mortalityModel-class: mortalityModel
In dfeehan/mortfit: Fit and evaluate mortality models

Description Details

this object represents a likelihood function for use with mortality data. generally, it will be the result of combining a likelihood with a hazard; for example, it could be the binomial model with the gompertz hazard embedded in it. many of the slots of this object correspond to those found in mortalityHazard-class

name

the model's name

loglik.fn

the log-likelihood function, which takes as its arguments

theta: a vector of parameter values
Dx: vector of deaths by age
Nx: vector of exposure by age
ages: vector of ages that corresponds to the entries of Nx and Dx

num.param

the number of parameters in the model

theta.default

default values for the parameters

theta.range

if not NULL, a range of plausible values that the parameters can assume

optim.default

a list of default parameters to pass into R's optim function

predict.fn

a function which, given values for the parameters and a dataset of exposures, produces predicted numbers of deaths by age. it takes as its arguments:

theta: a vector of parameter values
Nx: vector of exposure by age
age: vector of ages that the Nx correspond to
age.offset: the offset used for age. (See mortalityData-class.)

theta.start.fn

if not NULL, a function which takes a mortalityData object and returns very preliminary estimates of the parameters which can then be used in an optimization routine

simulate.fn

if not NULL, a function which takes as its arguments:

theta: a vector of parameter values
Nx: vector of exposure by age
age: vector of ages that the Nx correspond to
seed: if not NULL, the value to seed the pseudorandom number generator with
means: TODO

eval.fn

if not NULL, a function which produces an evaluation of the model's fit to a given dataset. it takes as its arguments:

fit.obj: the mortalityFit object that resulted from fitting the model to a dataset
observed: the mortalityData object with the observed data
fittedValues: the mortalityPrediction object with fitted values, usually resulting from a call to predict.fn

hazard