modeller_from_loglik: Create your own modelling function, given a log-likelihood...
In strongio/keanu: Enter the model.matrix

Description Usage Arguments Value

This function takes a log-likelihood function, and returns a modelling function, much like stats:lm or stats:glm. This modelling function will find coefficients that maximize the likelihood.

1 2	modeller_from_loglik(loglik_fun, penalty_fun = function(...) return(0), non_parameter_args = character())

`loglik_fun`	A function whose first argument is the target variable, and whose subsequent arguments are parameters of the distribution which gives the log-likelihood. (You can indicate arguments that are not parameters with `non_parameter_args`.)
`penalty_fun`	Optional. A function which, when given a dataframe with columns `c('parameter', 'term', 'beta')`, yields a penalty term (a positive number to be added to the function to be minimized by `stats::optim`).
`non_parameter_args`	A character-vector indicating arguments to `loglik_fun` that are not parameters. (This is in addition to the first argument, which is always assumed to be the target variable, not a parameter.)