shorth


minimum fraction of data to be covered by scale estimator.
if Alpha == 0.5, the <code>shorth</code> is calculated


Alpha


generalized length of shortest-half sample


robust

routines for univariate and multivariate outlier detection
with a focus on parametric methods, but support for some
methods based on resistant statistics

Vince Carey

parody

Parametric And Resistant Outlier DYtection

shorth function

minimum fraction of data to be covered by scale estimator.
if Alpha == 0.5, the <code>shorth</code> is calculated

generalized length of shortest-half sample

one-dimensional MVE (min. vol. ellipsoid)

a 2-vector with endpoints of the shortest Alpha-sample

see previous return component
<code>L$shorth[2]-L$shorth[1]</code>

length.shorth

midpt.shorth

mean of values in the shorth, studied by Andrews et al (1972)
as a location estimator

meanshorth

correction factor to be applied to achieve
approximate unbiasedness and diminish small-sample
parity dependence; <code>L["shorth"]] * L[["correction"]]</code> is approximately
unbiased for the Gaussian standard deviation, for
0 < Alpha < 1.

correction.parity.dep

correction factor to be applied along with
correction.parity.dep when Alpha = .5; empirically
derived bias correction useful for 10 < N < 2000 and
possibly beyond.  To use, divide: 
<code>(L[["shorth"]] * L[["correction"]] / L[["bias.corr"]])</code> is approximately
unbiased for Gaussian standard deviation, when Alpha=.5.

shorth: one-dimensional MVE (min. vol. ellipsoid)

Description

Usage

Arguments

Value

References