STATISTICS AND COMPUTING

STATISTICS AND COMPUTING

Return to homepage of Donald E. Myers

The use of computing is ubiquitous in statistics and it is essential for the practicing statistician to know and use various software packages. Computing is necessary not only for numerical computations but also for graphics and the manipulation of large data sets. In addition to the activity within various computing and statistical organizations, this emphasis has also resulted in the formation and scheduling of an annual conference on statistics and computing STATISTICS AND THE INTERFACE

Two statistical packages are in common use in statistics classes in the Department of Mathematics, these are MINITAB and S-Plus. While S-PLUS is a more sophisticated package and includes more extensive graphics, MINITAB is much easier to learn.

For a comparison of several packages see MINITAB,SPSS,GLIM
EXCEL is not a statistics package but incorporates some statistical functions, for a description and tutorial on these see MINITAB and EXCEL

STATLIB is a excellent site for a variety of information and links for statistics and computing.

The R PROJECT

R is `GNU S', a freely available language and environment for statistical computing and graphics which provides a wide variety of statistical and graphical techniques: linear and nonlinear modelling, statistical tests, time series analysis, classification, clustering, etc. Please consult the project homepage for further information. There are Windows, Unix and Linux versions available for downloading.

ACCESSING STATISTICS SOFTWARE

At the University of Arizona, MINITAB is available on the cluster of unix machines at CCIT. There is also a student edition of MINITAB that can be purchased at the bookstore (cost approximately $60), this requires access to a PC compatible. S-PLUS is available on a server at CCIT and also a server in the Department of Mathematics (the latter requires an account on the departmental system). Several other statistics packages are also available on the CCIT system , these include SAS and SPSS. A course on the latter is offered through CCIT, EDP 548 Statistical Packages. EDP 696a Multivariate Analysis All registered students, faculty and staff can obtain an account on the CCIT system using the Computer Accounts facility.

SPATIAL STATISTICS

The Spatial Newsletter site can be found at Spatial News

The University of Arizona is now a member of the University Consortion on Geographic Information Sciences, its webpage can be found at UCGIS For geostatistics in particular but also spatial statistics the AI-GEOSTATS website is useful, there is also a listserv. AI-GEOSTATS

SPATIAL STATISTICS SOFTWARE

There are at least two kinds of spatial statistics software available at the University of Arizona. GIS packages such as ARC/INFO and ARCVIEW are accessible via a site license in the College of Agriculture and the Department of Geography GIS USERS' GROUP. For a quick introduction to ARC/INFO see Tutorial.For another one see ARC/Info Tutorial. For tutorials and other information ARCVIEW

VARIOWIN 2.21 can now be downloaded free from this site (the book is out of print). Be sure to note the acknowledgement that should be included in any publication wherein Variowin has been used for some portion of the analyses.

GRASS is a public domain package and is installed in several departments on campus including the Department of Mathematics. ARCVIEW is available as a MS-Windows program from ARCVIEW. GIS packages generally do not include statistical computation components but S-PLUS includes a "window" to ARC/INFO.

GRASS and R is a good source for connections between GIS and geostatistics.

GEO-EAS is a public domain geostatistics package available in PC compatible (DOS) version, this can be run in a DOS window under Windows. Both UNIX/Linux versions are now available. There is spatial statistics component available for S-PLUS.

OTHER FREE AND PUBLIC DOMAIN SOFTWARE

GENERAL STATISTICAL PACKAGES

BAYESIAN STATISTICAL SOFTWARE

TIME SERIES

To return to home page of Donald E. Myers