Skip to contents

Distribution of the r-largest observations

Source: R/univdist.R
rlarg.Rd

Likelihood, score function and information matrix for the r-largest observations likelihood.

Arguments

par

vector of loc, scale and shape

dat

an n by r sample matrix, ordered from largest to smallest in each row

method

string indicating whether to use the expected ('exp') or the observed ('obs' - the default) information matrix.

nobs

number of observations for the expected information matrix. Default to nrow(dat) if dat is provided.

r

number of order statistics kept. Default to ncol(dat)

Usage


rlarg.ll(par, dat, u, np)
rlarg.score(par, dat)
rlarg.infomat(par, dat, method = c('obs', 'exp'), nobs = nrow(dat), r = ncol(dat))

Functions

  • rlarg.ll: log likelihood

  • rlarg.score: score vector

  • rlarg.infomat: observed or expected information matrix

References

Coles, S. (2001). An Introduction to Statistical Modeling of Extreme Values, Springer, 209 p.

Smith, R.L. (1986). Extreme value theory based on the r largest annual events, Journal of Hydrology, 86(1-2), 27--43, http://dx.doi.org/10.1016/0022-1694(86)90004-1.

Author

Leo Belzile