1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
|
\name{mc}
\alias{mc}
\alias{mc.default}
\title{Medcouple, a Robust Measure of Skewness}
\description{
Compute the \sQuote{medcouple}, a \emph{robust} concept and estimator
of skewness. The medcouple is defined as a scaled median difference
of the left and right half of distribution, and hence \emph{not} based
on the third moment as the classical skewness.
}
\usage{
mc(x, \dots)
\method{mc}{default}(x, na.rm = FALSE, doReflect = (length(x) <= 100),
eps1 = .Machine$double.eps, eps2 = .Machine$double.xmin,
maxit = 100, trace.lev = 0, full.result = FALSE, \dots)
}
\arguments{
\item{x}{a numeric vector}
\item{na.rm}{logical indicating how missing values (\code{\link{NA}}s)
should be dealt with.}
\item{doReflect}{logical indicating if the internal MC should also be
computed on the \emph{reflected} sample \code{-x}, with final result
\code{(mc.(x) - mc.(-x))/2}. This makes sense since the internal
MC, \code{mc.()} computes the himedian() which can differ slightly from
the median.}%% only whenever sum(x <= med) * sum(x >= med) is even
\item{eps1,eps2}{tolerance in the algorithm; only change with care!}
\item{maxit}{maximul number of iterations; typically a few should be
sufficient.}
\item{trace.lev}{integer specifying how much diagnostic output the
algorithm (in C) should produce. No output by default, most output
for \code{trace.lev = 5}.}
\item{full.result}{logical indicating if the full return values (from
C) should be return, as a list \code{attr(*, "mcComp")}.}
\item{\dots}{potentially further arguments passed to other methods.}
}
% \details{
% ~~ If necessary, more details than the description above ~~
% }
\value{
a number between -1 and 1, which is the medcouple, \eqn{MC(x)}.
For \code{r <- mc(x, full.result = TRUE, ....)}, then
\code{attr(r, "mcComp")} is a list with components
\item{medc}{the medcouple \eqn{mc.(x)}.}
\item{medc2}{the medcouple \eqn{mc.(-x)} if \code{doReflect=TRUE}.}
\item{eps,eps2}{tolerances used.}
\item{iter,iter2}{number of iterations used.}
\item{converged,converged2}{logical specifying \dQuote{convergence}.}
}
\references{
Guy Brys, Mia Hubert and Anja Struyf (2004)
A Robust Measure of Skewness;
\emph{JCGS} \bold{13} (4), 996--1017.
Mia Hubert and E. Vandervieren (2006)
An adjusted boxplot for skewed distributions.
}
\author{Guy Brys; modifications by Tobias Verbeke and bug fixes and
extensions by Martin Maechler.
}
\seealso{\code{\link{Qn}} for a robust measure of scale (aka
\dQuote{dispersion}), ....
}
\examples{
mc(1:5) # 0 for a symmetric sample
x1 <- c(1, 2, 7, 9, 10)
mc(x1) # = -1/3
data(cushny)
mc(cushny) # 0.125
stopifnot(mc(c(-20, -5, -2:2, 5, 20)) == 0,
mc(x1, doReflect=FALSE) == -mc(-x1, doReflect=FALSE),
all.equal(mc(x1, doReflect=FALSE), -1/3, tolerance = 1e-12))
}
\keyword{robust}
\keyword{univar}
|
