pareto2.goftest: Goodness-of-fit test for the Pareto-II distribution

Description

Performs goodness-of-fit test for the Pareto-II distribution basing on MLE or MMSE estimates (Zhang, Stevens, 2009) and the Anderson-Darling or Kolmogorov test.

Usage

pareto2.goftest(x, k, s, method=c("anderson-darling", "kolmogorov"))

Arguments

a non-negative numeric vector of data values.

scale parameter, $k>0$ or NULL.

shape parameter, $s>0$ or NULL.

method

either "anderson-darling" or "kolmogorov".

Value

The list of class htest with the following components is passed as a result: ll{ statistic the value of the test statistic. p.value the p-value of the test. alternative a character string describing the alternative hypothesis. method a character string indicating what type of test was performed. data.name a character string giving the name(s) of the data. }

Details

This method, proposed e.g. by Zhang and Stevens (2009), uses either the function ad.test from package ADGofTest or ks.test to compute the selected test.

It is known that the tests have low powers.

If k and s are NULL, it bases on bayesian MMS estimators, see pareto2.zsestimate. If s is not NULL, then the unbiased maximum likelihood estimator is used to determine the scale parameter (see pareto2.mlekestimate) iff it is not given.

References

Zhang J., Stevens M.A., A New and Efficient Estimation Method for the Generalized Pareto Distribution, Technometrics 51(3), 2009, 316-325.