R remove duplicates based on other columns

问题

I want to remove duplicates based on similarities or differences of other columns.

All the duplicated ID should be completely removed but just if they have DIFFERENT colours. It doesn't matter if they have different subgroups as well. If they have the same ID AND the same colour, just the first one should be kept.

At the end I want to have a list of all ID which are single-colour only (independent of subgroup). All the multicoloured ID should be removed.

Here and example:

   id colour   subgroup
1   1    red   lightred
2   2   blue  lightblue
3   2   blue   darkblue
4   3    red   lightred
5   4    red    darkred
6   4    red    darkred
7   4   blue  lightblue
8   5  green  darkgreen
9   5  green  darkgreen
10  5  green lightgreen
11  6    red    darkred
12  6   blue   darkblue
13  6  green lightgreen

At the end it should look like this:

  id colour  subgroup
1  1    red  lightred
2  2   blue lightblue
4  3    red  lightred
8  5  green darkgreen

The data I used for this example:

id = c(1,2,2,3,4,4,4,5,5,5,6,6,6)
colour = c("red","blue","blue","red","red","red","blue","green","green","green","red","blue","green")
subgroup = c("lightred","lightblue","darkblue","lightred","darkred","darkred","lightblue","darkgreen","darkgreen","lightgreen","darkred","darkblue","lightgreen")
data = data.frame(cbind(id,colour,subgroup))

Thanks for your help!

回答1:

library(tidyverse)
data%>%
  group_by(id)%>%
  filter(1==length(unique(colour)),!duplicated(colour))
# A tibble: 4 x 3
# Groups:   id [4]
  id    colour subgroup 
  <fct> <fct>  <fct>    
1 1     red    lightred 
2 2     blue   lightblue
3 3     red    lightred 
4 5     green  darkgreen

Using Base R:

 subset(data,as.logical(ave(colour,id,FUN=function(x)length(unique(x))==1& !duplicated(x))))
  id colour  subgroup
1  1    red  lightred
2  2   blue lightblue
4  3    red  lightred
8  5  green darkgreen

回答2:

I got a small data.table solution. It first filters on all non-duplicated id, colour combinations and then selects all combinations, where only one id, colour combination exists.

library(data.table)
dt.data <- data.table(data)
dt.data[!duplicated(dt.data, by = c("id", "colour"))
                       ,.(colour, subgroup, .N)
                       , by = list(id)][N==1, .(id
                                               , colour
                                               , subgroup)]

来源：https://stackoverflow.com/questions/51095249/r-remove-duplicates-based-on-other-columns

标签

duplicates