Ranking dataframe in reverse date order

Question

I'm trying to rank a data frame based on reverse order for dates. For example, see below:

user	start_time
ID	2022-05-31 18:13:43
BB	2020-11-07 00:42:36
RL	2021-11-25 01:01:15
ID	2020-05-30 03:30:19
KF	2021-05-18 00:26:32
MT	2021-07-02 22:44:01
BB	2021-08-01 16:06:37
RL	2021-02-02 20:54:08
BB	2022-06-25 22:52:56
KF	2021-04-24 01:45:18

The reverse order rank should be:

user	start_time	rank
ID	2022-05-31 18:13:43	1
BB	2020-11-07 00:42:36	3
RL	2021-11-25 01:01:15	1
ID	2020-05-30 03:30:19	2
KF	2021-05-18 00:26:32	1
MT	2021-07-02 22:44:01	1
BB	2021-08-01 16:06:37	2
RL	2021-02-02 20:54:08	2
BB	2022-06-25 22:52:56	1
KF	2021-04-24 01:45:18	2

see https://stackoverflow.com/questions/6246159/how-to-sort-a-data-frame-by-date — sharpie, Jul 20 '22 at 06:38

score 1 · Answer 1 · answered Jul 20 '22 at 06:41

One

do.call(
  rbind,
  lapply(split(df,df$user),function(x){
    cbind(x,"rank"=order(x$start_time,decreasing=T))
  })
)

      user          start_time rank
BB.2    BB 2020-11-07 00:42:36    3
BB.7    BB 2021-08-01 16:06:37    2
BB.9    BB 2022-06-25 22:52:56    1
ID.1    ID 2022-05-31 18:13:43    1
ID.4    ID 2020-05-30 03:30:19    2
KF.5    KF 2021-05-18 00:26:32    1
KF.10   KF 2021-04-24 01:45:18    2
MT      MT 2021-07-02 22:44:01    1
RL.3    RL 2021-11-25 01:01:15    1
RL.8    RL 2021-02-02 20:54:08    2

score 1 · Answer 2 · answered Jul 20 '22 at 06:49

Using dplyr,

library(dplyr)
df %>%
  group_by(user) %>%
  mutate(rank = order(start_time, decreasing = T))

   user  start_time           rank
   <chr> <chr>               <int>
 1 ID    2022-05-31 18:13:43     1
 2 BB    2020-11-07 00:42:36     3
 3 RL    2021-11-25 01:01:15     1
 4 ID    2020-05-30 03:30:19     2
 5 KF    2021-05-18 00:26:32     1
 6 MT    2021-07-02 22:44:01     1
 7 BB    2021-08-01 16:06:37     2
 8 RL    2021-02-02 20:54:08     2
 9 BB    2022-06-25 22:52:56     1
10 KF    2021-04-24 01:45:18     2

Ranking dataframe in reverse date order

2 Answers2