Question

Note: The title may be misleading. If you understand my problem and think of something more descriptive – please change it.

I’ve got a strange situation where the responses from a survey are all character, rather than numeric. It seems that R, really doesn’t like this. Let’s say I asked a question:

Q. In what area do you work? 
East
West
Central
North
South
None of the above

But respondents were only from the east, west and central.

dat <- rep(c("East", "West", "Central"),100)

Now, for presentation purposes, it’s important that I include North, south and None of the above, even if they are none. However, factoring those elements in is challenging.

Let’s try:

fac1 <- factor(dat, labels=c("East","West","Central","North","South","None of the above"))

Error in factor(dat, labels = c("East", "West", "Central", "North", "South",  : 
  invalid labels; length 6 should be 1 or 3

Basically, what i’d like to do is factor this data with the missing values. So that when I type something like summary(fac1) it shows them having 0 responses in that category.

There has to be an easier way to do this!

Answer 1

Almost there. You need to use the levels argument:

fac1 <- factor(dat, levels=c("East","West","Central","North","South","None of the above"))
str(fac1)
 Factor w/ 6 levels "East","West",..: 1 2 3 1 2 3 1 2 3 1 ...

---

The difference between levels and labels is this:

• levels defines the factor levels in your data
• labels allows you to rename the factor levels in one go.

For example:

fac2 <- factor(
  dat, 
  levels=c("East","West","Central","North","South","None of the above"),
  labels=c("E", "W", "C", "N", "S", "Other")
)
str(fac2)
Factor w/ 6 levels "E","W","C","N",..: 1 2 3 1 2 3 1 2 3 1 ...