How can I swap the levels?

Question

I have the following code in R:

a <- c("No", "Yes", "No", "Yes", "Yes")
af <- as.factor(a)

Then

levels(af)

Returns

[1] "No" "Yes"

The problem is that caret::confusionMatrix considers the first factor as the positive one, if you don't pass a positive argument.

How can I swap the levels, so that the first one is "Yes"?

Maybe a duplicate of [Reorder levels of a factor without changing order of values](https://stackoverflow.com/q/2375587/4752675) — G5W, May 29 '21 at 18:05

score 1 · Accepted Answer · answered May 29 '21 at 18:04

1

Use the factor() function:

a <- c("No", "Yes", "No", "Yes", "Yes")
af <- factor(a, levels = c("Yes","No"))

answered May 29 '21 at 18:04

VitaminB16

1,174
1
3
17

score 0 · Answer 2 · answered May 29 '21 at 18:09

0

Use forcats library

library(forcats)
a <- c("No", "Yes", "No", "Yes", "Yes")
af <- as.factor(a)

af
[1] No  Yes No  Yes Yes
Levels: No Yes

fct_recode(af, Yes = "No", No = "Yes")
[1] Yes No  Yes No  No 
Levels: Yes No

answered May 29 '21 at 18:09

Pete Kittinun

593
3
15

I don't think this is right. You swapped all the values from Yes to No and viceversa. I want to swap the order after `Levels: `, I think. I don't want to physically change the values inside the variable – robertspierre May 29 '21 at 18:15
ok, I misunderstood the question. – Pete Kittinun May 29 '21 at 18:18
I think `caret::confusionMatrix` somehow considers the order in `Levels: ` – robertspierre May 29 '21 at 18:21

How can I swap the levels?

2 Answers2