Class breaks
Finds class breaks in a distribution
classBreaks(x, n, type = c("equal", "quantile", "std", "geometric"))x |
A vector to find breaks for |
n |
Number of breaks |
type |
Statistic used to find breaks c("equal", "quantile", "std", "geometric") |
A vector containing class break values the length is n+1 to allow for specification of ranges
The robust std method uses sqrt(sum(x^2)/(n-1)) to center the data before deriving "pretty" breaks.
Jeffrey S. Evans <jeffrey_evans@tnc.org>
y <- rnbinom(100, 10, 0.5)
classBreaks(y, 10)
classBreaks(y, 10, type="quantile")
opar <- par(no.readonly=TRUE)
par(mfrow=c(2,2))
d <- density(y)
plot(d, type="n", main="Equal Area breaks")
polygon(d, col="cyan")
abline(v=classBreaks(y, 10))
plot(d, type="n", main="Quantile breaks")
polygon(d, col="cyan")
abline(v=classBreaks(y, 10, type="quantile"))
plot(d, type="n", main="Robust Standard Deviation breaks")
polygon(d, col="cyan")
abline(v=classBreaks(y, 10, type="std"))
plot(d, type="n", main="Geometric interval breaks")
polygon(d, col="cyan")
abline(v=classBreaks(y, 10, type="geometric"))
par(opar)
( y.breaks <- classBreaks(y, 10) )
cut(y, y.breaks, include.lowest = TRUE, labels = 1:10)Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.