在Python中获取日期时间数组的唯一值

Question

我在Python里有一个叫做 dates 的数组，里面的内容是：

[datetime.datetime(2012, 1, 11, 17, 24, 12, 676000), datetime.datetime(2012, 2, 3, 11, 25, 17, 73000), datetime.datetime(2012, 2, 3, 14, 9, 23, 699000), datetime.datetime(2012, 2, 4, 9, 15, 26, 644000), datetime.datetime(2012, 2, 4, 17, 14, 36, 65000), datetime.datetime(2012, 2, 5, 6, 18, 31, 139000), datetime.datetime(2012, 2, 5, 14, 55, 28, 62000), datetime.datetime(2012, 2, 5, 18, 28, 59, 379000), datetime.datetime(2012, 2, 6, 12, 24, 21, 768000), datetime.datetime(2012, 2, 6, 17, 32, 46, 675000), datetime.datetime(2012, 2, 14, 11, 33, 6, 74000), datetime.datetime(2012, 2, 14, 11, 36, 48, 11000), datetime.datetime(2012, 2, 16, 8, 54, 14, 175000), datetime.datetime(2012, 2, 16, 18, 33, 9, 200000), datetime.datetime(2012, 2, 20, 8, 41, 2, 550000), datetime.datetime(2012, 2, 20, 9, 4, 37, 446000), datetime.datetime(2012, 2, 20, 10, 10, 42, 950000), datetime.datetime(2012, 2, 20, 21, 21, 21, 986000), datetime.datetime(2012, 2, 21, 9, 1, 8, 429000), datetime.datetime(2012, 2, 21, 12, 5, 20, 475000), datetime.datetime(2012, 2, 21, 13, 23, 25, 281000), datetime.datetime(2012, 2, 21, 15, 4, 29, 366000), datetime.datetime(2012, 2, 21, 15, 12, 21, 729000), datetime.datetime(2012, 2, 21, 15, 29, 10, 723000), datetime.datetime(2012, 2, 21, 18, 10, 24, 822000), datetime.datetime(2012, 2, 22, 10, 42, 11, 689000), datetime.datetime(2012, 2, 22, 13, 10, 1, 309000), datetime.datetime(2012, 2, 22, 20, 28, 34, 260000), datetime.datetime(2012, 2, 27, 17, 53, 19, 225000), datetime.datetime(2012, 2, 28, 8, 13, 57, 139000), datetime.datetime(2012, 3, 2, 7, 55, 11, 505000), datetime.datetime(2012, 3, 2, 21, 6, 35, 270000), datetime.datetime(2012, 3, 5, 8, 10, 47, 76000), datetime.datetime(2012, 3, 5, 9, 15, 15, 448000), datetime.datetime(2012, 3, 7, 18, 15, 35, 401000), datetime.datetime(2012, 3, 15, 8, 6, 56, 968000), datetime.datetime(2012, 3, 16, 15, 34, 10, 59000), datetime.datetime(2012, 3, 20, 18, 19, 13, 687000), datetime.datetime(2012, 3, 22, 8, 50, 28, 983000), datetime.datetime(2012, 3, 23, 8, 26, 5, 468000), datetime.datetime(2012, 3, 27, 7, 50, 14, 474000), datetime.datetime(2012, 3, 27, 15, 14, 35, 59000), datetime.datetime(2012, 4, 5, 7, 23, 1, 374000), datetime.datetime(2012, 4, 6, 13, 8, 59, 578000), datetime.datetime(2012, 4, 6, 13, 34, 24, 843000), datetime.datetime(2012, 4, 6, 15, 35, 40, 538000), datetime.datetime(2012, 4, 10, 7, 0, 37, 455000), datetime.datetime(2012, 4, 10, 7, 12, 37, 199000), datetime.datetime(2012, 4, 10, 7, 39, 16, 366000), datetime.datetime(2012, 4, 10, 7, 55, 51, 228000), datetime.datetime(2012, 4, 11, 7, 53, 31, 699000), datetime.datetime(2012, 4, 11, 15, 32, 21, 582000), datetime.datetime(2012, 4, 13, 10, 22, 4, 673000), datetime.datetime(2012, 4, 16, 7, 17, 20, 578000), datetime.datetime(2012, 11, 29, 16, 5, 21, 53000), datetime.datetime(2012, 11, 29, 16, 6, 15, 244000), datetime.datetime(2013, 1, 25, 9, 45, 48, 921000), datetime.datetime(2013, 2, 4, 18, 1, 1, 418000), datetime.datetime(2013, 2, 5, 6, 14, 55, 728000), datetime.datetime(2013, 2, 5, 17, 2, 11, 959000), datetime.datetime(2013, 2, 7, 6, 4, 8, 629000), datetime.datetime(2013, 2, 7, 18, 6, 47, 247000), datetime.datetime(2013, 2, 8, 5, 36, 55, 702000), datetime.datetime(2013, 2, 8, 8, 51, 46, 261000), datetime.datetime(2013, 2, 12, 5, 56, 37, 233000), datetime.datetime(2013, 2, 12, 16, 6, 25, 126000), datetime.datetime(2013, 2, 13, 7, 45, 33, 448000), datetime.datetime(2013, 2, 13, 10, 43, 15, 749000), datetime.datetime(2013, 2, 14, 6, 10, 27, 562000), datetime.datetime(2013, 2, 14, 16, 44, 45, 469000), datetime.datetime(2013, 2, 15, 6, 3, 12, 787000), datetime.datetime(2013, 2, 15, 14, 8, 40, 281000), datetime.datetime(2013, 2, 17, 11, 46, 41, 983000), datetime.datetime(2013, 2, 20, 15, 32, 52, 455000), datetime.datetime(2013, 2, 21, 16, 0, 40, 165000), datetime.datetime(2013, 2, 22, 9, 12, 55, 32000), datetime.datetime(2013, 2, 22, 15, 11, 45, 979000), datetime.datetime(2013, 2, 25, 6, 52, 49, 991000), datetime.datetime(2013, 2, 25, 8, 52, 8, 947000), datetime.datetime(2013, 2, 25, 9, 27, 7, 716000), datetime.datetime(2013, 2, 25, 9, 33, 21, 121000), datetime.datetime(2013, 2, 26, 7, 15, 0, 135000), datetime.datetime(2013, 2, 26, 16, 15, 39, 693000), datetime.datetime(2013, 2, 27, 6, 33, 23, 745000), datetime.datetime(2013, 2, 27, 17, 28, 47, 793000), datetime.datetime(2013, 2, 28, 5, 43, 32, 479000), datetime.datetime(2013, 2, 28, 17, 22, 15, 510000), datetime.datetime(2013, 3, 1, 6, 54, 21, 676000), datetime.datetime(2013, 3, 1, 15, 47, 19, 912000), datetime.datetime(2013, 3, 4, 17, 39, 55, 809000), datetime.datetime(2013, 3, 5, 6, 40, 35, 101000), datetime.datetime(2013, 3, 5, 17, 5, 4, 324000), datetime.datetime(2013, 3, 6, 6, 39, 42, 235000), datetime.datetime(2013, 3, 6, 16, 6, 29, 410000), datetime.datetime(2013, 3, 7, 6, 32, 56, 197000), datetime.datetime(2013, 3, 7, 17, 31, 39, 249000), datetime.datetime(2013, 3, 8, 6, 56, 44, 369000), datetime.datetime(2013, 3, 11, 7, 17, 20, 748000), datetime.datetime(2013, 3, 11, 17, 27, 43, 102000), datetime.datetime(2013, 3, 12, 7, 10, 24, 751000), datetime.datetime(2013, 3, 12, 10, 23, 44, 759000), datetime.datetime(2013, 3, 12, 15, 42, 20, 461000), datetime.datetime(2013, 3, 13, 7, 12, 40, 494000), datetime.datetime(2013, 3, 13, 12, 7, 24, 986000), datetime.datetime(2013, 3, 14, 6, 52, 10, 779000), datetime.datetime(2013, 3, 14, 16, 39, 12, 776000), datetime.datetime(2013, 3, 15, 7, 4, 26, 454000), datetime.datetime(2013, 3, 15, 16, 40, 37, 98000), datetime.datetime(2013, 3, 18, 6, 53, 56, 937000), datetime.datetime(2013, 3, 18, 16, 53, 26, 914000), datetime.datetime(2013, 3, 19, 6, 34, 41, 813000), datetime.datetime(2013, 3, 19, 17, 19, 59, 721000), datetime.datetime(2013, 3, 20, 6, 57, 37, 141000), datetime.datetime(2013, 3, 20, 15, 15, 43, 458000), datetime.datetime(2013, 3, 21, 15, 36, 12, 949000), datetime.datetime(2013, 3, 22, 6, 57, 21, 973000), datetime.datetime(2013, 3, 22, 15, 36, 14, 388000), datetime.datetime(2013, 3, 25, 7, 0, 43, 602000), datetime.datetime(2013, 3, 25, 18, 27, 0, 693000), datetime.datetime(2013, 3, 26, 17, 20, 48, 194000), datetime.datetime(2013, 3, 27, 7, 11, 17, 665000), datetime.datetime(2013, 3, 27, 18, 27, 41, 894000), datetime.datetime(2013, 3, 28, 7, 2, 8, 624000), datetime.datetime(2013, 3, 28, 11, 12, 22, 53000), datetime.datetime(2013, 4, 3, 5, 45, 23, 995000), datetime.datetime(2013, 4, 4, 6, 5, 39, 243000), datetime.datetime(2013, 4, 8, 6, 4, 34, 667000), datetime.datetime(2013, 4, 8, 17, 6, 8, 718000), datetime.datetime(2013, 4, 9, 6, 2, 32, 813000), datetime.datetime(2013, 4, 9, 15, 16, 46, 622000), datetime.datetime(2013, 4, 10, 5, 26, 16, 694000), datetime.datetime(2013, 4, 10, 18, 50, 54, 809000), datetime.datetime(2013, 4, 11, 15, 12, 29, 376000), datetime.datetime(2013, 4, 12, 6, 9, 38, 925000), datetime.datetime(2013, 4, 12, 14, 42, 32, 607000), datetime.datetime(2013, 4, 15, 10, 0, 59, 995000), datetime.datetime(2013, 4, 15, 10, 11, 42, 16000), datetime.datetime(2013, 4, 16, 6, 8, 3, 838000), datetime.datetime(2013, 4, 16, 15, 27, 35, 147000), datetime.datetime(2013, 4, 17, 6, 4, 44, 272000), datetime.datetime(2013, 4, 17, 15, 23, 0, 924000), datetime.datetime(2013, 4, 18, 6, 9, 55, 454000), datetime.datetime(2013, 4, 18, 15, 5, 43, 601000), datetime.datetime(2013, 4, 19, 6, 0, 38, 132000), datetime.datetime(2013, 4, 19, 16, 35, 26, 14000), datetime.datetime(2013, 4, 19, 17, 44, 17, 116000), datetime.datetime(2013, 4, 19, 17, 51, 48, 43000), datetime.datetime(2013, 4, 19, 17, 54, 30, 44000), datetime.datetime(2013, 4, 21, 14, 58, 56, 363000), datetime.datetime(2013, 4, 21, 15, 8, 11, 276000), datetime.datetime(2013, 4, 23, 6, 24, 57, 124000), datetime.datetime(2013, 4, 23, 15, 44, 30, 503000), datetime.datetime(2013, 4, 25, 6, 13, 9, 269000), datetime.datetime(2013, 4, 25, 15, 41, 11, 370000), datetime.datetime(2013, 4, 26, 6, 2, 17, 877000), datetime.datetime(2013, 4, 27, 16, 17, 34, 97000), datetime.datetime(2013, 4, 27, 18, 20, 57, 975000), datetime.datetime(2013, 4, 29, 10, 17, 41, 746000), datetime.datetime(2013, 4, 29, 16, 45, 18, 65000), datetime.datetime(2013, 4, 30, 6, 13, 2, 333000), datetime.datetime(2013, 4, 30, 15, 3, 22, 343000), datetime.datetime(2013, 5, 1, 7, 22, 40, 401000), datetime.datetime(2013, 5, 1, 11, 16, 38, 525000), datetime.datetime(2013, 5, 2, 6, 7, 7, 749000), datetime.datetime(2013, 5, 3, 12, 48, 22, 617000), datetime.datetime(2013, 5, 6, 6, 1, 1, 168000), datetime.datetime(2013, 5, 6, 14, 56, 48, 236000), datetime.datetime(2013, 5, 7, 16, 47, 4, 597000), datetime.datetime(2013, 5, 8, 15, 26, 52, 105000), datetime.datetime(2013, 5, 10, 6, 10, 39, 379000), datetime.datetime(2013, 5, 13, 6, 9, 57, 990000), datetime.datetime(2013, 5, 13, 19, 56, 15, 354000), datetime.datetime(2013, 5, 15, 16, 39, 9, 127000), datetime.datetime(2013, 5, 16, 5, 59, 27, 609000), datetime.datetime(2013, 5, 16, 14, 18, 33, 253000), datetime.datetime(2013, 5, 17, 6, 20, 11, 853000), datetime.datetime(2013, 5, 21, 15, 38, 10, 53000), datetime.datetime(2013, 5, 22, 5, 59, 8, 126000), datetime.datetime(2013, 5, 22, 15, 48, 55, 877000), datetime.datetime(2013, 5, 23, 5, 47, 4, 779000), datetime.datetime(2013, 5, 23, 16, 59, 16, 948000), datetime.datetime(2013, 5, 24, 10, 57, 34, 831000), datetime.datetime(2013, 5, 24, 12, 29, 17, 332000), datetime.datetime(2013, 5, 27, 17, 0, 14, 513000), datetime.datetime(2013, 6, 20, 7, 28, 45, 975000), datetime.datetime(2013, 6, 20, 13, 31, 13, 228000), datetime.datetime(2013, 6, 21, 6, 18, 47, 789000), datetime.datetime(2013, 7, 1, 6, 12, 3, 640000), datetime.datetime(2013, 7, 1, 14, 33, 9, 251000), datetime.datetime(2013, 7, 2, 14, 59, 0, 421000), datetime.datetime(2013, 7, 3, 6, 12, 58, 282000), datetime.datetime(2013, 7, 3, 17, 23, 38, 745000), datetime.datetime(2013, 7, 5, 13, 40, 44, 719000), datetime.datetime(2013, 7, 9, 14, 51, 27, 348000), datetime.datetime(2013, 7, 10, 5, 12, 3, 104000)]

这个数组应该按日期排序。我想知道这个数组里有多少天。如果同一天有多个日期，我只想算作1天。

我可以手动去做，逐个查看每个日期，然后用一个临时变量来计数，但有没有更快、更合适的方法来“去重”天数呢？

谢谢！

Answer 1

你可以使用 datetime.date() 这个方法来处理日期：

s = {d.date() for d in dates}
print len(s)

因为日期是可以被哈希的，所以你可以把它们放进一个 set 里，这样是没问题的……

---

另外，你还可以统计每个日期出现的次数：

import collections
print collections.Counter(d.date() for d in dates)

或者，甚至可以做一个以日期为键的 datetime 实例列表：

import collections
d = collections.defaultdict(list)
for dt in dates:
    d[dt.date()].append(dt)

不过，我想因为输入数据是排好序的，你也可以用 itertools.groupby 来做类似的事情：

for date, dt_group in itertools.groupby(dates, key=lambda dt: dt.date()):
    print date, list(dt_group)

Answer 2

如果你只是想计算有多少个不同的日期，可以使用下面的代码：

print len({(i.day,i.month,i.year) for i in dates})

这样做是为了确保我们比较的是确实相同的日期，而不仅仅是相同的日期数字。比如，11月1日和12月1日的 .day 数字都是1，但它们显然不是同一天。