R - 根据多列识别重复行并根据日期删除它们

Question

1. 我想识别 ID 列中的重复项，但前提是 Wave==2 （在下面的示例中，只有 'C' 在第 2 波中重复）。

2. 然后我想根据Date 选择最新的副本并将其从数据框df 中删除。

我该怎么做？

structure(list(ID = c("E", "G", "C", "B", "D", "E", "A", "D", 
"F", "F", "C", "A", "B", "C", "A"), Wave = c(2L, 1L, 1L, 2L, 
2L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 1L, 2L, 1L), Date = c("25/02/2020", 
"18/02/2020", "14/02/2020", "21/02/2020", "24/02/2020", "16/02/2020", 
"12/02/2020", "15/02/2020", "17/02/2020", "26/02/2020", "22/02/2020", 
"20/02/2020", "13/02/2020", "23/02/2020", "11/02/2020")), class = "data.frame", row.names = c(NA, 
-15L))

Answer 1

您可以使用slice 选择Wave = 2 所在的最新行。

library(dplyr)

df %>%
  mutate(Date = lubridate::dmy(Date)) %>%
  group_by(ID, Wave) %>%
  slice(if(first(Wave) == 2) which.max(Date) else seq_len(n()))

#   ID     Wave Date      
#   <chr> <int> <date>    
# 1 A         1 2020-02-12
# 2 A         1 2020-02-11
# 3 A         2 2020-02-20
# 4 B         1 2020-02-13
# 5 B         2 2020-02-21
# 6 C         1 2020-02-14
# 7 C         2 2020-02-23
# 8 D         1 2020-02-15
# 9 D         2 2020-02-24
#10 E         1 2020-02-16
#11 E         2 2020-02-25
#12 F         1 2020-02-17
#13 F         2 2020-02-26
#14 G         1 2020-02-18

Answer 2

这是filter的选项

library(dplyr)
library(lubridate)
df1 %>% 
    arrange(ID, Wave, dmy(Date)) %>%
    group_by(ID, Wave) %>% 
    filter((row_number() == 1 & first(Wave) == 2)|first(Wave) != 2)
# A tibble: 14 x 3
# Groups:   ID, Wave [13]
#   ID     Wave Date      
#   <chr> <int> <chr>     
# 1 A         1 11/02/2020
# 2 A         1 12/02/2020
# 3 A         2 20/02/2020
# 4 B         1 13/02/2020
# 5 B         2 21/02/2020
# 6 C         1 14/02/2020
# 7 C         2 22/02/2020
# 8 D         1 15/02/2020
# 9 D         2 24/02/2020
#10 E         1 16/02/2020
#11 E         2 25/02/2020
#12 F         1 17/02/2020
#13 F         2 26/02/2020
#14 G         1 18/02/2020