ftailgpd: Fit tail GPD to threshold exceedances where scale is smoothly...

In gbstat/tailqr: 'Fit Mixture of Quantile Regression and GPDs'

Description

Fit tail GPD to exceedances where the log(shape) parameter varies as a linear combination of basis functions.

Usage

ftailgpd(y, dates = NULL, df.lf = NULL, df.hf = NULL, Xb = NULL,
  b.formula = ~b.lf + b.hf + b.lf:b.hf, winter.ind = NULL,
  use.winter.shape = FALSE, return.basis = FALSE)

Arguments

dates	vector of dates of each observation
df.lf	natural spline degrees of freedom on date
df.hf	periodic basis spline degrees of freedom on the day of the year
Xb	basis design matrix for tails (log(scale) of GPD). Each basis function is in a separate column.
b.formula	formula for natural spline basis and periodic basis splines. Must be written in terms of b.lf and b.hf
winter.ind	should a separate shape paramter be estimated for the winter months?
return.basis	(logical) should the basis matrix Xb be returned?

Value

list par: estimated parameters (last column corresponds to shape) first several correspond to log(scale) basis terms Xb: basis matrix

gbstat/tailqr documentation built on May 8, 2019, 5:42 p.m.