This function estimates the point process characterisation from exceedances above a threshold.
fitpp(data, threshold, noy = length(data) / 365.25, start, ...,
std.err.type = "observed", corr = FALSE, method = "BFGS", warn.inf = TRUE)
This function returns a list with components:
A vector containing the estimated parameters.
A vector containing the standard errors.
A vector containing the parameters of the model that have been held fixed.
A vector containing all parameters (optimized and fixed).
The deviance at the maximum likelihood estimates.
The correlation matrix.
Components taken from the
list returned by optim
- for the mle
method.
The threshold passed to argument threshold
.
The number and proportion of exceedances.
The data passed to the argument data
.
The exceedances, or the maxima of the clusters of exceedances.
The scale parameter for the fitted generalized Pareto distribution.
The standard error type - for 'mle'
only. That is Observed
Information matrix of Fisher.
Logical. Specify if the threshold is a varying one -
'mle'
only. For other methods, threshold is always constant
i.e. var.thresh = FALSE
. Not implemented yet.
A numeric vector.
A numeric value giving the threshold for the GPD.
Numeric. The number of year of observation.
A named list that gives the starting values for the optimization routine. Each list argument must correspond to one parameter to be estimated. May be missing.
Other optional arguments to be passed to the
optim
function, allow hand fixed parameters (only
- see the Note section.
A character string. If "observed", the standard errors are derived from the observed Fisher information matrix. If "none", standard errors are not computed.
Logical. Does the asymptotic correlation matrix has to be
computed? Default is "not computed" - e.g. FALSE
.
A character string specifying which numerical
optimization procedure has to be used. See optim
for
more details.
Logical. If TRUE
(default), users will be
warned if the log-likelihood is not finite at starting values - as
it may cause some problem during the optimation stage.
Mathieu Ribatet
Coles, S. (2001) An Introduction to Statistical Modelling of Extreme Values. Springer Series in Statistics. London.
Embrechts, P and Kluppelberg, C. and Mikosch, T (1997) Modelling Extremal Events for Insurance and Finance. Springers.
Pickands, J. (1975) Statistical Inference Using Extreme Order Statistics. Annals of Statistics. 3:119--131.
x <- rgpd(1000, 0, 1, 0.2)
fitpp(x, 0)
Run the code above in your browser using DataLab