how to order columns by names containing special symbols R-CodePudding

I have dataframe below:

df <- data.frame(aa = rep(1,4),
                ae = rep(2,4),
                dd = rep(3,4),
                `aa%` = rep(11,4),
                `ae%` = rep(22,4),
                `dd%` = rep(33,4))

  aa ae dd aa. ae. dd.
1  1  2  3  11  22  33
2  1  2  3  11  22  33
3  1  2  3  11  22  33
4  1  2  3  11  22  33

I want to order the columns to become like

  aa aa. ae ae. dd dd.
1  1  11  2  22  3  33
2  1  11  2  22  3  33
3  1  11  2  22  3  33
4  1  11  2  22  3  33

so I did

library(dplyr)
library(gtools)
df %>%
  select(1, mixedorder(names(.)[-1]))

but this gives

  aa dd aa. ae ae.
1  1  3  11  2  22
2  1  3  11  2  22
3  1  3  11  2  22
4  1  3  11  2  22

How can I get the output with desired order of columns?

CodePudding user response：

Use mixedsort instead of mixedorder

library(gtools)
library(dplyr)
df %>%
    select(mixedsort(names(.)))
   aa aa. ae ae. dd dd.
1  1  11  2  22  3  33
2  1  11  2  22  3  33
3  1  11  2  22  3  33
4  1  11  2  22  3  33

The issue with mixedorder by removing the first column is that it returns an index starting from 1, which needs to start from 2 i.e. add 1

df %>%
   select(1, mixedorder(names(.)[-1]) 1)
  aa aa. ae ae. dd dd.
1  1  11  2  22  3  33
2  1  11  2  22  3  33
3  1  11  2  22  3  33
4  1  11  2  22  3  33

CodePudding user response：

A possible solution:

library(dplyr)

df %>% 
  select(sort(names(.)))

#>   aa aa. ae ae. dd dd.
#> 1  1  11  2  22  3  33
#> 2  1  11  2  22  3  33
#> 3  1  11  2  22  3  33
#> 4  1  11  2  22  3  33

CodePudding user response：

A simplier way to go:

 df[,order(colnames(df))] 

  aa aa. ae ae. dd dd.
1  1  11  2  22  3  33
2  1  11  2  22  3  33
3  1  11  2  22  3  33
4  1  11  2  22  3  33

CodePudding user response：

With dplyr we could:

library(dplyr)

df %>% select(order(colnames(df)))

  aa aa. ae ae. dd dd.
1  1  11  2  22  3  33
2  1  11  2  22  3  33
3  1  11  2  22  3  33
4  1  11  2  22  3  33

CodePudding user response：

Another option:

library(dplyr)
df %>% 
  select(sort(tidyselect::peek_vars()))

Output:

  aa aa. ae ae. dd dd.
1  1  11  2  22  3  33
2  1  11  2  22  3  33
3  1  11  2  22  3  33
4  1  11  2  22  3  33