marginalOptim: Fit two 4-parameter log-logistic functions with common baseline

Description

This function is an alternative to non-linear least squares and provides optimization framework with optim function. It is however noticeably slower than NLS methods and can be especially time consuming in large datasets, in particular if bootstrap statistics are calculated.

Usage

marginalOptim(data, transforms = NULL, start, model, ...)

Value

Variance-covariance matrix which is returned by optim

is based on the fact that minimization of sum-of-squared residuals leads essentially to a maximum likelihood estimator and so variance-covariance matrix can be estimated using inverse Hessian evaluated at the optimal parameters. In some cases, so obtained variance-covariance matrix might not be positive-definite which probably means that estimates are unstable because of either a poor choice of initial values or poor properties of the data itself.

Arguments

data: Dose-response dataframe. Marginal data will be extracted from it automatically.
transforms: Transformation functions. If non-null, transforms is a list containing 5 elements, namely biological and power transformations along with their inverse functions and compositeArgs which is a list with argument values shared across the 4 functions. See vignette for more information.
start: Starting parameter values. If not specified, they will be obtained from initialMarginal.
model: List with model parameters. Typically, this is an output from constructFormula.
...: Further parameters passed to optim function