Introduction

The primary purpose of this project is to adapt open source software to run upon 64-bit, distributed environments.  The choice of which open source software to develop and the priority to assign to the development was dictated by civil engineering applications.  Initially, three applications had been identified; there were

1. Traffic Safety Analysis – examines the linkage between automotive accidents and the characteristics of accident locations.
2. Propagation of Uncertainty in Flood Models using Radar-Rainfall Analysis – seeks to predict how reliable flood predictions will be given knowledge on reliability of input data (rainfall).
3. Rainfall-Accident Relationship – analyzes the impact of rainfall upon the rate and types of automotive accidents.

The initial selection of open source software were WEKA, GRASS and R, which are, respectively, a JAVA-based machine learning package, a GIS system, and a statistical programming language.  The initial technical objectives are as follows:

1. Integrate and evaluate the usage of WEKA, GRASS, and R into the analysis efforts of Dr. Sun and Dr. Habib
2. Modify WEKA, GRASS, and R to support 64-Bit Intel Itanium processors and AMD Opterons
3. Deploy onto

4. Develop distributed versions of subcomponents identified as beneficial.
5. Generate Packages of Modified Software

Information relevant to the project:

Big Red:  (a) Instructions on use and b) Bigred's Webpage

GNU R: Information about the efforts to install GNU R.

mpiJava: Information about mpiJAVA (at least, what is relevant to this project).

WEKA:  Information about WEKA.

GRASS GIS: Information about GRASS.

CVS Repository Information