On Fri, Nov 7, 2008 at 6:16 PM, Alan Gauld <[EMAIL PROTECTED]> wrote:
> True, but its pretty rare that timing issues are a reason for me to choose a > data structure I would guess you commonly choose a dict or set over a list when you need fast tests for membership. Failure to choose dict when appropriate is certainly a common cause of performance problems. > But the geneal point is a good specific example (and I was struggling to > think of one!) where you might choose a non standard list over the vanilla > version. The array module is another case where performance is improved over > the standard lists. The standard lib also includes collections.deque (O(1) insertion and deletion at both ends) and heapq (binary priority queue). Third party implementations of b-tree, avltree and trie are available which have better performance than list and dict for some usage. Kent _______________________________________________ Tutor maillist - Tutor@python.org http://mail.python.org/mailman/listinfo/tutor
