inla.likelihood {INLA}R Documentation

Compute log-likelihood and CDF for new data using the likeilhood from a inla-object

Description

This function computes the log-likeilhood and computes the CDF for new data using the likelihood from a inla-object

Usage

     inla.likelihood(y, type = c("loglik", "CDF"), linear.predictor = numeric(0), 
                     family = "gaussian", theta = NULL, E = 1, scale = 1,  
                     Ntrials = 1,  strata = 1,
                     link.model = "default", link.covariates = NULL)
 

Arguments

y

Data point/vector

type

Return log-liklihood or the CDF?

linear.predictor

A vector of linear predictors for the liklihood/CDF to be evaluated at

family

The family

theta

Vector of hyperparameters in the internal scale, that goes into this likelihood

E

Constrant E for Poisson-like likehoods

scale

scaling, like for Gaussian likehood

Ntrails

Number of trials in Binomial-like likelihood

link.model

The link-model used

link.covariates

Possible covariates that goes into the link.model

Details

Details goes here

Value

value goes here

Author(s)

Havard Rue hrue@r-inla.org


[Package INLA version 22.11.22 Index]