Closed jeanbraun closed 2 years ago
Found a bug in NA_MPI.f where the dimension of an array was not consistent across routines. This seems to have fixed the problem encountered on some clusters when the parallel optimisation/inversion was used (using MPI)
Found a bug in NA_MPI.f where the dimension of an array was not consistent across routines. This seems to have fixed the problem encountered on some clusters when the parallel optimisation/inversion was used (using MPI)