I have a dataframe that looks like this:

structure(list(Date = structure(c(1630544400, 1630548000, 1630551600, 
1630555200, 1630558800, 1630562400, 1630566000, 1630569600, 1630573200, 
1630576800, 1630580400, 1630630800, 1630634400, 1630638000, 1630641600, 
1630645200, 1630648800, 1630652400, 1630656000, 1630659600), tzone = "America/Chicago", class = c("POSIXct", 
"POSIXt")), daytime = c("Night", "Night", "Night", "Night", "Morning", 
"Morning", "Morning", "Morning", "Morning", "Morning", "Morning", 
"Night", "Night", "Night", "Night", "Morning", "Morning", "Morning", 
"Morning", "Morning")), row.names = c(NA, -20L), class = c("tbl_df", 
"tbl", "data.frame"))

I would like to create another column to group the night and morning sequentially so the output would look like this:

   Date                daytime        nightcount
   <dttm>              <chr>          <dbl>
 1 2021-09-01 20:00:00 Night            1
 2 2021-09-01 21:00:00 Night            1  
 3 2021-09-01 22:00:00 Night            1
 4 2021-09-01 23:00:00 Night            1
 5 2021-09-02 00:00:00 Morning          1
 6 2021-09-02 01:00:00 Morning          1
 7 2021-09-02 02:00:00 Morning          1
 8 2021-09-02 03:00:00 Morning          1
 9 2021-09-02 04:00:00 Morning          1
10 2021-09-02 05:00:00 Morning          1
11 2021-09-02 06:00:00 Morning          1
12 2021-09-02 20:00:00 Night            2
13 2021-09-02 21:00:00 Night            2  
14 2021-09-02 22:00:00 Night            2
15 2021-09-02 23:00:00 Night            2  
16 2021-09-03 00:00:00 Morning          2
17 2021-09-03 01:00:00 Morning          2
18 2021-09-03 02:00:00 Morning          2
19 2021-09-03 03:00:00 Morning          2
20 2021-09-03 04:00:00 Morning          2

Is there an easy solution for this using dplyr?

>Solution :

You can create a logical value when "Morning" turns to "Night" and then use cumsum to sum these logical values across rows:

library(dplyr)

df |>
  mutate(nightcount = cumsum(daytime == "Night" & lag(daytime, default = "Morning") == "Morning"))