在递归函数中将大列表作为参数传递时的性能? [英] Performance when passing huge list as argument in recursive function?
问题描述
我正在使用 Python 并且我有一个递归函数,它将一个巨大的列表作为参数之一:
I am using Python and I have a recursive function that takes a huge list as one of the arguments:
# Current implementation
def MyFunction(arg1, arg2, my_huge_list)
...
...
MyFunction(new_arg1, new_arg2, my_huge_list)
如上所示,MyFunction
使用相同 列表my_huge_list
递归调用;与其他参数不同,这不会改变.而且,这个列表是巨大的.我的一个朋友建议我可以把 my_huge_list
当作一个全局变量来提高性能,否则这个庞大的列表可能会在每次迭代中一遍又一遍地复制.
As you can see above, MyFunction
is called recursively using the same list my_huge_list
; this doesn't change, unlike the other arguments. And, again, this list is huge. A friend of mine suggested that I could treat my_huge_list
as a global variable to improve the performance, as otherwise this huge list may be copied over and over in every iteration.
# Friend's suggestion
MyHugeList=[a,b,c, ...and many many other elements... ]
def MyFunction(arg1, arg2)
global MyHugeList
...
...
MyFunction(new_arg1, new_arg2)
使用如上所示的全局变量是否比原始版本提高了算法的性能?我的程序运行了数周,所以从长远来看,即使是轻微的改进也可能是有价值的.
Does using a global variable as shown above improve the performance of the algorithm over the original version? My program runs for weeks, so even a slight improvement may be valuable in the long run.
推荐答案
列表将通过引用传递,因此不再需要传输 1 项列表与 100000 项列表:
The list will be passed by reference, so it doesn't take any longer to transfer a 1-item list vs. a 100000 item list:
def null(x): return x
longlist = range(100000)
shortlist = range(1)
longerlist = range(1000000)
%timeit null(shortlist)
10000000 loops, best of 3: 124 ns per loop
%timeit null(longlist)
10000000 loops, best of 3: 137 ns per loop
%timeit null(longerlist)
10000000 loops, best of 3: 125 ns per loop
较长的列表中有 100k 和 1M 的条目,但与较短的列表相比,作为参数传递的时间并不长.
The longer lists have 100k and 1M entries in them, and yet don't take significantly long to pass as arguments than shorter lists.
可能还有其他方法可以提高性能;这可能不是其中之一.
there may be other ways to improve performance; this probably isn't one of them.
这篇关于在递归函数中将大列表作为参数传递时的性能?的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!