[Users] Problem with the latest version of ET and example Bondi par file

Petra Suková psukova at cft.edu.pl
Mon Aug 17 05:32:33 CDT 2015


Hello,

I am a new user of ET and during my initial tests with the latest version of 
ET and the example parameter file for Bondi inflow I encountered an error, 
which terminates the run.
I have discovered, that the problem is connected with the new variable 
defined in thorn EinsteinAnalysis/Hydro_Analysis in this way:

CCTK_REAL Hydro_Analysis_core_rho_centroid_gf[4] type = GF Timelevels = 3 
tags='checkpoint="no" tensortypealias="4u"'

This variable was not in the previous version of ET.
During execution of the file interpolate.c from 
CactusNumerical/RotatingSymmetry90, the check of the number of variables and 
vector length for this variable does not pass:

else if (CCTK_EQUALS (tensortypealias, "4u")
               || CCTK_EQUALS (tensortypealias, "4d"))
    {
      assert ((numvars == (DIM+1)*vectorlength));
    }

because numvars = 4 and vectorlength=4.

I fixed the problem in the way that I added another condition (similarly to 
the case of usual three-vector u or d, which is just above)

else if (CCTK_EQUALS (tensortypealias, "4u")
               || CCTK_EQUALS (tensortypealias, "4d"))
    {
      assert ((numvars == (DIM+1)*vectorlength)||
              (numvars == (DIM+1) && vectorlength == (DIM+1)));
     }

It works for me, however, I am not sure, if it will not cause problems in 
some other cases. Maybe the quantity Hydro_Analysis_core_rho_centroid_gf 
should be defined rather as a group of four variables and not as a 
four-vector.

Best regards,
Petra 


---
Tato zpráva byla zkontrolována na viry programem Avast Antivirus.
https://www.avast.com/antivirus



More information about the Users mailing list