Hello Andr?,
The problem you quote is with the integrator Runge Kutta of order 8 that you chose for the solution of the orbitals equations of motion. The energy and occupations seem to be undefined (NaN). In the future, please also post the input file "MCTDHB.inp" to make the analysis easier.
I think you specified the wrong file to read from, right? Anyways, I think we resolved this problem by renaming your *orbs.dat and *coefs.dat files?
I will take your post as a request to revise the input handling, especially, I would like to deprecate the restart from ASCII files with GUESS='DATA' in the near future.
Could you specify, what fixed the issue?
Regards,
Axel