This is a likelihood model generator based on the convention used in R, where we have a name for a distribution, like normal, weibull, exp, etc. and then we have a set of functions that are named p<name>, d<name>, q<name>, r<name>, etc., where the first letter is the first letter of the distribution name. For instance, for the normal distribution, we have pnorm, dnorm, qnorm, rnorm, respectively for the cdf, pdf, quantile function, and random variate generator.

This function generates a likelihood model based on a distribution name. The distribution name is used to generate the log-likelihood function, the score function, and the Hessian of the log-likelihood function for either exact (pdf), left-censored (cdf), right-censored (1 - cdf), or interval-censored (cdf(b) - cdf(a)) observations.

likelihood_name(dist_name, ob_col, censor_col)

Arguments

dist_name

The name of the distribution.

ob_col

The name of the column in a data frame that contains the observations.

type_col

The name of the column in a data frame that contains the censoring type. The values in this column determine the type of censoring.

• exact or NA: exact observation

• left: left-censored observation

• right: right-censored observation

• interval: interval-censored observation