The Adaptive Sparse Grid Integration Project

  1. Introduction
  2. Documentation
  3. Downloads
  4. Copyrights and Usage
  5. Contact Information

Introduction

This project implements the sparse grid algorithm introduced by Gerstner and Griebel (T. Gerstner and M. Griebel. Dimension-Adaptive Tensor-Product Quadrature. Computing, 71(1):65-87, 2003). A suite of other integrators (especially standard sparse grid integration and Monte Carlo) is included for comparison purposes.

An effort has been made to divide the different parts of the algorithm into modules. For example, different error estimation and refinement strategies can easily be mixed and matched. It is also easy to use different quadrature formulas. Many quadrature formulas, including Patterson, Gauss-Hermite and Clenshaw-Curtis, are included with the project. It is possible to visually observe the integration process with visualizer plug-ins.

Requirements
Java Standard Edition 1.6 or higher

Documentation

Documentation for the code follows the philosophy outlined in the official javadoc guidelines. It is available in two forms, the programming guide and the API Specification. For a general overview, you should read the programming guide first.

The programming guide provides an introduction to the scope and aims of project, gives usage guidelines and code examples for this project.

The API specification outlines concisely the purpose, implementation and specifics of the classes. It is used for reference and for detailed description. The API specification is provided in javadoc format, and is automatically generated from the source code.
Documentation
Programming Guide
HTML
API Specification Javadoc Online
API Specification Javadoc in ZIP Archive
Benchmark for Genz Test Functions
Excel File

Downloads

All downloads includes the class files, source files and programming documentation for this project. The downloads also include the class files and source codes of any required other projects.

Downloads
2.0 RC1 JAR Archive

Copyrights and Usage

All source code contained in the de.torstennahm.* packages is placed in the public domain, no rights reserved. Some projects may include other required code. For this code, restrictions may apply. Please check the programming guide for any restrictions.

Contact Information

For contact information, please see the main page.