Replacing numbers with characters in a text column in R-CodePudding

I would like to replace some numbers in the text column of my data. The numbers are either 8 or 9 digits and in two formats. This is snapshot of the data:

df <- data.frame(
  notes = c(
    'my number is 123-41-567',
    "321 12 788 is valid",
    'why not taking 987-012-678',
    '120 967 325 is correct'
  )
)

df %>% select(notes)

                       notes
1    my number is 123-41-567
2        321 12 788 is valid
3 why not taking 987-012-678
4     120 967 325 is correct

I need to replace them all with a term such as aaaaa. Hence, the data should look like:

           notes
1     my number is aaaaa
2        aaaaa is valid
3   why not taking aaaaa
4     aaaaa is correct

Thank you in advance!

CodePudding user response：

Assuming the examples really do cover all possible cases (I would be careful). You can do this with the following regular expression:

\\d{3}( |-)\\d{2,3}( |-)\\d{3}

Here's the code for replacing:

library(dplyr)
library(stringr)

df %>% 
    mutate(
        notes = str_replace_all(notes, '\\d{3}( |-)\\d{2,3}( |-)\\d{3}', 'XXXXXX')
    )

                  notes
1   my number is XXXXXX
2       XXXXXX is valid
3 why not taking XXXXXX
4     XXXXXX is correct