Python 在时间跨度中搜索

我有这个带有时间戳的列表，我希望能够根据用户输入（小时作为一个工作流：

• 使用

  datetime

  模块，使用

  strtime

  方法将字符串转换为

  datetime

  对象：您将获得

  datetime

  对象的列表
• 通过从最后一个列表中减去此列表的每个条目来计算时间增量
• 您可以使用

  timedelta

  的

  seconds

  属性来确定一个点与参考点之间的秒数：将其与

  3*3600

  （3h）进行比较，以确定哪些项目属于正确的时间段

---

类似的方法应该可以奏效

ls =  ['2002-03-31 19:30:41', '2002-03-31 19:30:41', '2002-03-31 19:43:49', '2002-03-31     19:43:50', '2002-03-31 19:50:05', '2002-03-31 19:50:06', '2002-03-31 19:50:06', '2002-03-31 19:50:44', '2002-03-31 19:50:45', '2002-03-31 19:50:45', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:53:05', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:17', '2002-03-31 19:54:14', '2002-03-31 19:54:14']

# target is one of the items in the list
target = datetime.strptime('2002-03-31 19:53:17', '%Y-%m-%d %H:%M:%S')
for l in ls:
    print datetime.strptime(l, '%Y-%m-%d %H:%M:%S') - target

印刷品

-1 day, 23:37:24
-1 day, 23:37:24
-1 day, 23:50:32
-1 day, 23:50:33
-1 day, 23:56:48
-1 day, 23:56:49
-1 day, 23:56:49
-1 day, 23:57:27
-1 day, 23:57:28
-1 day, 23:57:28
-1 day, 23:58:33
-1 day, 23:58:33
-1 day, 23:58:33
-1 day, 23:59:08
-1 day, 23:59:08
-1 day, 23:59:08
-1 day, 23:59:48
-1 day, 23:59:49
-1 day, 23:59:49
-1 day, 23:59:49
0:00:00
0:00:57
0:00:57

datetime.strtime（l、%Y-%m-%d%H:%m:%S'）-target

返回一个

timedelta

对象（）。您可以访问

timedelta

对象

days

、

seconds

和

microseconds

属性，并将它们与所需的时间跨度进行比较。例如，要获取从某个参考点开始不到一小时发生的所有事件的所有索引：

less_than_an_hour = []
for i,l in enumerate(ls):
    if (datetime.strptime(l, '%Y-%m-%d %H:%M:%S') - target).seconds < 3600:
        less_than_an_hour.append(i)

少于一小时=[]
对于枚举中的i，l（ls）：
如果（datetime.strTime（l，'%Y-%m-%d%H:%m:%S'）-目标）。秒<3600：
少于一小时。附加（i）

您非常幸运，因为您的时间戳以最简单的排序顺序排列，您可以将整个“转换为时间值”，只需进行字符串比较：

times =  ['2002-03-31 19:30:41', '2002-03-31 19:30:41', '2002-03-31 19:43:49', '2002-03-31     19:43:50', '2002-03-31 19:50:05', '2002-03-31 19:50:06', '2002-03-31 19:50:06', '2002-03-31 19:50:44', '2002-03-31 19:50:45', '2002-03-31 19:50:45', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:53:05', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:17', '2002-03-31 19:54:14', '2002-03-31 19:54:14']
activity= ['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w']

start = '2002-03-31 16:54:14'
end = '2002-03-31 19:54:14'

for time, activity in zip(times, activity):
    if time >= start and time <= end:
        print time, activity

次数=['2002-03-31 19:30:41', '2002-03-31 19:30:41', '2002-03-31 19:43:49', '2002-03-31     19:43:50', '2002-03-31 19:50:05', '2002-03-31 19:50:06', '2002-03-31 19:50:06', '2002-03-31 19:50:44', '2002-03-31 19:50:45', '2002-03-31 19:50:45', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:53:05', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:17', '2002-03-31 19:54:14', '2002-03-31 19:54:14']
活度=['a'、'b'、'c'、'd'、'e'、'f'、'g'、'h'、'i'、'j'、'k'、'l'、'm'、'n'、'o'、'p'、'q'、'r'、's'、't'、'u'、'v'、'w']
开始='2002-03-3116:54:14'
完＝‘2002-03-3119:54:14’
对于时间，zip中的活动（时间，活动）：
如果时间>=开始和时间我会：


将时间戳列表转换为
datetime
对象：

times = [datetime.datetime.strptime(t, '%Y-%m-%d %H:%M:%S') for t in times]


使用查找用户请求的开始时间。如果您也将用户输入转换为
datetime
对象，则使用
bisect
比使用线性搜索快得多：

start = datetime.datetime(2002, 3, 31, 19, 53, 17)
startindex = bisect.bisect_left(times, start)


使用函数将两个列表合并为一个，以显示与您的范围匹配的条目：

end = datetime.datetime(2002, 4, 1, 07, 53, 17)

merged = itertools.izip(times, activity)
fromstart = itertools.islice(merged, startindex)
untilend = itertools.takewhile(lambda e: e[0] <= end, fromstart)

用户多久搜索一次此信息？时间戳列表有多大？@MartijnPieters不太常见，执行时间不重要。大约14000个时间戳。谢谢！我如何转换用户输入，比如说13天（从午夜开始）要开始？请查看
datetime
模块。您可以使用类似
start=（datetime.date.today（）-datetime.timedelta（days=13））.strftime（“%Y-%m-%d 00:00:00”）
的内容预先计算开始日期，然后将
if
语句更改为
if time>=start:…
。
end = datetime.datetime(2002, 4, 1, 07, 53, 17)

merged = itertools.izip(times, activity)
fromstart = itertools.islice(merged, startindex)
untilend = itertools.takewhile(lambda e: e[0] <= end, fromstart)

>>> import itertools
>>> import datetime
>>> import bisect
>>> times =  ['2002-03-31 19:30:41', '2002-03-31 19:30:41', '2002-03-31 19:43:49', '2002-03-31     19:43:50', '2002-03-31 19:50:05', '2002-03-31 19:50:06', '2002-03-31 19:50:06', '2002-03-31 19:50:44', '2002-03-31 19:50:45', '2002-03-31 19:50:45', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:51:50', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:52:25', '2002-03-31 19:53:05', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:06', '2002-03-31 19:53:17', '2002-03-31 19:54:14', '2002-03-31 19:54:14']
>>> activity= ['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w']
>>> times = [datetime.datetime.strptime(t, '%Y-%m-%d %H:%M:%S') for t in times]
>>> start = datetime.datetime(2002, 3, 31, 19, 53, 17)
>>> end = datetime.datetime(2002, 4, 1, 07, 53, 17)
>>> startindex = bisect.bisect_left(times, start)
>>> merged = itertools.izip(times, activity)
>>> fromstart = itertools.islice(merged, startindex)
>>> untilend = itertools.takewhile(lambda e: e[0] <= end, fromstart)
>>> for time, activity in untilend:
...     print time, activity
... 
2002-03-31 19:30:41 a
2002-03-31 19:30:41 b
2002-03-31 19:43:49 c
2002-03-31 19:43:50 d
2002-03-31 19:50:05 e
2002-03-31 19:50:06 f
2002-03-31 19:50:06 g
2002-03-31 19:50:44 h
2002-03-31 19:50:45 i
2002-03-31 19:50:45 j
2002-03-31 19:51:50 k
2002-03-31 19:51:50 l
2002-03-31 19:51:50 m
2002-03-31 19:52:25 n
2002-03-31 19:52:25 o
2002-03-31 19:52:25 p
2002-03-31 19:53:05 q
2002-03-31 19:53:06 r
2002-03-31 19:53:06 s
2002-03-31 19:53:06 t