This package provides a number of functions in R for the generalised lambda distribution. These are:

`rgl`Random numbers`qgl`Quantiles`pgl`Probabilities - since*F(x)*doesn't exist in closed form for the gld, these are calculated by numerically solving the distribution function (i.e. inverting the quantile function)`dqgl`the density quantile function f(Q(u)) exists in closed form. This function provides it.`dgl`the density function is then provided by using`dqgl`together with`pgl`.`plotgld`Plots density functions for the gld, making use of`dqgl`.`plotglc`Plots distribution functions (a.k.a. cumulative probability functions) for the gld`qqgl`Q-Q plots for the gld.`starship`Estimate the gld using the starship method of King & MacGillivray (1999). There are also diagnostic plots for the fits.

The package works for three parametersations, including the FMKL and the RS, but the default is the
Freimer, Mudholkar, Kollia and Lin (FMKL) parameterisation, which is
*strongly recommended*.

Download the package from from CRAN or email me if you have problems at CRAN.

The package is based on my previously released C code for starship estimation and includes a number of improvements on that implementation.

Changelog

| | | |

| | | |

comments to: robert.king@newcastle.edu.au