Wrong math result with calling C extension in Python

Question

I am calling the pow() function of C's math library. My installation is Python3.5 on Windows 10 but I tried the same program in Python2.7 with same result. The return from the function is not as expected. Not sure what's happening. The code is given below. The result I get is 1.0.

from ctypes import *

libc = cdll.msvcrt
libc.pow.argtypes = [c_double, c_int]
libc.pow.restype = c_double
libc.pow(2.3, 2)

Will_Panda · Answer 1 · 2016-10-10T06:50:31.600

1

There is a type mismatch. Use c_double for power as well. Like this. It should work.

libc.pow.argtypes = [c_double, c_double]

edited Oct 10 '16 at 06:50

answered Oct 10 '16 at 06:46

Will_Panda

534
10
26

1

IIRC, your "also" addendum is superfluous. – Oct 10 '16 at 06:48
@Hurkyl Thanks. Edited. – Will_Panda Oct 10 '16 at 06:50

score 1 · Accepted Answer · answered Oct 10 '16 at 06:49

1

Double check your types, I'm pretty sure the prototype for pow is

double pow(double a, double b);

So changes this line

libc.pow.argtypes = [c_double, c_int]

To

libc.pow.argtypes = [c_double, c_double]

Whenever you call c functions from python always make sure the types you are inputting from python are correct, don't rely on casting like you would in c/c++

answered Oct 10 '16 at 06:49

Javonne Martin

76
4

That's correct. Looks like an interface bug. It should catch this as wrong argument type instead of returning wrong result. – coder.in.me Oct 10 '16 at 07:05

score 0 · Answer 3 · answered Oct 10 '16 at 06:49

0

pow in C libray takes double, double as input arguments. That's why.

from ctypes import *

libc = cdll.msvcrt
libc.pow.argtypes = [c_double, c_double]
libc.pow.restype = c_double
libc.pow(2.3, 2.0)

answered Oct 10 '16 at 06:49

stefan_ah

54
4

Wrong math result with calling C extension in Python

3 Answers3