Automated Parameter Estimation for Complex Models

This R package, calibrar, has been designed for the parameter estimation (or calibration) of a wide range of ecological models, including complex and stochastic models. The package combines various optimization functionalities in a single interface, enabling the implementation of the latest advancements in complex model calibration. The package provides support for multiple phases and box constrained optimisation with the possibility of using several algorithms available in R. In particular, by using a “black-box” approach, the package allows the calibration of models implemented in any programming language. It provides a generic interface with models and allows the construction of the objective function, within R, without requiring any changes in the models’ code. Parallel support for computationally intensive models is also provided, and can be used with high performance computing systems in a simple manner, including the capability to restart an unfinished optimisation for models with a long runtime.

It implements generic functions that can be used for fitting any type of models, especially those with non-differentiable objective functions, with the same syntax as base::optim. It supports multiple phases estimation (sequential parameter masking), constrained optimization (bounding box restrictions) and automatic parallel computation of numerical gradients. Some common maximum likelihood estimation methods and automated construction of the objective function from simulated model outputs is provided.
See for more details.


# The easiest way to get calibrar is to install it from CRAN:

# Alternatively, install the stable development version from OSMOSE drat repository:
install.packages("calibrar", repo="")

# Or the development version from GitHub:
# install.packages("remotes")


For a quick introduction, check the worked the examples available from the package:

For a more detailed explanation of the package philosophy, you can read the pre-print calibrar: an R package for fitting complex ecological models.


If you find any bug, have questions about the documentation or requests for enhancements, please open an issue.

Contributions are accepted as pull requests. Please note that the calibrar package is released with a Contributor Code of Conduct. By contributing to this project, you agree to abide by its terms.