Switch from pypy to CPython for sections of code

Question

Is there any feature in pypy that allows for a standard CPython 2.7 interpreter to be run for a designated section of code? I have a function that has pandas code within it (its a performance intensive function, benefiting greatly by pandas), all references to pandas are contained within that function.

Obviously pypy can't interpret pandas code due to pandas' C-bound nature. Is there a way that I can "switch over" to a standard interpreter just for this function? The codebase as a whole greatly benefits from a pypy interpreter.

@user2357112 Is there a way I can use the `multiprocessing.Process(target=fn)` interface to accomplish this? I'd rather not completely shell the entire module out, just the function itself. — jab, Apr 20 '17 at 20:59

score 0 · Answer 1 · answered Apr 20 '17 at 22:21

0

Try running the pandas code with PyPy. Compatibility is rather good these days.

answered Apr 20 '17 at 22:21

Ronan Lamy

577
3
9

How would you install a package like pandas under pypy? Installing via pip complains as soon as you get to c-extenisons. – jab Apr 20 '17 at 22:36
Are you on the latest version of pypy (5.7.1)? I don't think it works with previous releases. – Ronan Lamy Apr 20 '17 at 23:35

Switch from pypy to CPython for sections of code

1 Answers1