Numpy probably. > Naturally far more efficient Most data science languages are interpreted with a few higher-level routines. Same with J, for instance - it ha

Numpy probably. > Naturally far more efficient Most data science languages are interpreted with a few higher-level routines. Same with J, for instance - it has adequate performance as well.