CODEWITHSUNDEEP
rnumericalphanumeric
purple_plop
purple_plop

Reputation: 280

Column can only be numerical or alphanumerical values R

Columns can only be numerical or alphanumerical values (i.e. acronym followed by numbers). However, several of these columns have symbols and dates that shouldn't be there. I want the desired output where all values that are not alphanumerical or numerical get flagged as '1'.

i.e.

1234           0
ABC1234        0
#              1
12/13/17       1
$              1
ABC            1

I am looking for code that isn't specific to the ex data above, rather generalized to apply to several columns.

Edited: Clarification

Upvotes: 0

Views: 44

Answers (1)

akrun
akrun

Reputation: 887851

We can use grepl to create a logical output which can be coerced to binary

+(grepl("[^[:alnum:]]", v1))
#[1] 0 0 1 1 1

If it needs to be a letter followed by number

+(!grepl("^[A-Za-z]*\\d+$", v1))
#[1] 0 0 1 1 1

If this is to check every column

df1[] <- lapply(df1, function(x) +(grepl("[^[:alnum:]]", x)))

If the intention is to find any value that are not alpha numeric in a column

v2 <-  sapply(df1, function(x) any(grepl("[^[:alnum:]]", x)))

data

v1 <- c(1234, "ABC1234", "#", "12/13/17", "$")

Upvotes: 2

Related Questions