Renaming factor levels

Categorical Data in the Tidyverse

Emily Robinson

Data Scientist

Introduction to FiveThirtyEight dataset

# A tibble: 1,040 x 27
   RespondentID travel_amount      do_recline       height 
          <dbl> <fct>              <fct>            <fct>  
 1   3436139758 Once a year or le... NA               NA     
 2   3434278696 Once a year or le... About half th... "6'3\""
 3   3434275578 Once a year or le... Usually          "5'8\""
 4   3434268208 Once a year or le... Always           "5'11\""
 5   3434250245 Once a month or l... About half th... "5'7\""
# ... with 1,035 more rows, and 23 more variables:
#   children_sub_18 <fct>, middle_arm_rest_three <fct>,
#   middle_arm_rest_two <fct>, window_shade_control <fct>,
#   rude_move_seats <fct>, rude_talk <fct>,
#   times_get_up <fct>, recliner_obligation <fct>,
#   rude_recline <fct>, eliminate_recline <fct>,
#   rude_switch_seats_friend <fct>,
#   rude_switch_seats_family <fct>, rude_bathroom <fct>,
#   rude_walking <fct>, rude_baby <fct>,
#   rude_unruly_children <fct>, personal_electronics <fct>,
#   smoking <fct>, gender <fct>, age <fct>, income <fct>,
#   education <fct>, location <fct>

fct_recode()

levels(flying_etiquette$middle_arm_rest_three)

[1] "Other (please specify)"                                     
[2] "The arm rests should be shared"                             
[3] "The people in the aisle and window seats get both arm rests"
[4] "The person in the middle seat gets both arm rests"          
[5] "Whoever puts their arm on the arm rest first"

ggplot(flying_etiquette, 
  aes(x = fct_infreq(middle_arm_rest_three))) +
    geom_bar() + 
    labs(x = "Arm rest opinions")

Crowded text

ggplot(flying_etiquette, aes(x = fct_infreq(middle_arm_rest_three))) +
    geom_bar() + 
    labs(x = "Arm rest opinions")

A bar chart with count on the y-axis and arm rest opinions on the x-axis. It's impossible to read the x-axis labels because the text runs together.

Extraneous text

ggplot(flying_etiquette, aes(x = fct_infreq(middle_arm_rest_three))) +
    geom_bar() + 
    coord_flip() + 
    labs(x = "Arm rest opinions")

The same bar chart as before, now flipped. The y-axis labels for arm rest opinions are all readable.

Changing with fct_recode()

flying_etiquette %>% 
    mutate(middle_arm_rest_three = fct_recode(middle_arm_rest_three,
    "Other" = "Other (please specify)",
    "Everyone should share" = "The arm rests should be shared",
    "Aisle and window people" = 
    "The people in the aisle and window seats get both arm rests",
    "Middle person" = "The person in the middle seat gets both arm rests",
    "Fastest person" = "Whoever puts their arm on the arm rest first"
  )) %>%
  count(middle_arm_rest_three)

# A tibble: 6 x 2
  middle_arm_rest_three       n
  <fct>                   <int>
1 Everyone should share     587
2 Middle person             119
3 Fastest person             87
4 Other                      45
5 Aisle and window people    18
6 NA                        184

Renaming a couple levels

flying_etiquette %>% 
    mutate(middle_arm_rest_three = fct_recode(middle_arm_rest_three,
    "Everyone should share" = "The arm rests should be shared")) %>%
     count(middle_arm_rest_three)

# A tibble: 6 x 2
  middle_arm_rest_three                                   n
  <fct>                                               <int>
1 Other (please specify)                                 45
2 Everyone should share                                 587
3 The people in the aisle and window seats get both …    18
4 The person in the middle seat gets both arm rests     119
5 Whoever puts their arm on the arm rest first           87
6 NA                                                    184

Renaming unknown levels

flying_etiquette %>% 
    mutate(middle_arm_rest_three = fct_recode(middle_arm_rest_three,
    "Everyone should share" = "arm rests should be share")) %>%
     count(middle_arm_rest_three)

# A tibble: 6 x 2
  middle_arm_rest_three                                   n
  <fct>                                               <int>
1 Other (please specify)                                 45
2 The arm rests should be shared                        587
3 The people in the aisle and window seats get both …    18
4 The person in the middle seat gets both arm rests     119
5 Whoever puts their arm on the arm rest first           87
6 NA                                                    184
Warning message:
Unknown levels in `f`: arm rests should be share

Let's practice!

Categorical Data in the Tidyverse