the only SaaS-based application performance monitoring 
> service
> that delivers powerful full stack analytics. Optimize and monitor your
> browser, app, & servers with just a few lines of code. Try New Relic
> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
> _______________________________________________
> mesa-users mailing list
> mesa-users at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/mesa-users
-------------- next part --------------
A non-text attachment was scrubbed...
Name: star_adipls_support.f
Type: application/x-extension-fortran
Size: 107086 bytes
Desc: not available
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20130424/5d97f730/attachment.bin>