Question

我有一个很大的开始结束时间序列数据库，并且想要合并在“ 00:00:00”“结束”（第三列）和在“ 00:00”“开始”（第二列）的所有时间条目：下一行为00”。

当前尝试在末尾正确的地方删除附加的“ 00:00:00”。

#Before
df<-data.frame(id=1,start=c("17/07/2017 19:30","17/07/2017 22:30",
                            "18/07/2017 00:30","18/07/2017 19:00","18/07/2017 21:30",
                            "19/07/2017 00:00","19/07/2017 02:30","19/07/2017 21:00",
                            "19/07/2017 22:30","20/07/2017 00:00",
                            "20/07/2017 03:00","20/07/2017 22:00","20/07/2017 23:00",
                            "21/07/2017 00:00","21/07/2017 01:30"),
               end=c("17/07/2017 22:00","18/07/2017 00:00","18/07/2017 02:30",
                     "18/07/2017 21:00","19/07/2017 00:00",
                     "19/07/2017 02:00","19/07/2017 04:30","19/07/2017 21:30",
                     "20/07/2017 00:00","20/07/2017 02:00",
                     "20/07/2017 04:00","20/07/2017 22:30","21/07/2017 00:00",
                     "21/07/2017 01:00","21/07/2017 02:00"))

#After
dfoutput<-data.frame(id=1,start=c("17/07/2017 19:30","17/07/2017 22:30",
                                  "18/07/2017 00:30","18/07/2017 19:00","18/07/2017 21:30",
                                  "19/07/2017 02:30","19/07/2017 21:00",
                                  "19/07/2017 22:30","20/07/2017 03:00","20/07/2017 22:00",
                                  "20/07/2017 23:00","21/07/2017 01:30"),
                     end=c("17/07/2017 22:00","18/07/2017 00:00",
                           "18/07/2017 02:30","18/07/2017 21:00","19/07/2017 02:00",
                           "19/07/2017 04:30","19/07/2017 21:30",
                           "20/07/2017 02:00","20/07/2017 04:00","20/07/2017 22:30",
                           "21/07/2017 01:00","21/07/2017 02:00"))

当前ifelse结果删除正确的最后一行中的“ 00:00:00”。

Answer 1

我在dplyr中提出了两种解决方案：

df %>% 
  group_by(id) %>% 
  mutate(index = (substr(start,12,16) == "00:00" & 
                  substr(lag(end),12,16) == "00:00"),
         start2 = if_else(index, lag(start), start),
         end2 = if_else(lead(index), lead(end), end)) %>% 
  mutate(start = if_else(is.na(start2), start, start2),
         end = if_else(is.na(end2), end, end2)) %>% 
  filter(!index) %>% 
  select(-index, -start2, -end2)

或

df %>% 
  group_by(id) %>% 
  arrange(start, end) %>% 
  mutate(index1 = ((substr(start,12,16) == "00:00" & 
                     substr(lag(end),12,16) == "00:00") |
                    (substr(lead(start),12,16) == "00:00" & 
                       substr(end,12,16) == "00:00")),
         index2 = cumsum(index1 != c(F, lag(index1)[-1]))) %>% 
  group_by(id, index2) %>% 
  mutate(start = if_else(index1, first(start), start),
           end = if_else(index1, last(end), end)) %>% 
  ungroup %>% 
  select(-index1, -index2) %>%
  unique()

都给：

#> # A tibble: 12 x 3
#>       id start            end             
#>    <dbl> <fct>            <fct>           
#>  1     1 17/07/2017 19:30 17/07/2017 22:00
#>  2     1 17/07/2017 22:30 18/07/2017 00:00
#>  3     1 18/07/2017 00:30 18/07/2017 02:30
#>  4     1 18/07/2017 19:00 18/07/2017 21:00
#>  5     1 18/07/2017 21:30 19/07/2017 02:00
#>  6     1 19/07/2017 02:30 19/07/2017 04:30
#>  7     1 19/07/2017 21:00 19/07/2017 21:30
#>  8     1 19/07/2017 22:30 20/07/2017 02:00
#>  9     1 20/07/2017 03:00 20/07/2017 04:00
#> 10     1 20/07/2017 22:00 20/07/2017 22:30
#> 11     1 20/07/2017 23:00 21/07/2017 01:00
#> 12     1 21/07/2017 01:30 21/07/2017 02:00

合并发生在下一行的时间点

1 个答案: