Comment (by Cheng-Hsin Cheng):

Hi all, I finished checking NRPyEllipticET and I have some questions and comments about the code. To keep it brief I am leaving out the “conformally_flat_BBH” when referring to the file names in the src folder.

In NRPyEllipticET folder

• Please add a README in the Cactus format with the author, maintainer, license, and brief description.

• interface.ccl

  • uuGF: Please add a description for it. Although it might be familiar for someone who has used TwoPunctures, a new user might not know what is the purpose of it.

• param.ccl:

  • info_output_freq: I think a better description would be e.g. "Print progress of relaxation time evolution every info_output_freq iterations"
  • N2: allows a "forbidden value" of -1 to make sure it is set explicitly in the parfile, but the default is set to 16 rather than -1. Was the value of -1 to be removed (since N0 and N1 don't have a forbidden value like N2)?
  • initialize_offdiagonal_metric_to_zero: This looks to be unused in the code. Should it be removed from the param.ccl?

C source files

• Hyberbolic_Relaxation.c

  • Using CCTK_ERROR (or CCTK_INFO) to error out might be preferable to using printf and calling exit(1).
  • Instead of hard-coding the value for integration_radius, could you add it as a thorn parameter?

• conformally_flat_BBH_driver_bcstruct.c

  • Line 23: It took me a while to understand what the comment meant. I think it might be clearer to rewrite as "in the case that a boundary point is..."
  • Lines 59-61 are commented out. I don't know if set_bcstruct handles grids without the outer boundary, but if it doesn't, should this be left uncommented?

• set_bcstruct.c

  • I really like how at the top the

    --
    Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2616/add-nrpyellipticet-to-the-einstein-toolkit