The R Project SVN R

Rev

Rev 80324 | Show entire file | Ignore whitespace | Details | Blame | Last modification | View Log | RSS feed

Rev 80324 Rev 88376
Line 10... Line 10...
10
    low = FALSE, high = FALSE)
10
    low = FALSE, high = FALSE)
11
}
11
}
12
\alias{mad}
12
\alias{mad}
13
\arguments{
13
\arguments{
14
  \item{x}{a numeric vector.}
14
  \item{x}{a numeric vector.}
15
  \item{center}{Optionally, the centre: defaults to the median.}
15
  \item{center}{optionally, a number, the centre: defaults to the median.}
16
  \item{constant}{scale factor.}
16
  \item{constant}{scale factor.}
17
  \item{na.rm}{if \code{TRUE} then \code{NA} values are stripped
17
  \item{na.rm}{if \code{TRUE} then \code{NA} values are stripped
18
    from \code{x} before computation takes place.}
18
    from \code{x} before computation takes place.}
19
  \item{low}{if \code{TRUE}, compute the \sQuote{lo-median}, i.e., for even
19
  \item{low}{if \code{TRUE}, compute the \sQuote{lo-median} of the absolute
20
    sample size, do not average the two middle values, but take the
20
    deviations, i.e., for even sample size, do not average the two middle
21
    smaller one.}
21
    values, but take the smaller one.}
-
 
22
 
22
  \item{high}{if \code{TRUE}, compute the \sQuote{hi-median}, i.e., take the
23
  \item{high}{if \code{TRUE}, compute the \sQuote{hi-median} of \eqn{|x - <center>|}, i.e., take the
23
    larger of the two middle values for even sample size.}
24
    larger of the two middle values for even sample size.}
24
}
25
}
25
\description{
26
\description{
26
  Compute the median absolute deviation, i.e., the (lo-/hi-) median of
27
  Compute the median absolute deviation, i.e., the (lo-/hi-) median of
27
  the absolute deviations from the median, and (by default) adjust by a
28
  the absolute deviations from the median (or another \code{center}), and (by default) adjust by a
28
  factor for asymptotically normal consistency.
29
  factor for asymptotically normal consistency.
29
}
30
}
30
\details{
31
\details{
31
  The actual value calculated is \code{constant * cMedian(abs(x - center))}
32
  The actual value calculated is \code{constant * cMedian(abs(x - center))}
32
  with the default value of \code{center} being \code{median(x)}, and
33
  with the default value of \code{center} being \code{median(x)}, and