1

I am trying to add a column called "Visited" which looks at an existing column called "Visits" and if "visits" = NA, then i want "visited" to = 0, but if "visits" > 0, "visited" should = 1. I am getting an error which states "Error in mutate(Visited = if (Visits == "NA") { : object 'Visits' not found". Thank you for all advice!! Here is my code.

mutate(Visited = 
  if(Visits == "NA") {
  replace("NA", 0)
  } else {
  replace(1)
  }
)````
Callahan McGovern
  • 13
  • 2
  • 1
    When you feel that your question has been resolved, please choose one of the questions (based on preference or whatever criteria you choose) and [accept it](https://stackoverflow.com/help/someone-answers); doing so not only provides a little perk to the answerer with some points, but also provides some closure for readers with similar questions. Though you can only accept one answer, you have the option to up-vote as many as you think are helpful. (If there are still issues, you will likely need to edit your question with further details.) – r2evans May 14 '21 at 18:10

3 Answers3

4

Some issues:

  1. You cannot use if in a mutate like this: I'm inferring that your data is more than one row, in which case Visits == "NA" will be a logical vector, length greater than 1. The if conditional must have length-1. What you likely need is a vectorized conditional such as ifelse or replace.

    There are a few things to realize: vectorized conditionals do not short-circuit (&& and || do short-circuit, & and | do not, and you cannot just interchange them); and ifelse has issues with classes other than logical, integer, numeric, and character.

  2. Your use of replace is incorrect: it requires three arguments, it infers nothing. You cannot use just replace(0) hoping that it will know to look for a conditional outside of its call.

  3. There is a big difference between the R symbol NA (which can be numeric, logical, string, etc) and the string "NA". There are times when mis-read data gives you strings of "NA", but typically it's not. Note that NA == . anything is going to be NA (not true/false), since NA can be interpreted as "can be anything" as well as "not applicable". Because of this, if you have NAs in your code, then . == "NA" is going to first internally coerce the data to strings, which does not convert NA to "NA", and then look for the literal "NA", not what you want/need. I hope that makes sense.

  4. The error message suggests that you are not passing in data. mutate(Visited = ...) works fine if the call to mutate is in a dplyr/magrittr "pipe" (%>%), but by itself mutate requires its first argument to be a data.frame, as in mutate(mydata, Visited=...).

Here are some equivalent alternatives that should work for you:

mydata %>%
  mutate(
    Visited1 = ifelse(!is.na(Visits) & Visits > 0, 1, 0),
    Visited2 = replace(rep(1, n()), is.na(Visits) | Visits <= 0, 0),
    Visited3 = +(!is.na(Visits) & Visits > 0)
  )

The third takes advantage of R's coercion from logical to integer with the +(.) shortcut.

You pick which you prefer.

r2evans
  • 108,754
  • 5
  • 72
  • 122
2

ifelse should do the trick. Note: df can be replaced by your dataframes name:

df$Visited = ifelse(is.na(df$Visits), 0, 1)

If you prefer dplyr:

library(dplyr)
df = df %>%
        mutate(Visited = ifelse(is.na(Visits), 0, 1))
bird
  • 2,343
  • 1
  • 4
  • 22
0
library(dplyr)
df %>%
  mutate(Visited = if_else(is.na(Visits), 0, 1))
  Visits Visited
1     NA       0
2      2       1
3      1       1
4     NA       0
5      5       1

Data:

df <- data.frame(
  Visits = c(NA, 2, 1, NA, 5)
)
Chris Ruehlemann
  • 15,379
  • 3
  • 11
  • 27