[ET Trac] #2823: Include TOVola: An enhanced TOV solver with Tabulated EOS support

[Maxwell Rizzo]

#2823: Include TOVola: An enhanced TOV solver with Tabulated EOS support

 Reporter: David Boyer
   Status: open
Milestone: ET_2024_11
  Version: 
     Type: enhancement
 Priority: major
Component: EinsteinToolkit thorn

Comment (by Maxwell Rizzo):

ReadMe/Documentation/Parfiles

1. In Output, hdf5 data is also supported through CarpetIOHDF5/IOHDF5 with Kuibit.
2. All 3 par files \(Piecewise, Simple, Tabulated\) have `IO::out_fileinfo="none" `and `CarpetIOASCII::compact_format = no` , both options Kuibit does not like. I recommend commenting out the first option and changing the second to yes to enable Kuibit support for ASCII files here.
3. In the documentation.tex, in the section “Solving the ODE system” multiple different thorns are mentioned which gives off the impression that these thorns are critical in the operation of TOVola which I don’t think is the case. In the spirit of generality to other gravity/hydro solvers than what is used in the example parfiles this language could be refined. Some language dedicated towards how TOVola could be generalized with other solvers not used in the example parfiles might be helpful for users.

Param.ccl

1. `TOVola_error_limit` could use description of what it actually is doing.
2. `TOVola_Max_Interpolation_Stencil` comment could be refined, it has the same comment as the previous parameter.

General Driver

1. Since only the stencil size and the amount of layers in a simple polytrope are checked in the `TOVola_Parameter_Checker`, the inline comments and cactus output could maybe be refined to reflect that, both in the main driver and the schedule.ccl. The current output and comments may lead users to believe more parameters are being checked than just the two.
2. Looks like the logic in `switch (TOVola_TOV_Populate_Timelevels)` for case 3 and case 2 could be combined, code looks entirely identical except for one variable.

Overall, I compiled TOVola into the ETK 2024\_05/ LevLandau release with no changes. I compiled the documentation.tex and the corresponding HTML. Additionally, I ran the two provided tests, simple and piecewise polytrope TOVola stars successfully. I visualized the initial output with Kuibit after the above changes to the parfiles were made. I went through the main driver routine and its supporting files, and can say the code is well documented and formatted. I apologize for my C knowledge with Cactus being slightly inadequate to say more, I will have to defer to Roland’s comments above regarding memory/Cactus infrastructure. I found TOVola to meet my standards and the quality/formal criteria listed in the Einstein Toolkit documentation for new components.

--
Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2823/include-tovola-an-enhanced-tov-solver-with
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.einsteintoolkit.org/pipermail/trac/attachments/20241022/2c64867f/attachment.htm>