log_lik: Calculate the log-likelihood

In kmlcov: Clustering longitudinal data using the likelihood as a metric of distance

Description

The log-likelihood is calculated with taking into account the type of data ('gaussian', 'binomial', ... etc) and the link function.

Usage

  log_lik(y, n, mu, wt, family, nparam, disp_mod)

Arguments

y	Observed values.
n	Vector of '1's and same length as y.
mu	Predicted values.
wt	Weights.
family	An object of class family.
nparam	Number of parameters of the model.
disp_mod	Dispersion of the 'glm' model.

Details

This function calculates the log-likelihood for the exponential family, it uses the 'AIC' function to realise this operatin.

Value

The log-likelihood of an individual (trajectory).

Note

Meant to be used internally.

kmlcov documentation built on May 2, 2019, 11:38 a.m.