lci: Create a Lee-Carter model for improvment rates

In amvillegas/iMoMo: Mortality Improvement Rate Modelling

Description

Utility function to initialise a iMoMo object representing a Lee-Carter model on improvment rates.

Usage

lci(type = c("indirect", "direct"), const = c("sum", "last", "first"))

Arguments

type	defines the type of estimation method to be used. "indirect" would estimate the equivalent mortality rate model and the transform the model into improvement rates. "direct" would estimate the model directly on the improvement rate model.
const	defines the constraint to impose to the period index of the model to ensure identifiability. The alternatives are "sum"(default), "last" and "first" which apply constraints ∑_tκ_t = 0, κ_n = 0 and κ_1 = 0, respectively.

Details

The created model has predictor structure

\log(μ_{xt}/μ_{x,t-1}) = α_x + β_xκ_t.

To ensure identifiability one of the following constraints is imposed

∑_tκ_t = 0,\,κ_1 = 0,\, κ_n = 0

depending on the value of const, and

∑_xβ_x = X.

where X is the number of ages in the data

Value

An object of class "StMoMo".

See Also

iMoMo

amvillegas/iMoMo documentation built on Sept. 18, 2020, 11:25 p.m.