Evaluating and improving open source software for nonlinear statistical modeling in ecology

The National Center for Ecological Analysis and Synthesis (NCEAS) recently began this project to compare and evaluate three open source statistical software packages (OpenBUGS, AD Model Builder, and R).  The resulting analysis will be published as a web resource and through peer-reviewed literature.  Stay tuned for more information as results are released!



Increasingly, non‐linear and complex models are applied as a tool for improving understanding of ecological systems. These statistical models are often used to test hypotheses and make inferences about ecological theories and management decisions based on available data. This explosion in the application of such models is due to rapid and current development of methodology to carryout statistical inference of complex nonlinear models and improvements in computer power (faster and multiple processors). While there are many tools available for statistical inference that differ in their effectiveness for specific applications, no formal comparisons have been conducted between various software packages. It is therefore important to identify which tools are most appropriate for given applications and to demonstrate how such tools can be used most effectively. We evaluate three open source software packages commonly used to carry out statistical inference of complex nonlinear models: OpenBUGS, AD Model Builder, and R. To test the strengths and weaknesses of each package, we will bring together experts in all three software packages and apply a common set of ecological models. Working directly with NCEAS informatics staff, we will produce a web‐based guide regarding the utility of each package for particular applications that includes annotated model code for each package, the data sets used in the applications, and peer‐reviewed articles. We will also identify how the different packages can be modified to improve their applicability to an array of complex nonlinear models that are essential for advancing ecological research. As statistical models are becoming increasingly more complex and ecologists are faced with a myriad of software options, the results of this project will provide support for ecologists and analysts across a broad spectrum of specialties.