mlePP-class: Class '"mlePP"' for results of maximum likelihood estimation...

In NHPoisson: Modelling and Validation of Non Homogeneous Poisson Processes

Description

This class encapsulates the output from the maximum likelihood estimation of a Poisson process where the intensity is modeled as a linear function of covariates.

Objects from the Class

Objects can be created by calls of the form new("mlePP", ...), but most often as the result of a call to fitPP.fun.

Slots

call:

Object of class "language". The call to fitPP.fun.

coef:

Object of class "numeric". The estimated coefficientes of the model.

fullcoef:

Object of class "numeric". The full coefficient vector, including the fixed parameters of the model. It has an attribute, called 'TypeCoeff' which shows the names of the fixed parameters.

vcov:

Object of class "matrix". Approximate variance-covariance matrix of the estimated coefficients. It has an attribute, called 'CalMethod' which shows the method used to calcualte the inverse of the information matrix: 'Solve function', 'Cholesky', 'Not possible' or 'Not required' if modCI=FALSE.

min:

Object of class "numeric". Minimum value of objective function, that is the negative of the loglikelihood function.

details:

Object of class "list". The output returned from optim. If nlminb is used to minimize the function, it is NULL.

minuslogl:

Object of class "function". The negative of the loglikelihood function.

nobs:

Object of class "integer". The number of observations.

method:

Object of class "character". It is a bit different from the slot in the extended class mle: here, it is the input argument minfun of fitPP.fun instead of the method used in optim (this information already appears in details).

detailsb:

Object of class "list".The output returned from nlminb. If optim is used to minimize the function, it is NULL.

npar:

Object of class "integer". Number of estimated parameters.

inddat:

Object of class "numeric". Input argument of fitPP.fun.

lambdafit:

Object of class "numeric". Vector of the fitted intensity \hat λ(t).

LIlambda:

Object of class "numeric". Vector of lower limits of the CI.

UIlambda:

Object of class "numeric". Vector of upper limits of the CI.

convergence:

Object of class "integer". A code of convergence. 0 indicates successful convergence.

posE:

Object of class "numeric". Input argument of fitPP.fun.

covariates:

Object of class "matrix". Input argument of fitPP.fun.

tit:

Object of class "character". Input argument of fitPP.fun.

tind:

Object of class "logical". Input argument of fitPP.fun.

t:

Object of class "numeric". Input argument of fitPP.fun.

Extends

Class "mle", directly.

Methods

Most of the S4 methods in stats4 for the S4-class mle can be used. Also a mle method for the generic function extractAIC and a version of the profile mle method adapted to the mlePP objects are available:

coef

signature(object = "mle")

logLik

signature(object = "mle")

nobs

signature(object = "mle")

show

signature(object = "mle")

summary

signature(object = "mle")

update

signature(object = "mle")

vcov

signature(object = "mle")

confint

signature(object = "mle")

extractAIC

signature(object = "mle")

profile

signature(fitted = "mlePP")

Some other generic functions related to fitted models, such as AIC or BIC, can also be applied to mlePP objects.

Note

Let us remind that, as in all the S4-classes, the symbol @ must be used instead of $ to name the slots: mlePP@covariates, mlepp@lambdafit, etc.

See Also

fitPP.fun, mle

Examples

showClass("mlePP")

NHPoisson documentation built on Feb. 19, 2020, 5:07 p.m.