evmr: Extreme Value Modeling for r-Largest Order Statistics
Tools for extreme value modeling based on the r-largest
order statistics framework. The package provides functions for
parameter estimation via maximum likelihood, return level
estimation with standard errors, profile likelihood-based
confidence intervals, random sample generation, and entropy
difference tests for selecting the number of order statistics r.
Several r-largest order statistics models are implemented,
including the four-parameter kappa (rK4D), generalized logistic
(rGLO), generalized Gumbel (rGGD), logistic (rLD), and Gumbel
(rGD) distributions. The rK4D methodology is described in
Shin et al. (2022) <doi:10.1016/j.wace.2022.100533>, the rGLO
model in Shin and Park (2024) <doi:10.1007/s00477-023-02642-7>,
and the rGGD model in Shin and Park (2025)
<doi:10.1038/s41598-024-83273-y>. The underlying distributions
are related to the kappa distribution of Hosking (1994)
<doi:10.1017/CBO9780511529443>, the generalized logistic
distribution discussed by Ahmad et al. (1988)
<doi:10.1016/0022-1694(88)90015-7>, and the generalized Gumbel
distribution of Jeong et al. (2014)
<doi:10.1007/s00477-014-0865-8>. Penalized likelihood approaches
for extreme value estimation follow Martins and Stedinger (2000)
<doi:10.1029/1999WR900330> and Coles and Dixon (1999)
<doi:10.1023/A:1009905222644>. Selection of r is supported using
methods discussed in Bader et al. (2017)
<doi:10.1007/s11222-016-9697-3>. The package is intended for
hydrological, climatological, and environmental extreme value
analysis.
Documentation:
Downloads:
Linking:
Please use the canonical form
https://CRAN.R-project.org/package=evmr
to link to this page.
These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.
[aut,
cre],
Jeong-Soo Park