impute: Data augmentation for the Brown-Resnick process

In lbelzile/mgp: Multivariate generalized Pareto models

Data augmentation for the Brown-Resnick process

Description

This function returns a matrix of data on the data scale (if map == TRUE) or else unit Pareto observations. Data that fall below a marginal threshold are censored and imputed according to the conditional distribution. For riskr == "sum", this require an accept-reject scheme to ensure the points with the simulated components still lies in the risk region defined by the marginal and dependence parameters.

Usage

impute(dat, thresh, mthresh, loc = 1, scale = 1, shape = 1,
  lambdau = 1, riskr = c("max", "sum"), par, map = FALSE, ...)

Arguments

dat	matrix of observations
thresh	functional threshold for the maximum
mthresh	vector of individuals thresholds under which observations are censored
loc	vector of location parameter for the marginal generalized Pareto distribution
scale	vector of scale parameter for the marginal generalized Pareto distribution
shape	vector of shape parameter for the marginal generalized Pareto distribution
lambdau	vector of marginal rate of marginal threshold exceedance.
riskr	string giving the risk region, either max or sum
par	list of parameters: alpha for the logistic model, Lambda for the Brown–Resnick model or else Sigma and df for the extremal Student.
map	logical; should data be returned on the unit Pareto scale? Default to FALSE
...	additional arguments (see Details)

Value

a matrix of observations with imputed values in place of censored components.

lbelzile/mgp documentation built on Aug. 5, 2023, 2:34 a.m.