PySAL: Open Source Python Library for Spatial Analytical Functions

PySAL is a collaborative effort between Luc Anselin and Sergio Rey to develop a cross-platform library of spatial analysis functions (see figure) written in Python. It is a work-in-progress that combines the development efforts on PySpace and STARS - Space Time Analysis of Regional Systems. Both will continue to exist and exploit a common library of functions.

PySpace is an open source software development effort that is part of PySAL to implement spatial statistical methods in general and spatial regression analysis in particular using Python and Numerical Python. Current activities focus on a set of classes and methods to carry out diagnostics for spatial correlation in linear regression models and to estimate spatial lag and spatial error specifications.

The geospatial semantic web project builds on the GeoDa and PySAL software projects. A major distinction between this software and the web services is that these services are intended for automatic discovery by other services and not primarily targeted at human users. This requires the development of appropriate geospatial processing ontologies to facilitate the discovery and properly describe the data requirements, assumptions and computational functionality of the analytical services. This ontology is implemented in Web Ontology Language (OWL) and tested using web services programmed in Python and OWL-S (Semantic Markup for Web Services), SAWSDL (Semantic Annotations for Web Services Description Language), and the Web Service Modeling Ontology (WSMO) are compared as the framework to support discovery.

As soon as PySAL code is available, it will be announced on the Openspace mailing list and on this website. Spatial regression code in Python continues to be under development and will be announced in the same forums.