Sets, trees, contrasts and other enhancements for Releases 6 and 7
by
Roger Payne
VSN International, 5 The Waterhouse, Waterhouse Street, Hemel Hempstead, Herts HP1 1ES, UK.

This talk will review some of the new facilities that were provided in the server of the 6th Edition of GenStat for Windows (Release 6.1), and give advance highlights of Release 7.1 (due to be completed in April 2003).

The aim of Release 6.1 was not only to strengthen and consolidate the existing facilities, but also to expand the scope of the system.

At the simplest level the enhancements included extensions to the ANOVA and regression algorithms to remove the previous limitation of no more than 30 factors or variates in the models to be fitted. Now up to 1022 can be included, but this can easily be extended if required - let me know! In regression, the POL and REG functions were extended to work on factors, and they can now be included in interactions. The meaning of the REG function was clarified, so that now its contrasts are always orthogonalized for the main effects of the variate or factor (even if the matrix third argument is set). Unorthogonalized contrasts are now fitted using the COMP function (previously available only in ANOVA), which has an identical syntax to REG. The accumulated analysis-of-variance (or deviance) table was extended to give full details of submodel terms. The RKEEP directive has three new parameters: SUMMARY and ACCUMULATED save the summary and accumulated analysis-of-variance (or deviance) tables respectively, and the STATISTICS parameter allows statistics to be saved for any current y-variate (rather than only the first, as with the existing STATISTICS option). The PREDICT directive can now calculate SEDs and LSDs. Also, procedures were written to provide a user-friendly interface (also accessible by menu) to John Nelder and Youngjo Lee’s HG system for fitting hierarchical generalized linear models.

There were extensions to the saving of information from ANOVA. The AKEEP directive has a new option to control the type of residual that is saved (either simple or standardized), and three new parameters, to save symmetric matrices of sed's of means and combined means, and se's of combined means.

New tree data structures were added, together with a suite of directives and functions to enable them to be defined and manipulated. These provide the basic utilities for a set of higher-level procedures for constructing particular types of tree. Most users will call only the procedures, but the other utilities provide a tool kit for user-defined extensions and enhancements.

Another new area in Release 6.1 was provided by the directives SETCALCULATE and SETRELATE, which perform Boolean operations on sets of items, which may be stored in pointers, variates, factors or texts. A related directive, SET2FORMULA, provides convenient ways of constructing a model formula for a set of factors stored in a pointer.

Most of the new statistical facilities provided so far for Release 7.1 have been implemented as procedures. These include procedures for analysis and display of circular data, for statistical process control and for survival analysis.

Date received: October 1, 2002

Copyright © 2002 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cajn-47.