从 R 的特定列中删除单元格中的重复项？

Question

I have a data frame which contains repeated characters or numbers. I want to create a new df which only contains unique cells in each of these columns. Below is a visual of what I am trying to achieve. Any ideas would be highly appreciated.

Answer 1

Here's a regex solution (based on mock data in the absence of reproducible data):

library(stringr)
df[,1:3] <- lapply(df[,1:3], function(x) str_extract_all(x, "(\\b\\w+\\b)(?!.*\\1)"))

The solution drwas on negative lookahead ( (?....) ) and backreference ( \\1 ): the pattern (\\b\\w+\\b)(?..*\\1) is used to str_extract_all alphanumeric strings unless they are repeated later in the string, which effectively captures all unique values:

Result:

df
               Title   Length    Prediction
1             George 555, 666           111
2 Alice, Peter, Kate 123, 444 333, 777, 222

Data:

df <- data.frame(
  Title = c("George,George,George", "Kate,Alice,Kate,Peter,Kate"),
  Length = c("555,555,666", "123,123,444,123,444"), 
  Prediction = c("111,111,111", "222,333,222,777,222"), stringsAsFactors = F)

Answer 2

Do like this. Using the df created by ChrisRuehlemann

library(tidyverse)
df %>% mutate(across(everything(), ~str_split(., ",")),
              across(everything(), ~map(., ~unique(.x))))
               Title   Length    Prediction
1             George 555, 666           111
2 Kate, Alice, Peter 123, 444 222, 333, 777

Or one-liner

mutate(df, across(everything(), ~map(str_split(., ","), ~unique(.x))))

               Title   Length    Prediction
1             George 555, 666           111
2 Kate, Alice, Peter 123, 444 222, 333, 777