Question

我有一个带有不规则时间戳的时间序列数据集。如何将其转换为以5分钟为间隔设置的正则化时间序列

Datetime           Value
11/6/2017 07:56:41  16
11/6/2017 08:01:22  16
11/6/2017 08:06:39  23
11/6/2017 08:11:56  23
11/6/2017 08:23:18  25
11/6/2017 08:29:11  31
11/6/2017 08:36:40  33
11/6/2017 08:42:05  39
11/6/2017 08:47:42  39
11/6/2017 08:53:08  37
11/6/2017 08:58:28  39
11/6/2017 09:03:50  39
11/6/2017 09:09:19  39

Answer 1

因为您没有显示任何代码，或者您从哪里获得数据，所以我只能提供快速而肮脏的建议。因此，没有使用datetime或math。

data = """
11/6/2017 07:56:41  16
11/6/2017 08:01:22  16
11/6/2017 08:06:39  23
11/6/2017 08:11:56  23
11/6/2017 08:23:18  25
11/6/2017 08:29:11  31
11/6/2017 08:36:40  33
11/6/2017 08:42:05  39
11/6/2017 08:47:42  39
11/6/2017 08:53:08  37
11/6/2017 08:58:28  39
11/6/2017 09:03:50  39
11/6/2017 09:09:19  39""".strip()

def rewrite_line(line):
    date, time, count = line.split()
    hours, minutes, seconds = time.split(':')
    minutes, seconds = (int(minutes)/5)*5, 0
    time = '%02s:%02d:%02d' % (hours, minutes, seconds)
    return '%s %s %s' % (date, time, count)

for line in data.splitlines():
    print rewrite_line(line)

应该打印如下：

11/6/2017 07:55:00 16
11/6/2017 08:00:00 16
11/6/2017 08:05:00 23
11/6/2017 08:10:00 23
11/6/2017 08:20:00 25
11/6/2017 08:25:00 31
11/6/2017 08:35:00 33
11/6/2017 08:40:00 39
11/6/2017 08:45:00 39
11/6/2017 08:50:00 37
11/6/2017 08:55:00 39
11/6/2017 09:00:00 39
11/6/2017 09:05:00 39

将不规则带时间戳的数据转换为等间隔时间序列

1 个答案: