Preprocessing: text analysis on many columns from a dataframe - r

Question

Using the following lines it is possible to preprocess text in a specific column of my dataframe:
#text to lower case
df$name <- tolower(df$name)
#remove all special characters
df$name <- gsub("[[:punct:]]", " ", df$name)
#remove long spaces
df$name <- gsub("\\s+"," ",str_trim(df$name))
I would like to implement this preprocessing rules in all columns (expect id) of a dataframe like this:
df <- data.frame(id = c("A","B","C"), D = c("mytext 11","mytext +", "!!"), E = c("text","stg","1.2"), F = c("press","remove","22"))

Answer 1

If you want to do something multiple times, it is often useful to define a function.
For example, you could do the following:
library(stringr)
df <- data.frame(id = c("A","B","C"), D = c("mytext 11","mytext +", "!!"),
E = c("text","stg","1.2"), F = c("press","remove","22"))
# create a function so we can apply this multiple times easily.
process <- function(my_vector)
{
my_vector <- tolower(my_vector)
#remove all special characters
my_vector <- gsub("[[:punct:]]", " ", my_vector)
#remove long spaces
my_vector <- gsub("\\s+"," ",str_trim(my_vector))
# return result
return(my_vector)
}
# for all columns except 'id', apply our function.
for(x in setdiff(colnames(df),"id"))
{
df[[x]]=process(df[[x]])
}

Answer 2

You can use dplyr::mutate_at() to mutate multiple columns; in this case, all columns except for id:
mydf %>%
mutate_at(.vars = vars(-id),
.funs = processText)
Where processText is a function containing your desired code:
processText <- function(str) {
str %>%
str_to_lower() %>%
str_replace_all(pattern="[[[:punct:]]]|[\\s+]", replacement=" ", .) %>%
str_trim()
}
The output is as follows:
id D E G
1 A mytext 11 text press
2 B mytext stg remove
3 C 1 2 22