Variables will be return by the order in which they appear. Even factors are shown by their order of appearance in the vector.
There are 2 methods for counting vectors. The default method uses
base::tabulate() (the workhorse for base::table() with a call to
pseudo_id() to transform all inputs into integers. The logical method
counts TRUE, FALSE and NA values, which is much quicker.
Usage
counts(x, ...)
# S3 method for class 'data.frame'
counts(x, cols, sort = FALSE, ..., .name = "freq")
props(x, ...)
# Default S3 method
props(x, sort = FALSE, na.rm = FALSE, ...)
# S3 method for class 'data.frame'
props(x, cols, sort = FALSE, na.rm = FALSE, ..., .name = "prop")Arguments
- x
A vector or
data.frame- ...
Arguments passed to other methods
- cols
A vector of column names or indexes
- sort
Logical, if
TRUEwill sort values (not counts) before returning. For factors this will sort by factor levels. This has no effect for logical vectors, which already return in the order ofFALSE,TRUE,NA.- .name
The name of the new column
- na.rm
If
TRUEwill remove NA values from proportions
Value
A named vector of integers or doubles (for counts, and props,
respectively) or data.frame with columns for each column chosen and the
.name chosen for the summary
Examples
x <- sample(1:5, 10, TRUE)
counts(x)
#> 5 4 1 2 3
#> 5 2 1 1 1
props(x)
#> 5 4 1 2 3
#> 0.5 0.2 0.1 0.1 0.1
x <- quick_df(list(
a = c("a", "c", "a", "c", "d", "b"),
b = c("a", "a", "a", "c", "c", "b"),
c = c("a", "a", "a", "c", "b", "b")
))
counts(x, "a")
#> a freq
#> 1 a 2
#> 2 c 2
#> 3 d 1
#> 4 b 1
counts(x, c("a", "b", "c"))
#> a b c freq
#> 1 a a a 2
#> 2 c a a 1
#> 3 c c c 1
#> 4 d c b 1
#> 5 b b b 1
props(x, 2)
#> b prop
#> 1 a 0.5000000
#> 2 c 0.3333333
#> 3 b 0.1666667
props(x, 1:3)
#> a b c prop
#> 1 a a a 0.4
#> 2 c a a 0.2
#> 3 c c c 0.2
#> 4 d c b 0.2
#> 5 b b b 0.2
props(c(1, 1, 3, NA, 4))
#> 1 3 4 <NA>
#> 0.4 0.2 0.2 0.2
props(c(1, 1, 3, NA, 4), na.rm = TRUE)
#> 1 3 4 <NA>
#> 0.50 0.25 0.25 NA