Troubleshooting

Most common error messages

Error: “ WARNING : first cell is in diffusion approximation zone. Increase spatial grid resolution”

It is an optical depth problem, which require a proper sampling of the grid. The solution is to increase n_rad_in which is the number of subdivision of the first cell The solution is to increase n_rad_in. You may need to try a few values empirically to find one that works. Around 30 or 40 is recommended, ie something like that: 150 60 1 30 n_rad, nz (or n_theta), n_az, n_rad_in to define the grid geometry.

When I try to run mcfost, I get the error message “Exec format error. Binary file not executable”. What’s going on?

You’re trying to run a Linux version of MCFOST on a Mac, or vice versa, or some other incorrect operating system combination. You probably downloaded or copied the wrong file by accident, and should just obtain a new copy from the download site.

I get random segmentation faults

Make sure you have increased the stacksize:

$ export OMP_STACKSIZE=512M
$ ulimit -s unlimited

Note

There is a hard limit at 64kB for the stacksize on MacOS. This will prevent post-processing of large phantom dumps (more than 2 millions particles). The solution is either to move to linux, or to re-compile the code with gfortran (which relies less on the stack than ifort), but this will result in some speed penalties.

Geometry

The disc orientation is correct, but the red and blue shifted sides are flipped.

The easiest way to change the sign of the rotation without changing the orientation is to change the inclination i to 180-i and the PA to PA+180, and if the model is 3d, the azimuth (or planet_azimuth) should be changed at 180-azimuth.

The PA and red/blue sides are correct, but the disk is inclined the wrong way.

The easiest fix n is to change the inclination i to i-180 and the PA to PA+180, and if the model is 3d, the azimuth (or planet_azimuth) should be changed at 180-azimuth.