bob@statsci.com (Robert Treder) (06/08/90)
ANNOUNCING VERSION 1.0 OF S-PLUS FOR DOS
========================================
Statistical Sciences, Inc. (STATSCI), based in Seattle, Washington, has just
released Version 1.0 of S-PLUS for DOS. S-PLUS is a powerful system and an
easily extensible interactive language used for graphical data analysis,
statistics and mathematical computing.
GENERAL DESCRIPTION OF S-PLUS for DOS
=====================================
S-PLUS is a fully supported superset of New S. New S was developed by
Becker, Chambers and Wilks at AT&T Bell Laboratories in an ongoing research
effort that now spans a decade. Complete access to DOS tools is available
from within S-PLUS. S-PLUS provides over 500 functions which may be easily
extended by defining new functions written in the S-PLUS language and/or
by incorporating user-supplied Fortran or C programs.
STATSCI's ENHANCEMENTS AND EXTENSIONS
=====================================
CLASSICAL INFERENCE
--------------------
S-PLUS now contains for the first time a basic set of classical methods
for estimation and testing. These methods are available in an interactive
and graphical display form which makes S-PLUS much more widely useful
for both teaching and research. The methods contained in Version 1.0
include:
t-test (paired and unpaired One-way ANOVA with and without covariate
equal and unequal) Kruskal-Wallis test
Wilcoxon signed rank test Generalized Cochran Q test
Binomial test for p Mantel-Haenszel test for association
Chisquare test for r X s Stuart-Maxwell test for association
Fisher's exact test for 2 X 2 t-test for zero correlation
McNemar's Test for symmetry Kendall's Tau and test for zero correlation
Bivariate normality test Spearman's rank correlation and test
Univariate normality test for zero correlation
Additionally, the binomial, hypergeometric and Wilcoxon distributions have
been added to the list of distributions represented in S-PLUS.
MODERN REGRESSION TECHNIQUES
----------------------------
Regression model linearization (ACE)
Additivity and Variance Stabilization for regression (AVAS)
Projection Pursuit Regression
Least Median of Squared Residuals (LMS) Regression
Kernel Smoothers, Friedman and Stuetzle's "Super Smoother"
TIME SERIES
-----------
Univariate ARIMA models
Multivariate AR models
Robust Fits for univariate AR models
Complex Demodulation
Spectral Analysis
General specification of filters and smoothers
Robust filters and smoothers
ROBUST METHODS
--------------
Bisquare estimates of location and scale
LMS regression
Generalized M-estimates of AR parameters in time series
Robust estimates of location and scale in time series
Robust filters and smoothers for time series
GENERALIZED LINEAR MODELS (GLIM)
--------------------------------
This is a general implementation of McCullagh and Nelder's
Generalized Linear Models including gaussian, binomial, poisson,
gamma, and inverse gaussian error structures and identity, log,
logit, probit, square root, inverse and loglog link functions.
Model fitting includes offset and model sequence specification.
SURVIVAL ANALYSIS
-----------------
Survival Curves including Kaplan-Meier and Fleming-Harrington
Cox Proportional Hazards Models
G-rho family of tests for differences between survival curves
NEW "HELP" FILES
----------------
Revised "help" files format with additional information and references,
a permuted index for all the functions, optional pagers for scrolling,
and optional pop-up help windows.
=========================================================================
SUPPORT
=======
Software and documentation updates
S-PLUS hot-line phone support
Statistical consulting
On-site training courses
SYSTEM REQUIREMENTS
===================
386 IBM-PC, PS/2 or compatible with 387 math chip
2 MB RAM (4 MB recommended)
Hard Drive (recommend 10-15 MB available)
PC-DOS 3.1 or above
TECHNICAL NEWSLETTER
====================
An S/S-PLUS technical newsletter series will commence in May.
Anyone wishing to receive this may do so by sending their surface
mail address to STATSCI's electronic or surface mail address
provided below.
DISCOUNTS
=========
Special introductory discounts are now available until July 31, 1990.
Additional discounts are available for educational organizations.
FOR MORE INFORMATION
====================
Email: mktg@statsci.com
or uunet!statsci!mktg.
Surface: STATSCI
P.O. Box 85625 Phone: (206)322-8707
Seattle, WA 98145-1625 USA Fax: (206)322-0738