Several test suite updates (astrobear changeset 701 + 702)

I finished fixing the RadiativeInstability files so that they now only produce 1 chombo at the correct computational time. Then, I updated the reference chombos that are in the /data/tests directory on Clover. I also noticed some errors while doing the post processing on this problem. I fixed the bear2fix.data files in astrobear and the clover directory.

Another thing that I noticed while doing the chombo comparison for RadiativeInstability was a relative error of infinity! I realized that the y-momentum is zero for this problem. Therefore, I rewrote bear2fix's i_compare.f90 to avoid this divide by zero issue. It now replaces an infinite error with one of three things:

1) zero error if the new chombo also has zero momentum component

2) an error equal to the new chombo's momentum component if the new momentum component is very small (which would lead to an error within the 'successful' tolerance)

3) an error relative to the average of the reference chombo's momentum components (which could lead to a 'success' or 'fail')

Lastly, the OrbitingParticles problem was added to the test suite. Producing a useful image was the most difficult part here. I "tricked" bear2fix into plotting the sink particle potential. To do this I told bear2fix to plot By. The particle potential is stored in the same location in the q array that By would be if there were magnetic fields.

All the relevant aforementioned changes were pushed to scrambler_devel.

Comments

No comments.