glmProvision: Claims reserving using glm modeling.

In acanadis/provision: Statistical Models for Claim Reserving

Description

Claims reserving using glm modeling.

Usage

glmProvision(lossData, peMethod = "formula",
fam = 1, link = 0, B = 1000, seed = NULL)

Arguments

lossData	Matrix of incremental losses cij, for i = 1,...,k origin years (rows) and for j = 1,...,k development years (columns); filled with NAs for i + j > k.
peMethod	Method to be used, can be formula or bootstrap. Defaults to "formula".
fam	Index of power variance function as defined in tweedie. Defaults to Poisson.
link	Index of power link function as defined in tweedie. Defaults to logarithmic link.
B	Number of iterations to perform in the bootstrapping procedure. Defaults to 1000.
seed	Seed to make bootstrap reproducible.

Value

A list of 5 elements with:

• triangle Input data (lossData).

• glm.triangle Fitted values by glm modeling.

• model glm model.

• bootstrap.losses (Only with peMethod = "bootstrap") Three-dimensional array of lower triangles from bootstrap procedure.

• summary Summary table.

• params List of output parameters.

Examples

data("TaylorData")
glmProvision(lossData = TaylorData$lossData)

acanadis/provision documentation built on May 5, 2019, 3:54 p.m.