Currently, the csu_kdp and csu_fhc sub-modules use f2py to compile fortran subroutines, which has vastly sped up the KDP routines and moderately sped up the FHC routines. However, this can lead to compiler issues, especially on Windows machines. We should convert the problematic code to Cython to get hopefully similar performance but fewer compiling issues.
Currently, the csu_kdp and csu_fhc sub-modules use f2py to compile fortran subroutines, which has vastly sped up the KDP routines and moderately sped up the FHC routines. However, this can lead to compiler issues, especially on Windows machines. We should convert the problematic code to Cython to get hopefully similar performance but fewer compiling issues.