从集合中随机选出n项

import random #如果剩余选择空间个数乘以随机数大于等于剩余选择数，则选择该数，否则跳过 def sample(sets, n): res = [] m = t = 0 N = len(sets) for item in sets: u = random.random() if (N-t)*u >= (n-m): t += 1 else: res.append(item) m += 1 t += 1 if m >= n: break return res #this come from python cookbook def sample2(sets, n): N = len(sets) res = [] for samp in xrange(n, 0, -1): cumprob = 1.0 u = random.random() while u < cumprob: cumprob -= cumprob * samp / N N -= 1 res.append(sets[len(sets)-N-1]) return res print sample(xrange(100), 5) print sample2(xrange(100), 5)