.BG
.FN quantile
.TL
quantile: Compute Empirical Quantiles
.CS
quantile(x, probs)
.PP
.AG x
vector of data.
.AG probs
vector of desired probability levels.
Values must be between 0 and 1 inclusive.
Defaults to `c(0,.25,.5,.75,1)', producing a "five number summary"
that gives the minimum, lower quartile, median, upper quartile, and maximum
of `x'.
.RT
vector of empirical quantiles corresponding to the
`probs' levels in the sorted `x' data.
.PP
The algorithm linearly interpolates between order statistics of `x',
assuming that the `i'th order statistic is the `(i-.5)/len(x)' quantile.
The algorithm uses partial sorting, hence is quickly able to find
a few quantiles of even large datasets.
.EX
quantile(mydata)	# five number summary
quantile(xxx,c(.33,.67))  # 33% and 67% points of xxx
.KW sorting
.KW math
.WR
