Cython使用及pycharm

文件总览

.
+-- main.py
+-- fib
|   +-- __init__.py
|   +-- cfib.c
|   +-- cfib.h
|   +-- fib.pyx
|   +-- setup.py

说明：

1. main.py主程序
2. __init__.py是模块引用文件
3. cfib.c用c写的fib函数
4. cfib.h是cfib的头文件，供fib.pyx参照
5. fib.pyx是Cython文件
6. setup.py是用来编译Cython的

代码

cfib.c：

#include "cfib.h"

unsigned long fib(unsigned long n) {
    unsigned long a=0, b=1, i, tmp;
    for (i=0;i<n;++i) {
        tmp = a; a = a + b; b = tmp;
    }
    return a;
}

cfib.h：

#ifndef __CFIB_H__
#define __CFIB_H__

unsigned long fib(unsigned long n);

#endif

cfib.c和cfib.h就是标准的c程序写法，就不赘述，主要拿来比performance的

fix.pyx：定义引用的c函数

cdef extern from "cfib.h":
    unsigned long _fib "fib"(unsigned long n)

def fib_c(n):
    '''Returns the nth Fibonacci number.'''
    return _fib(n)

def fib_cython(n):
    '''Returns the nth Fibonacci number.'''
    a, b = 0, 1
    for i in range(n):
        a, b = a + b, a
    return a

def fib_cython_optimized(unsigned long n):
    '''Returns the nth Fibonacci number.'''
    cdef unsigned long a=0, b=1, i
    for i in range(n):
        a, b = a + b, a
    return a

__init__.py：

from .fib import *

setup.py：

from distutils.core import setup, Extension
from Cython.Build import cythonize

setup(ext_modules = cythonize(Extension(name="fib", sources=["cfib.c", "fib.pyx"])))

Pycharm

打开PyCharm，到settings => Tools => External Tools，然後点+，新增一个工具，设置如下：

Porgram：$ModuleSdkPath$即python可执行文件路径
Arguments：$FilePath$ build_ext --inplace
Working directory：$FileDir$

这样基本上就大功告成了，然后回到Project对setup.py点右键，选External Tools，选CythonBuild就可以看到开始跑了

如果设置对的话，就可以看到编译成功的信息

当然为了避免每次编译pyx文件的时候都点击的话，可以直接设置该tool的快捷键。

最后就可以来跑跑看啦，main.py如下：

def fib_python(n):
    '''Returns the nth Fibonacci number.'''
    a, b = 0, 1
    for i in range(n):
        a, b = a + b, a
    return a

if __name__ == '__main__':
    print("##### check result #####")
    import fib
    print("fib(47) in python:", fib_python(47))
    print("fib.fib_c(47):", fib.fib_c(47))
    print("fib.fib_cython(47):", fib.fib_cython(47))
    print("fib.fib_cython_optimized(47):", fib.fib_cython_optimized(47))

    print("\n##### performace benchmark #####")
    import timeit
    python_setup = "from __main__ import fib_python"
    cython_setup = "import fib"
    print("Python code: ", timeit.timeit('fib_python(47)', setup=python_setup), "seconds")
    print("Cython code: ", timeit.timeit('fib.fib_cython(47)', setup=cython_setup), "seconds")
    print("Optimized Cython code: ", timeit.timeit('fib.fib_cython_optimized(47)', setup=cython_setup), "seconds")
    print("C code: ", timeit.timeit('fib.fib_c(47)', setup=cython_setup), "seconds")

输出结果：

C:\Users\*****\Miniconda3\envs\idp\python.exe D:/Dropbox/Data/programming/Python/Cython/1_fib/main.py
##### check result #####
fib(47) in python: 2971215073
fib.fib_c(47): 2971215073
fib.fib_cython(47): 2971215073
fib.fib_cython_optimized(47): 2971215073

##### performace benchmark #####
Python code:  2.9352622053858113 seconds
Cython code:  1.7331176511158422 seconds
Optimized Cython code:  0.14643933094340067 seconds
C code:  0.11884286952119272 seconds

Process finished with exit code 0

参考

Pycharm and Cython