ADMB, which stands for Automatic Differentiation Model Builder, is a reliable application that helps you create and compute non-linear statistical models. Automatic Differentiation refers to the set of techniques used to evaluate the derivative of a function that is specified by a computer program.
It exploits the fact that a computer program will always execute a sequence of elementary arithmetic operations and functions, regardless of its complexity. The application is vastly used worldwide by academic institutions for ecological and statistical modeling. The application requires any version of Visual C++ in order to properly function.
Fast model creator
The application runs in Command Prompt and allows you to create objects for DLL and ADMB-BE, based on your input data.
Furthermore, it helps you understand your model’s layout by offering you the chance to create object files that contain safe bounds and debugging symbols.
You can now visualize each step taken to create your statistic model, as well as repair any flaw or miscalculation that you might have added to your project.
A powerful object builder
ADMB utilizes the automatic differentiation capabilities of the AUTODIF C++ library, which was built specifically for this purpose. The application gives you the opportunity to study advanced algorithms, such as the Bayesian hierarchical model.
In addition, the program provides support for modeling random effects in a frequency framework by using high-class algorithms such as Laplace approximation and importance sampling.
Reviewed by Andrei Fercalo, last updated on March 13th, 2014
In a hurry? Add it to your Download Basket!
- Updated Visual Studio nmake build files.
- Improved Unix build files. Only outdated files are rebuilt.
- Note: Building from source will create distribution folder in build/dist instead of build/os-comiler-arch.
- Combined mulitple libraries to a single library 'libadmb.a'.
Application descriptionThe ADMB (Automatic Differentiation Model Builder) software suite is an environment for non-linear statistical modelin...