在dplyr中，使用分组过滤器删除Behavior等于"Push"且等于row_number() == n()的行（这意味着它是组中的最后一行）。

library(dplyr)

Events %>% 
  group_by(ID) %>% 
  filter(!(Behavior == "Push" & row_number() == n())) %>% 
  ungroup()

# A tibble: 13 × 3
   ID    Behavior Order
   <chr> <chr>    <dbl>
 1 A     Bucket       1
 2 A     Tube         2
 3 A     Bucket       3
 4 A     Tube         4
 5 B     Bucket       1
 6 B     Tube         2
 7 B     Bucket       3
 8 B     Tube         4
 9 C     Bucket       1
10 C     Tube         2
11 C     Push         3
12 C     Bucket       4
13 C     Tube         5

使用data.table的另一个选项
其中.SD是每组ID的数据子集。然后删除Behavior =="Push" AND Order ==. N的任何行。其中. N表示每组的总行数。

library(data.table)

# sample data
df = data.table(
  ID = c(rep("A",5),rep("B",4), rep("C",5)),
  Behavior = c("Bucket","Tube","Bucket","Tube","Push","Bucket","Tube","Bucket","Tube","Bucket","Tube","Push","Bucket","Tube")
)
df[, Order := 1:.N, by = ID]

# subsetting
df = df[, .SD[!(Behavior == "Push" & Order == .N)], by = ID]

> df
    ID Behavior Order
 1:  A   Bucket     1
 2:  A     Tube     2
 3:  A   Bucket     3
 4:  A     Tube     4
 5:  B   Bucket     1
 6:  B     Tube     2
 7:  B   Bucket     3
 8:  B     Tube     4
 9:  C   Bucket     1
10:  C     Tube     2
11:  C     Push     3
12:  C   Bucket     4
13:  C     Tube     5