ksEstimator: Generic Function for the Computation of the Kolmogorov...

In ROptEstOld: Optimally Robust Estimation - Old Version

Description

Generic function for the computation of the Kolmogorov(-Smirnov) minimum distance estimator.

Usage

ksEstimator(x, distribution, ...)

## S4 method for signature 'numeric,Binom'
ksEstimator(x, distribution, param, eps = .Machine$double.eps^0.5)

## S4 method for signature 'numeric,Pois'
ksEstimator(x, distribution, param, eps = .Machine$double.eps^0.5)

## S4 method for signature 'numeric,Norm'
ksEstimator(x, distribution, param, eps = .Machine$double.eps^0.5)

## S4 method for signature 'numeric,Lnorm'
ksEstimator(x, distribution, param, eps = .Machine$double.eps^0.5)

## S4 method for signature 'numeric,Gumbel'
ksEstimator(x, distribution, param, eps = .Machine$double.eps^0.5)

## S4 method for signature 'numeric,Exp'
ksEstimator(x, distribution, param, eps = .Machine$double.eps^0.5)

## S4 method for signature 'numeric,Gammad'
ksEstimator(x, distribution, param, eps = .Machine$double.eps^0.5)

Arguments

x	sample
distribution	object of class "Distribution"
...	additional parameters
param	name of the unknown parameter. If missing all parameters of the corresponding distribution are estimated.
eps	the desired accuracy (convergence tolerance).

Details

In case of discrete distributions the Kolmogorov distance is computed and the parameters which lead to the minimum distance are returned. In case of absolutely continuous distributions ks.test is called and the parameters which minimize the corresponding test statistic are returned.

Value

The Kolmogorov minimum distance estimator is computed. Returns a list with components named like the parameters of distribution.

Methods

x = "numeric", distribution = "Binom"

Binomial distributions.

x = "numeric", distribution = "Pois"

Poisson distributions.

x = "numeric", distribution = "Norm"

Normal distributions.

x = "numeric", distribution = "Lnorm"

Lognormal distributions.

x = "numeric", distribution = "Gumbel"

Gumbel distributions.

x = "numeric", distribution = "Exp"

Exponential distributions.

x = "numeric", distribution = "Gamma"

Gamma distributions.

Author(s)

Matthias Kohl [e-mail]

References

Rieder, H. (1994) Robust Asymptotic Statistics. New York: Springer.

Kohl, M. (2005) Numerical Contributions to the Asymptotic Theory of Robustness. Bayreuth: Dissertation.

See Also

Distribution-class

Examples

x <- rnorm(100, mean = 1, sd = 2)
ksEstimator(x=x, distribution = Norm()) # estimate mean and sd
ksEstimator(x=x, distribution = Norm(mean = 1), param = "sd") # estimate sd
ksEstimator(x=x, distribution = Norm(sd = 2), param = "mean") # estimate mean
mean(x)
median(x)
sd(x)
mad(x)

ROptEstOld documentation built on May 2, 2019, 12:51 p.m.