[mesa-users] mesa problems with the newly added OP opacity tables for a05 and a09 -- the old OP gs98 tables are fine

[Bill Paxton]

On Oct 17, 2013, at 5:52 PM, Aaron Dotter wrote:

> After some time and effort it seems that the OP server is producing tables with NaNs in them.  The solution is to not support new OP tables (from their web server, in the OPAL type1 format) until and unless OP fixes the problem.


Hi,

Thanks to Aaron for finding the source of this problem.   And my apologies for saying it was a problem with the OPAL software! 

Keep in mind that this is only a problem with the new OP data files --- the existing mesa OP_gs98 files are fine since they are built using older data from OP that isn't polluted by NaNs.

To fill in a few more details in case you are interested, I only find 6 NaNs per new OP input file and (I think) they are only at the extreme low density edge of the tables.  Out of curiosity, I just did a global replace of the NaNs by -1's and reran the preprocessor.  It seemed to do okay.  If you are so rash that you'd want to use these hacked OP opacities as a temporary solution until we get repaired OP input data, please let me know.

-Bill