Using dplyr + gsub on many columns

Question

I'm using dplyr and gsub to remove special characters. I'm trying to translate a code I had with base R.

Here's a fake example to resemble my data:

library(dplyr)

region = c("regi\xf3n de tarapac\xe1","regi\xf3n de tarapac\xe1")
provincia = c("cami\xf1a","iquique")
comuna = c("tamarugal","alto hospicio")

comunas = tibble(region,provincia,comuna)

This works for me:

comunas = comunas %>% 
  mutate(comuna = gsub("\xe1", "\u00e1", comuna), # a with acute
         comuna = gsub("<e1>", "\u00e1", comuna) # a with acute
  )

But now I want to apply the same to every column:

comunas = comunas %>% 
  mutate_all(funs(gsub("\xe1", "\u00e1", .), # a with acute
         gsub("<e1>", "\u00e1", .) # a with acute
  ))

And I see the last chunk has no effect. The idea is to obtain:

              region provincia        comuna
1 región de tarapacá    camiña     tamarugal
2 región de tarapacá   iquique alto hospicio

And any other needed change.

Any idea? many thanks in advance !

Please provide a reproducible example. We do not have access to comunas_casen_2015. http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example — Pierre Lapointe, Apr 17 '17 at 17:57
Make a small reproducible example. It doesn't have to include everything you have. — Pierre Lapointe, Apr 17 '17 at 18:03
comuna doesn't change. Is there anything to change in it? Also provide what you expect as a result. — Pierre Lapointe, Apr 17 '17 at 18:09
this doesn't need to change in every cell, but the idea is to apply the command to every cell because there's no special characters on every cell — pachadotdev, Apr 17 '17 at 18:12

score 29 · Accepted Answer · edited Aug 17 '23 at 21:55

2021 update

mutate_all is now replaced with across. Here are two ways to use gsub across many columns with the help of dplyr:

library(dplyr)

#data
region = c("regi\xf3n de tarapac\xe1","regi\xf3n de tarapac\xe1")
provincia = c("cami\xf1a","iquique")
comuna = c("tamarugal","alto hospicio")

comunas = data.frame(region,provincia,comuna,stringsAsFactors=FALSE)

#Without anonymous function
comunas %>%
  mutate(across(everything(), gsub, pattern = "\xe1|<e1>", replacement = "\u00e1"))

#With anonymous function
comunas %>%
  mutate(across(everything(),~ gsub("\xe1|<e1>","\u00e1", .)))

              region provincia        comuna
1 región de tarapacá    camiña     tamarugal
2 región de tarapacá   iquique alto hospicio

Original answer

This works for me:

region = c("regi\xf3n de tarapac\xe1","regi\xf3n de tarapac\xe1")
provincia = c("cami\xf1a","iquique")
comuna = c("tamarugal","alto hospicio")

comunas = data.frame(region,provincia,comuna,stringsAsFactors=FALSE)


comunas %>%
  mutate(region = gsub("\xe1", "\u00e1", region), # a with acute
         region = gsub("<e1>", "\u00e1", region) # a with acute
  )
  
  
comunas %>%
  mutate_all(funs(gsub("\xe1", "\u00e1", .), # a with acute
         gsub("<e1>", "\u00e1", .) # a with acute
  ))

              region provincia        comuna        region_gsub provincia_gsub   comuna_gsub
1 región de tarapacá    camiña     tamarugal región de tarapacá         camiña     tamarugal
2 región de tarapacá   iquique alto hospicio región de tarapacá        iquique alto hospicio

@ethantenison 'mutate_all' has been superseded in 'dplyr' and replaced with 'mutate' and 'across'. I updated my answer — Pierre Lapointe, Dec 08 '21 at 12:13

Using dplyr + gsub on many columns

1 Answers1