下面是我的数据集，我正在尝试使用dplyr中当前突变值的前一行值。 first（start_time）不起作用，因为在此代码之前发生了group_by，因此第一个引用了组中的第一个不是我想要的。
这是我正在尝试的代码：
mutate(shift_start_time = if_else(diff==0 | diff > 21600,start_time,lag(shift_start_time)))
但是shift_start_time是当前变异的变量无法进行操作，我正在努力进行任何其他工作。
电流：
# A tibble: 703 × 4
start_time end_time diff shift_start_time
<dttm> <dttm> <time> <dttm>
1 2017-05-01 14:21:00 2017-05-01 14:42:18 0 secs 2017-05-01 14:21:00
2 2017-05-01 14:48:00 2017-05-01 15:08:17 342 secs 1970-01-01 00:00:00
3 2017-05-01 15:14:00 2017-05-01 15:35:31 343 secs 1970-01-01 00:00:00
4 2017-05-01 15:42:00 2017-05-01 15:45:36 389 secs 1970-01-01 00:00:00
5 2017-05-01 15:50:00 2017-05-01 16:51:47 264 secs 1970-01-01 00:00:00
6 2017-05-01 16:39:00 2017-05-01 16:51:22 -767 secs 1970-01-01 00:00:00
7 2017-05-01 17:37:00 2017-05-01 17:51:57 2738 secs 1970-01-01 00:00:00
8 2017-05-01 17:46:00 2017-05-01 17:58:25 -357 secs 1970-01-01 00:00:00
9 2017-05-01 17:59:00 2017-05-01 18:11:16 35 secs 1970-01-01 00:00:00
10 2017-05-01 18:08:00 2017-05-01 18:22:22 -196 secs 1970-01-01 00:00:00
必需：
# A tibble: 703 × 4
start_time end_time diff shift_start_time
<dttm> <dttm> <time> <dttm>
1 2017-05-01 14:21:00 2017-05-01 14:42:18 0 secs 2017-05-01 14:21:00
2 2017-05-01 14:48:00 2017-05-01 15:08:17 342 secs 2017-05-01 14:21:00
3 2017-05-01 15:14:00 2017-05-01 15:35:31 343 secs 2017-05-01 14:21:00
4 2017-05-01 15:42:00 2017-05-01 15:45:36 389 secs 2017-05-01 14:21:00
5 2017-05-01 15:50:00 2017-05-01 16:51:47 264 secs 2017-05-01 14:21:00
6 2017-05-01 16:39:00 2017-05-01 16:51:22 -767 secs 2017-05-01 14:21:00
7 2017-05-01 17:37:00 2017-05-01 17:51:57 2738 secs 2017-05-01 14:21:00
8 2017-05-01 17:46:00 2017-05-01 17:58:25 -357 secs 2017-05-01 14:21:00
9 2017-05-01 17:59:00 2017-05-01 18:11:16 35 secs 2017-05-01 14:21:00
10 2017-05-01 18:08:00 2017-05-01 18:22:22 -196 secs 2017-05-01 14:21:00