|Java Statistical Classes|
This page contains general information about this site, JSC and its author.
This site and the core JSC software were designed and programmed by Andrew Bertie.
The author has a MSc in statistics and computation, and a PhD in statistics. He has worked as a statistician in the fishing industry, civil service and NHS, and from 1982 to 2003 as a statistical programmer/software designer for the Open University; developing software and related materials mainly for statistics and mathematics courses, but also for other OU courses in science, education and social sciences. He is currently working as an independent consultant in the Department of Economics of the Open University - developing agent-based modelling software and analysing survey data.
Aims and Philosophy
The aim of this site is to promote the use of Java for statistical programming and teaching. The author believes that Java offers exciting opportunities to the statistical programmer and teacher. Java's object-orientated approach can be applied to statistical concepts, methods and data in powerful and expressive ways that would be difficult in older languages: it stimulates thinking about programming and its role in statistics.
The aim of the JSC library is to make Java more accessible to the statistical community, and to provide a reliable correct, modifible and reusable set of statistical classes to the general programmer.
In developing JSC, the author has used the latest published algorithms from reputable sources wherever possible. These have been incorporated into a coherent system of classes and interfaces which aspire to the principles of object-orientated design and effective Java usage. Accuracy, speed and robustness, however, have not been sacrificed in favour of elegant object-orientated solutions that would appeal to computer scientists.
While the author believes that the future of statistics is Bayesian, the JSC library will attempt to cover all schools of inference and approaches to data analysis. Suggestions for additions to the library are welcomed.
JSC uses the JAMA linear algebra package - a cooperative product of The MathWorks and the U.S. National Institute of Standards and Technology (NIST) which has been released to the public domain.
The use of other public domain Java classes, algorithms originally coded in other languages, and contributions by other programmers are acknowledged in the API documentation of individual classes and methods.
JSC and all the Java software available on this site were developed using Sun's Java 2 Platform Standard Edition v1.4.1. Versions earlier than 1.4 or from other providers are not supported. The Java 2 Software Development Kit (J2SDK) and Java Runtime Environment (JRE) can be downloaded free from the Sun Microsystems Java website.
This site was last updated 15th August, 2005.