如果我的行本身相同，如何不计算 datediff

Question

我在 dplyr 的 R 中遇到问题。我计算日期之间的 diffdate 并创建新列，但我想在状态与自身后面的一行（行 +1）相同时省略。它应该是 NA 或 0.

serial      status             date       days days2
312313124   Good Stock         20/01/2021 0    0
312313124   Under Assessment   29/01/2021 9    9
312313124   In Repair          03/02/2021 4    4
312313124   Under Assessment   06/02/2021 3    3 <- is correct, because between status: Under is other status
70453423040 Under Assessment   18/03/2021 0    0
70453423040 In Repair          25/03/2021 7    0
70453423040 In Repair          28/03/2021 3    0 <- should be NA or 0, because in the same serial, status before (1 line above) is the same
12131231    Good Stock         03/04/2021 6

我尝试根据我的数据重新分组，但没有用。我使用这个代码。:

df2 <- df %>%
  distinct() %>%
  group_by(Serial) %>%
  mutate(Days = c(NA, as.numeric(diff(Exported), units='days'))) %>%
  ungroup() %>%
  group_by(Serial, Status, Date) %>%
  mutate(Days2 = if_else(row_number() > 1 , NA, Days)) %>%
  ungroup()

我也尝试过，但是当我有重复项时，即使记录之间的差异大于 2，这段代码也会显示给我。

df3<- df%>%
  group_by(Serial, Status) %>%
  mutate(Days2 = +duplicated((paste(Serial, Status)))

Answer 1

您应该可以使用 lag / lead 来完成此操作。类似于以下内容：

df2 <- df %>%
  distinct() %>%
  group_by(Serial) %>%
  mutate(next_date = lead(date, 1, order_by = date),
         next_status = lead(status, 1, order_by = date)) %>%
  mutate(Days2 = ifelse(status != next_status, next_date - date, 0)

如果我的行本身相同，如何不计算 datediff

How to don't calculate datediff if I have the line following itself the same

r

group-by

dplyr

datediff