根据R中的规则从data.frame中删除行[重复]

Question

在 data.frame 中，如果在所有其他列中存在具有相同信息的另一行，我想自动删除 Column_E 上带有“NA”的行，例如：

Column_A   Column_B   Column_C    Column_D      Column_E
A121       NAME1      A321        2019-01-01    NA
A121       NAME1      A321        2019-01-01    2020-02-01
A123       NAME2      A322        2019-01-01    2020-01-01
A123       NAME2      A322        2019-01-01    NA
A124       NAME3      A323        2019-01-01    2019-01-01
A124       NAME4      A324        2019-01-01    NA

输出应该是：

Column_A   Column_B   Column_C   Column_D       Column_E
A121       NAME1      A321        2019-01-01    2020-02-01
A123       NAME2      A322        2019-01-01    2020-01-01
A124       NAME3      A323        2019-01-01    2019-01-01
A124       NAME4      A324        2019-01-01    NA

有什么想法吗？

Answer 1

您可以选择没有NA 值或组中只有1 行的行。

library(dplyr)
df %>%
  group_by(across(Column_A:Column_D)) %>%
  filter(!is.na(Column_E) | n() == 1)

# Column_A Column_B Column_C Column_D   Column_E  
#  <chr>    <chr>    <chr>    <chr>      <chr>     
#1 A121     NAME1    A321     2019-01-01 2020-02-01
#2 A123     NAME2    A322     2019-01-01 2020-01-01
#3 A124     NAME3    A323     2019-01-01 2019-01-01
#4 A124     NAME4    A324     2019-01-01 NA       

---

data.table 中的逻辑相同：

library(data.table)

setDT(df)
df[, .SD[!is.na(Column_E) | .N == 1], .(Column_A, Column_B, Column_C, Column_D)]

和基础R：

subset(df, ave(!is.na(Column_E),Column_A, Column_B, Column_C, Column_D, 
          FUN = function(x) x | length(x) == 1))

数据

df <- structure(list(Column_A = c("A121", "A121", "A123", "A123", "A124", 
"A124"), Column_B = c("NAME1", "NAME1", "NAME2", "NAME2", "NAME3", 
"NAME4"), Column_C = c("A321", "A321", "A322", "A322", "A323", 
"A324"), Column_D = c("2019-01-01", "2019-01-01", "2019-01-01", 
"2019-01-01", "2019-01-01", "2019-01-01"), Column_E = c(NA, "2020-02-01", 
"2020-01-01", NA, "2019-01-01", NA)), class = "data.frame", 
row.names = c(NA, -6L))