frange: Fast implementation of range.
In hadley/ggstat: Statistical Computations for Visualisation

Description Usage Arguments Value Examples

View source: R/frange.R

This is an efficient C++ implementation of range for numeric vectors: it avoids S3 dispatch, and computes both min and max in a single pass through the input.

1	frange(x, finite = TRUE)

`x`	A numeric vector, or an S3 object built on top of integer and double atomic vectors, as long as there is a method for `restore()`.
`finite`	If `TRUE` ignores missing values and infinities. Note that if the vector is empty, or only contains missing values, `frange` will return `c(Inf, -Inf)` because those are the identity values for `min` and `max` respectively.

A numeric vector of length two. If x contains no non-missing values, this will be c(Inf, -Inf).

# frange() always ignores missing values
frange(c(1:10, NA))

# frange() works with dates and times
frange(Sys.Date())
frange(Sys.time())

# frange() is much faster than range()
x <- runif(1e6)
system.time(range(x))
system.time(frange(x))