Sort a dataframe based on a specific column in R with dplyr-CodePudding

I have a dataframe that looks like this

df <- data.frame(time=seq(1,4,1),col1=c("a","b","d","c"), col2=c("d","a","c","b"))
df
#>   time col1 col2
#> 1    1    a    d
#> 2    2    b    a
#> 3    3    d    c
#> 4    4    c    b

^{Created on 2021-11-06 by the reprex package (v2.0.1)}

I want to sort my data frame based on col2 and look like this

  time col1 col2 
    3    d    d
    1    a    a
    4    c    c
    2    b    b

Any ideas or help is highly appreciated!

CodePudding user response：

Don't know, if this makes any sense, but you could do a self join:

library(tidyr)
library(dplyr)

df %>% 
  select(col2) %>% 
  inner_join(df %>% mutate(col2 = col1), by = "col2") %>% 
  select(time, col1, col2)

This returns

  time col1 col2
1    3    d    d
2    1    a    a
3    4    c    c
4    2    b    b

CodePudding user response：

A solution in base R:

df <- data.frame(time = match(df$col2,df$col1),  col1 = df$col2, col2=df$col2)

#>   time col1 col2
#> 1    3    d    d
#> 2    1    a    a
#> 3    4    c    c
#> 4    2    b    b