Generalized extreme value distribution constrained to positive shape parameters (GEV+)

Negative GEV shape parameters are considered artifacts from sample uncertainties. The GEV+ provides a fit to data series with shape parameters >= 0. If the L-skewness smaller or equal to ln(9/8)/ln2, a Gumbel distribution is fit, otherwise a GEV. Adapted from 10.1016/j.ejrh.2021.100906

Usage

fgevplus(x, method = c("Lmoments", "MLE"))

Arguments

x: A vector of block maxima.
method: A character string defining the method used for parameter estimation. Defaults to Lmoments.

Value

A list with components:

type: The type of the distribution.
method: Method used to fit the distribution.
params: A named vector with the fitted parameters loc. sacle and shape.
x: The data the distribution was fitted to.
fit: The fitting object, which is either NULL or of class fevd.

References

Moccia, B. et al. (2021) 'Spatial variability of precipitation extremes over Italy using a fine-resolution gridded product', Journal of Hydrology: Regional Studies. Elsevier Ltd, 37 (October). doi: 10.1016/j.ejrh.2021.100906.

Examples

z <- extRemes::revd(100, loc = 20, scale = 0.5, shape = -0.2)
fit <- fgevplus(z)
fit$params[3]
#> shape 
#>     0 

z <- extRemes::revd(100, loc = 20, scale = 0.5, shape = 0.2)
fit <- fgevplus(z)
fit$params[3]
#>      shape 
#> 0.03285595