On Nov 22, 9:21 am, n00m <n...@narod.ru> wrote: > Any comment: > > class Vector: > def __init__(self, x, y): > self.x = x > self.y = y > def __cmp__(self, v): > if self.x < v.x and self.y > v.y: > return -1 > return 0 > > def v_cmp(v1, v2): > if v1.x < v2.x and v1.y > v2.y: > return -1 > return 0 > > from random import randint > from time import time > > a = [] > for i in range(200000): > a += [Vector(randint(0, 500000), randint(0, 500000))] > b = a[:] > c = a[:] > > print 'Sorting...' > > t = time() > b.sort(cmp=v_cmp) > print time() - t > > t = time() > c.sort() > print time() - t > > print b == c > > >>> ===================================== RESTART ====== > > Sorting... > 0.906000137329 > 6.57799983025 > True
Do you get the same magnitude difference if you make Vector a new- style class? (I.e., use "class Vector(object)" instead of "class Vector ()".) Mark -- http://mail.python.org/mailman/listinfo/python-list
