Recoding categorical variables based on multiple 'or' conditions

Asked Sep 30 '14 at 02:53

Active Sep 30 '14 at 02:55

Viewed 46 times

I want to recode a variable called 'booktitle', based on the presence or absence of certain strings. If the field is blank or contains specific 'negative' words, I want to code my new variable "No". If none of the negative words are present, I want to code my new variable "Yes".

I tried the following:

newvar <- ifelse(
  booktitle==c("","n/a","na","none","do not","not really","not committed",
               "not a fan","NOne","not fan"),
  "No",
  "Yes"
)

and got an error message:

Error in $<-.data.frame(*tmp*, "myDF", value = logical(0)) : replacement has 0 rows, data has 411.

I also tried it using '|' instead of commas to represent 'or'.

edited Sep 30 '14 at 02:55

thelatemail

91,185
12
128
188

asked Sep 30 '14 at 02:53

user3614783

You need `%in%` not `==`. This has been asked here many times. – thelatemail Sep 30 '14 at 03:00
1

Also check out this: https://stat.ethz.ch/R-manual/R-devel/library/base/html/grep.html – j1897 Sep 30 '14 at 03:02
Thanks, this seems to work: newvar<-ifelse(grepl("n/a|na|none|do not|not really|not committed|not a fan|NOne|not fan",booktitle),"No","Yes") – user3614783 Oct 01 '14 at 02:07

Recoding categorical variables based on multiple 'or' conditions

0 Answers0