如何在R中删除具有单个唯一ID的行?

Question

我有如下编码的数据集。对于一组特定的治疗对、年、月、级别，我分配了唯一的ID。理想情况下，一个完整的"设置";有两行对应于相同的惟一ID。如果没有，我想要删除这些行。

这里，all "sets"除了ID2对应的ID外，其余的都是唯一的ID。在我的原始数据集中，我有成千上万这样的行。如何扫描以移除这些单元素?

tmt.pair <- c("A","A","A","B","B","B","B")
tmt <- c("1000 C","4000 C","1000 C","1000 C","4000 C","1000 C","4000 C")
year <- c("2021","2021","2021","2021","2021","2020","2020")
month <- c("A","A","A","J","J","O","O")
level <- c("Low","Low","Up","Low","Low","Low","Low")
site <- c(1,1,2,1,1,1,1)
val <- rnorm(7,5,1)
df <- data.frame(tmt.pair, year,month, level,tmt,val)
df$ID <- cumsum(!duplicated(df[1:4]))

tmt.pair year month level tmt    val       ID
1        A 2021     A   Low 1000 C 4.789715  1
2        A 2021     A   Low 4000 C 6.451113  1
3        A 2021     A    Up 1000 C 4.281171  2
4        B 2021     J   Low 1000 C 5.176668  3
5        B 2021     J   Low 4000 C 6.384432  3
6        B 2020     O   Low 1000 C 4.833731  4
7        B 2020     O   Low 4000 C 3.274355  4

Answer 1

使用base R可以这样做:

tab=table(df$ID)
df[ifelse(tab[df$ID]==1, FALSE, TRUE),] 

输出:

tmt.pair year month level    tmt      val ID
1        A 2021     A   Low 1000 C 5.156294  1
2        A 2021     A   Low 4000 C 4.395990  1
4        B 2021     J   Low 1000 C 5.714170  3
5        B 2021     J   Low 4000 C 6.075886  3
6        B 2020     O   Low 1000 C 7.249756  4
7        B 2020     O   Low 4000 C 5.197891  4

Answer 2

另一个使用data.table的选项:

library(data.table)

setDT(df)[,if(.N > 1) .SD, by=ID]

ID tmt.pair year month level    tmt      val
1:  1        A 2021     A   Low 1000 C 4.424811
2:  1        A 2021     A   Low 4000 C 4.556058
3:  3        B 2021     J   Low 1000 C 4.396996
4:  3        B 2021     J   Low 4000 C 3.906065
5:  4        B 2020     O   Low 1000 C 5.714706
6:  4        B 2020     O   Low 4000 C 4.891188

或者对于dplyr，我们只保留有多于一个观测值的IDs:

library(dplyr)
df %>%
group_by(ID) %>%
filter(n() > 1)