CDEFGAB 1010110

挖了太多坑,一点点填回来

使用装饰器为Python函数做缓存

python, 设计模式

写了个函数作为装饰器为Fibonacci数列做缓存,做了一下性能比较。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
#!/usr/bin/python
# coding: UTF-8

import datetime
now = datetime.datetime.now
from functools import wraps

def cache(func):
    caches = {}
    @wraps(func)
    def wrap(*args):
        if args not in caches:
            caches[args] = func(*args)
        return caches[args]
    return wrap

def fib(num):
    if num < 2:
        return 1
    return fib(num-1) + fib(num-2)

fib_with_cache = cache(fib)

start = now()
for i in range(10):
    fib(30)
end = now()
print "Fib without cache costs: %r" % (end - start)

start = now()
for i in range(10):
    fib_with_cache(30)
end = now()
print "Fib with cache costs: %r" % (end - start)

输出结果:
Fib without cache costs: datetime.timedelta(0, 6, 219000)
Fib with cache costs: datetime.timedelta(0, 0, 672000)

当然,正确的使用Fibonacci数列的方法应该是在函数体内使用循环,而不是使用递归。多层递归很容易对性能造成非常严重的影响。

Have a nice day!