search

will-henney / proplyd-cloudy

Ionization/dynamic models of proplyds
http://deprecated.github.com/proplyd-cloudy/
2 stars 0 forks source link

Some models fail on the linux machines #12

Open will-henney opened 12 years ago

will-henney commented 12 years ago

This seems to be due to problems with the ".in" files. Sometimes a comment line about the stellar atmosphere gets wrapped onto two lines, which causes Cloudy to get confused.

I have pushed change 5cc8f8d05fef1940fd5b40c0729b7623dff45c7b to the emission branch, which should fix this, but Nahiely needs to test it

nahieflores commented 12 years ago

I test it and it works very well but we have two more problems:

1.- The first one is with model_profile.py.

when more than two angles are used, the line fluxes are negatives. When only 0 and 90 degrees are used the line fluxes are mostly the same than those calculated in the mac.

2.- The second one is that some models still fail.

Something is wrong with Cloudy but another time just in linux machines. The problem is that in the second iteration the modl stops with: 079 1.0000e+00 2.1368e-02 1.0000e+00 9.9575e-01 1.0335e+00 1.00095 079 1.0000e+00 2.1368e-02 1.0000e+00 9.9575e-01 1.0335e+00 1.00095 Floating exception (core dumped)

Just the same model runs and finish without problems in mac machine.

nahieflores commented 12 years ago

apparently all is fine with the radians conversion, I prove that and is correct, but I supposed that is something the angles ...

will-henney commented 12 years ago

Do you still have this core dump problem on the linux machines?