set factor levels in specific order

Question

How to set levels in specific order? The desired order of levels are - "CA", "IN", "FL" "NY"

library(dplyr)

tbl <- tibble(states = c("FL", "NY", "CA", "IN")) %>%
  mutate(states_fct = factor(states))

score 2 · Accepted Answer · answered Sep 11 '21 at 16:56

2

Use the levels = argument.

tbl <- tibble(states = c("FL", "NY", "CA", "IN")) %>%
  mutate(states_fct = factor(states, levels = c("CA", "IN", "FL", "NY"))

answered Sep 11 '21 at 16:56

Ben Bolker

211,554
25
370
453

score 2 · Answer 2 · answered Sep 11 '21 at 16:57

2

You can use ordered to express that you want an ordered factor:

states = ordered(c("FL", "NY", "CA", "IN"), levels = c("CA", "IN", "FL", "NY"))
str(states)

answered Sep 11 '21 at 16:57

Bernhard

4,272
1
13
23

score 1 · Answer 3 · answered Sep 11 '21 at 17:10

We could use fct_relevel from forcats package:

library(dplyr)
library(forcats)
tbl <- tibble(states = c("FL", "NY", "CA", "IN")) %>%
    mutate(states_fct = factor(states)) %>% 
    mutate(states_fct = fct_relevel(states_fct, c("CA", "IN", "FL", "NY")))

levels(tbl$states_fct)

Output:

  states states_fct
  <chr>  <fct>     
1 FL     FL        
2 NY     NY        
3 CA     CA        
4 IN     IN        

> levels(tbl$states_fct)
[1] "CA" "IN" "FL" "NY"

set factor levels in specific order

3 Answers3