fit.hgld: Fit the Hurdle Generalized Lambda Distribution
In dmarcondes/HGLD: Fit the Hurdle Generalized Lambda Distribution

Description Usage Arguments Details Value References Examples

Fit the Hurdle Generalized Lambda Distribution to a dataset by the Numerical Maximum Likelihood Method.

fit.hgld(data, mixture = FALSE, clustering.m = clara,
  threshold = NULL, leap1 = 3, leap2 = 3, fun1 = "runif.sobol",
  fun2 = "runif.sobol", rs.leap = 3, fmkl.leap = 3,
  rs.init = c(-1.5, 1.5), fmkl.init = c(-0.25, 1.5),
  FUN = "runif.sobol", no = 10000)

`data`	A vector of data.
`mixture`	Whether a mixture of HGLD must be fitted.
`clustering.m`	Clustering method used in classifying the dataset into two parts when fitting a mixture of HGLD. Valid arguments include clara, fanny and pam from the cluster library, or threshold, if the data must be divided by a threshold. Default is clara. Or a logical vector specifying how data should be split. See fun.auto.bimodal.pml for more details.
`threshold`	The threshold to divide the data if clustering.m = "threshold".
`leap1`	Scrambling (0,1,2,3) for the sobol sequence for the first distribution fit when fitting a mixture of HGLD. See scrambling/leap argument for runif.sobol, runif.halton or QUnif of the GLDEX package.
`leap2`	Scrambling (0,1,2,3) for the sobol sequence for the second distribution fit when fitting a mixture of HGLD. See scrambling/leap argument for runif.sobol, runif.halton or QUnif of the GLDEX package.
`fun1`	A character string of either "runif.sobol" (default), "runif.halton" or "QUnif" for the first distribution fit when fitting a mixture of HGLD. See fun.auto.bimodal.pml for more details.
`fun2`	A character string of either "runif.sobol" (default), "runif.halton" or "QUnif" for the second distribution fit when fitting a mixture of HGLD. See fun.auto.bimodal.pml for more details.
`rs.leap`	Scrambling (0,1,2,3) for the sobol sequence for the RS generalized lambda distribution fit. See scrambling/leap argument for runif.sobol, runif.halton or QUnif of the GLDEX package. See fun.data.fit.ml for more details.
`fmkl.leap`	Scrambling (0,1,2,3) for the sobol sequence for the fmkl generalized lambda distribution fit. See scrambling/leap argument for runif.sobol, runif.halton or QUnif of the GLDEX package. See fun.data.fit.ml for more details.
`rs.init`	Initial values (lambda3 and lambda4) for the RS generalized lambda distribution. See fun.data.fit.ml for more details.
`fmkl.init`	Initial values (lambda3 and lambda4) for the fmkl generalized lambda distribution. See fun.data.fit.ml for more details.
`FUN`	A character string of either "runif.sobol" (default), "runif.halton" or "QUnif". See fun.data.fit.ml for more details.
`no`	Number of initial random values to find the best initial values for optimization. See fun.data.fit.ml for more details.

Given a dataset, estimate by the Numerical Maximum Likelihood Method the five parameters of the HGLD. Fit both the RS and the fmkl parametrizations. Also fit a mixture of HGLDs.

`par`	The estimate of the HGLD five parameters for both the RS and fmkl parametrizations if mixture = FALSE. Otherwise present the estimation of ten parameters: the zero probability mass, the four parameters of each GLD and the clustering parameter p.
`data`	The data used in the fit.
`mixture`	Whether a mixture of HGLD was fitted.

Marcondes, D.; Peixoto, C.; Maia, A. C.; A Survey of a Hurdle Model for Heavy-Tailed Data Based on the Generalized Lambda Distribution. (2017) arxiv1712.02183

Su, S.; Fitting Single and Mixture of Generalized Lambda Distributions to Data via Discretized and Maximum Likelihood Methods: GLDEX in R.(2007), Journal of Statistical Software: *21* 9.

set.seed(100)
data <- healthcare[sample(1:nrow(healthcare),30),]
fit <- fit.hgld(data$log_expense)

#mixture
set.seed(100)
data <- c(rcauchy(20,location = 10),rep(0,10),rcauchy(20))
fit2 <- fit.hgld(data = data,mixture = TRUE)