[ET Trac] #2616: Add NRPyEllipticET to the Einstein Toolkit

Leonardo Werneck trac-noreply at einsteintoolkit.org
Thu Sep 15 09:30:12 CDT 2022


#2616: Add NRPyEllipticET to the Einstein Toolkit

 Reporter: 
   Status: new
Milestone: 
  Version: 
     Type: enhancement
 Priority: major
Component: 

Comment (by Leonardo Werneck):

Hi all. Here are the updates made to `NRPyEllipticET` based on the reviewers' comments:

* A brief `README` file has been added;
* `uuGF`'s description has been updated in `interface.ccl`;
* `info_output_freq`'s comment has been updated in `param.ccl`;
* `N2`'s “forbidden value” has been removed, as it was unused;
* `initialize_offdiagonal_metric_to_zero` was indeed an old parameter and has been removed;
* All instances of `fprintf(stderr,...); exit();` have been replaced by either `CCTK_*ERROR` or `CCTK_*WARN`;
* We have added a new parameter called `residual_integration_radius`, with a proper description of what it does;
* Some of the algorithms are hard to describe using brief code comments. We recommend looking at the NRPy\+ Jupyter notebooks for the complete description of such algorithms;
* Thank you for pointing out that `CFL_FACTOR` was missing in `find_timestep.c`; this was actually a bug fix;
* Unfortunately the code in `wavespeed_gf_all_points.c` depends on `dt` \(the result of `find_timestep.c`\)  at every grid point and thus the two cannot be decoupled;
* The reason why the `rfm` parameters have the unused attribute is similar to the reason why parameters have the attribute, i.e., not every function that uses the `rfm` includes will use _all_ of the included parameters, which could lead to warnings;
* All instances of `__attribute__((unused))` have been replaced with `CCTK_ATTRIBUTE_UNUSED`;
* I have removed the empty files in `conformally_flat_BBH/rfm_files/` manually and am trying to figure out a way of automating this;
* `apply_bcs_curvilinear_radiation.c` and `set_Cparameters-SIMD.h` have been removed for now;
* `xx_to_Cart.c` and `set_Cparameters-nopointer.h` have _not_ been removed. Although currently unused by the thorn, they provide useful functionality that new users might want to use.

A general comment: the Jupyter notebook that generates `NRPyEllipticET` was very outdated. I am currently pushing on getting it up to speed, but at the moment we do not have a notebook that regenerates the latest version of the thorn.

--
Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2616/add-nrpyellipticet-to-the-einstein-toolkit
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.einsteintoolkit.org/pipermail/trac/attachments/20220915/2f7594c0/attachment-0001.html 


More information about the Trac mailing list