Lump together least/most common factor levels into "other"

Source: R/lump.R

Lump together least/most common factor levels into "other"

fct_lump(f, n, prop, w = NULL, other_level = "Other",
  ties.method = c("min", "average", "first", "last", "random", "max"))

fct_lump_min(f, min, w = NULL, other_level = "Other")

Arguments

f

A factor (or character vector).
n, prop

If both n and prop are missing, fct_lump lumps together the least frequent levels into "other", while ensuring that "other" is still the smallest level. It's particularly useful in conjunction with fct_inorder().

Positive n preserves the most common n values. Negative n preserves the least common -n values. It there are ties, you will get at least abs(n) values.

Positive prop preserves values that appear at least prop of the time. Negative prop preserves values that appear at most -prop of the time.
w

An optional numeric vector giving weights for frequency of each value (not level) in f.
other_level

Value of level used for "other" values. Always placed at end of levels.
ties.method

A character string specifying how ties are treated. See rank() for details.
min

Preserves values that appear at least min number of times.

See also

fct_other() to convert specified levels to other.

Examples

x <- factor(rep(LETTERS[1:9], times = c(40, 10, 5, 27, 1, 1, 1, 1, 1)))
x %>% table()
#> .
#>  A  B  C  D  E  F  G  H  I 
#> 40 10  5 27  1  1  1  1  1 
x %>% fct_lump() %>% table()
#> .
#>     A     D Other 
#>    40    27    20 
x %>% fct_lump() %>% fct_inorder() %>% table()
#> .
#>     A Other     D 
#>    40    20    27 

x <- factor(letters[rpois(100, 5)])
x
#>   [1] d c h f d g h f f d c b b f c h h e h h c f e f d b e c b f e c h d b f e
#>  [38] e f d c g f g d d e c a b d c g e f d f f g g e c f b g g b g e g i e d g
#>  [75] g d c d i d c b d c d d f f d e d h d d e f f h f e
#> Levels: a b c d e f g h i
table(x)
#> x
#>  a  b  c  d  e  f  g  h  i 
#>  1  9 13 21 14 19 12  9  2 
table(fct_lump(x))
#> 
#>     b     c     d     e     f     g     h Other 
#>     9    13    21    14    19    12     9     3 

# Use positive values to collapse the rarest
fct_lump(x, n = 3)
#>   [1] d     Other Other f     d     Other Other f     f     d     Other Other
#>  [13] Other f     Other Other Other e     Other Other Other f     e     f    
#>  [25] d     Other e     Other Other f     e     Other Other d     Other f    
#>  [37] e     e     f     d     Other Other f     Other d     d     e     Other
#>  [49] Other Other d     Other Other e     f     d     f     f     Other Other
#>  [61] e     Other f     Other Other Other Other Other e     Other Other e    
#>  [73] d     Other Other d     Other d     Other d     Other Other d     Other
#>  [85] d     d     f     f     d     e     d     Other d     d     e     f    
#>  [97] f     Other f     e    
#> Levels: d e f Other
fct_lump(x, prop = 0.1)
#>   [1] d     c     Other f     d     g     Other f     f     d     c     Other
#>  [13] Other f     c     Other Other e     Other Other c     f     e     f    
#>  [25] d     Other e     c     Other f     e     c     Other d     Other f    
#>  [37] e     e     f     d     c     g     f     g     d     d     e     c    
#>  [49] Other Other d     c     g     e     f     d     f     f     g     g    
#>  [61] e     c     f     Other g     g     Other g     e     g     Other e    
#>  [73] d     g     g     d     c     d     Other d     c     Other d     c    
#>  [85] d     d     f     f     d     e     d     Other d     d     e     f    
#>  [97] f     Other f     e    
#> Levels: c d e f g Other

# Use negative values to collapse the most common
fct_lump(x, n = -3)
#>   [1] Other Other h     Other Other Other h     Other Other Other Other b    
#>  [13] b     Other Other h     h     Other h     h     Other Other Other Other
#>  [25] Other b     Other Other b     Other Other Other h     Other b     Other
#>  [37] Other Other Other Other Other Other Other Other Other Other Other Other
#>  [49] a     b     Other Other Other Other Other Other Other Other Other Other
#>  [61] Other Other Other b     Other Other b     Other Other Other i     Other
#>  [73] Other Other Other Other Other Other i     Other Other b     Other Other
#>  [85] Other Other Other Other Other Other Other h     Other Other Other Other
#>  [97] Other h     Other Other
#> Levels: a b h i Other
fct_lump(x, prop = -0.1)
#>   [1] Other Other h     Other Other Other h     Other Other Other Other b    
#>  [13] b     Other Other h     h     Other h     h     Other Other Other Other
#>  [25] Other b     Other Other b     Other Other Other h     Other b     Other
#>  [37] Other Other Other Other Other Other Other Other Other Other Other Other
#>  [49] a     b     Other Other Other Other Other Other Other Other Other Other
#>  [61] Other Other Other b     Other Other b     Other Other Other i     Other
#>  [73] Other Other Other Other Other Other i     Other Other b     Other Other
#>  [85] Other Other Other Other Other Other Other h     Other Other Other Other
#>  [97] Other h     Other Other
#> Levels: a b h i Other

# Use weighted frequencies
w <- c(rep(2, 50), rep(1, 50))
fct_lump(x, n = 5, w = w)
#>   [1] d     c     h     f     d     Other h     f     f     d     c     Other
#>  [13] Other f     c     h     h     e     h     h     c     f     e     f    
#>  [25] d     Other e     c     Other f     e     c     h     d     Other f    
#>  [37] e     e     f     d     c     Other f     Other d     d     e     c    
#>  [49] Other Other d     c     Other e     f     d     f     f     Other Other
#>  [61] e     c     f     Other Other Other Other Other e     Other Other e    
#>  [73] d     Other Other d     c     d     Other d     c     Other d     c    
#>  [85] d     d     f     f     d     e     d     h     d     d     e     f    
#>  [97] f     h     f     e    
#> Levels: c d e f h Other

# Use ties.method to control how tied factors are collapsed
fct_lump(x, n = 6)
#>   [1] d     c     h     f     d     g     h     f     f     d     c     b    
#>  [13] b     f     c     h     h     e     h     h     c     f     e     f    
#>  [25] d     b     e     c     b     f     e     c     h     d     b     f    
#>  [37] e     e     f     d     c     g     f     g     d     d     e     c    
#>  [49] Other b     d     c     g     e     f     d     f     f     g     g    
#>  [61] e     c     f     b     g     g     b     g     e     g     Other e    
#>  [73] d     g     g     d     c     d     Other d     c     b     d     c    
#>  [85] d     d     f     f     d     e     d     h     d     d     e     f    
#>  [97] f     h     f     e    
#> Levels: b c d e f g h Other
fct_lump(x, n = 6, ties.method = "max")
#>   [1] d     c     Other f     d     g     Other f     f     d     c     Other
#>  [13] Other f     c     Other Other e     Other Other c     f     e     f    
#>  [25] d     Other e     c     Other f     e     c     Other d     Other f    
#>  [37] e     e     f     d     c     g     f     g     d     d     e     c    
#>  [49] Other Other d     c     g     e     f     d     f     f     g     g    
#>  [61] e     c     f     Other g     g     Other g     e     g     Other e    
#>  [73] d     g     g     d     c     d     Other d     c     Other d     c    
#>  [85] d     d     f     f     d     e     d     Other d     d     e     f    
#>  [97] f     Other f     e    
#> Levels: c d e f g Other

x <- factor(letters[rpois(100, 5)])
fct_lump_min(x, min = 10)
#>   [1] g     d     f     f     e     c     f     e     f     c     d     f    
#>  [13] e     g     e     c     g     f     c     d     c     c     f     e    
#>  [25] c     d     c     Other e     Other Other c     c     f     c     d    
#>  [37] f     c     g     f     e     Other d     Other c     Other e     d    
#>  [49] e     d     Other g     Other f     e     e     d     Other Other Other
#>  [61] g     Other e     Other d     f     d     g     f     d     c     Other
#>  [73] e     f     d     f     d     c     d     g     c     e     f     c    
#>  [85] f     Other Other g     d     g     g     f     c     c     d     Other
#>  [97] g     d     Other d    
#> Levels: c d e f g Other