As an LBJ student, you have access to a variety of statistical or mathematical software (listed below).  All that is required is that you add Austin Services (abbreviated AMS) added to your IF account (used for accessing some computer labs on campus) or departmental account.  Information on how to add this service is available at http://www.utexas.edu/its/wts.  ITS will charge you a small fee for disk services (generally about $13 per year).  For helpful details on how to work with a remote desktop, see the directions below.

How to Use the Windows Terminal Server

First, add Austin Services (abbreviated AMS) added to your IF or departmental account.  Information on how to add this service is available at http://www.utexas.edu/its/wts.  ITS will charge you a small fee for disk services (generally about $13 per year). 

When signing up, you will need to add “disk services” and then “Austin Terminal Services” to your account in order to use the statistical and mathematical applications on our time sharing system.

In the new system you will have a remote desktop, but it will look just like a local Windows desktop. You will need to configure your remote desktop as follows:

Configuring Your Remote Desktop Client for Windows
http://www.utexas.edu/its/wts/answers/rdc-windows.php

Configuring Your Remote Desktop Client for Macintosh OS X
http://www.utexas.edu/its/wts/answers/rdc-macintosh.php

To access statistical applications click on Start, then All Programs, then Statistics. For mathematical applications click Start, then All Programs, and then Mathematics.

You can switch from your remote desktop to your local desktop by using the minimize button on the remote desktop. You will see a gray band at the top of your remote desktop with wts.austin.utexas.edu in the center of it. To the right you will see ” - ” for minimize and “X” for exiting from the remote window. The minimize symbol will take you back to your regular desktop.

If you need technical assistance with signing up for AMS, it is available through the helpdesk at 475-9400. You can also send us e-mail questions at stats@ssc.utexas.edu.

AVAILABLE STATISTICAL SOFTWARE

SPSS
SPSS is a graphical user interface statistical package capable of performing a broad range of statistical analyses such as General Linear Models, Regression, Reliability, and Confirmatory Factor analysis.

SAS
The SAS package is a powerful tool to organize, manage, and analyze large amounts of complex data in a variety of formats.  The SAS system can also be used easily create and run macros, allow for the efficient processing and reporting of data.

AMOS
AMOS is a graphical user interface package for running Path Analyses and Structural Equation Models.  Contains built-in bootstrapping for handling missing data, and is callable from within SPSS.

Mplus
Mplus can be used to examine latent variables for Structural Equation Modeling and Confirmatory Factor Analysis, and can handle continuous and categorical outcomes.

HLM
Software that is designed to examine multilevel models – databases with multiple levels.  Such multiple level databases are often known as Hierarchical Linear Models, Random Coefficient Models, or Mixed Effect Models.

Lisrel
Lisrel can be used to analyze path models or structural equation models.  In addition, Lisrel can also perform Confirmatory Factor Analysis. 

Sudaan
Survey data that contains complex data structures can be analyzed with Sudaan.  It produces reliable estimates of population parameters based upon the sample design.  Sudaan can be used as a stand-alone package or as a callable procedure within SAS, allowing the user to take advantage of complex sample estimates.

Stata
Stata, while a package capable of performing standard statistical tests, is well known for its handling of clustered and stratified datasets.  The newer versions of Stata are also capable of handling Generalized Estimating Equations for longitudinal analysis.

AVAILABLE MATHEMATICAL SOFTWARE

CPLEX
This is a package developed by ILOG, Inc. whose purpose is primarily to find a set of optimal values for various model parameters such that certain expressions derived from them are maximized or minimized within possible constraints upon other expressions derived from them.  Constraints can be based on equalities, inequalities, or restrictions to integer values.  The programs are designed to handle very large scale problems.

ILOG Solver
This is used in conjunction with CPLEX and is essentially a set of library functions designed to solve particular classes of constraint problems.  The underlying code for these is C++.

IMSL Libraries
This package, from Visual Numerics, is also a set of statistical and mathematical library functions that are callable from programs written in C.  They cover areas of linear algebra, eigensystem analysis, interpolation, approximation, quadrature, differential equations, integral and discrete transforms, nonlinear equations, constrained and nonconstrained optimization, specialized functions, and random number generation among others.  There are also functions useful in statistical analysis such as regression and time series.

LabVIEW
This is a package from National Instruments that is used for creating VIs (virtual instruments) and algorithmic computational procedures from modular components and connections.  It is useful in modeling complex systems with circuits and feedback loops. It is also useful in interfacing analysis code or data acquisition instructions with hardware or instrumentation, although this would not normally be done from a terminal server.

Maple
This is a general purpose math package from Waterloo Maple with its own programming language.  Its primary use is for symbolic algebra and mathematics rather than numerical computation although there are procedures for converting symbolic expressions to numerical values.  It is useful for calculus, integral transforms, discrete transforms, differential geometry, linear programming, combinatorics, optimization problems, factorization, expression simplification, and many other areas. This is the package to use when there is interest in elegance and exact solutions in closed form.

Mathematica
A general purpose mathematical software package from Wolfram Research.  It is designed for both symbolic and numerical computation.  It also has its own scripting language and can be threaded with text in notebook format.  Both text and computational components can be represented by characters from bitmapped fonts, thus accommodating Greek letters and various standard mathematical notation symbols commonly used in mathematical notation and display.  There is a base kernel with which elementary mathematical operations can be performed to arbitrarily long precision such as a thousand decimal places, and add-on packages that can be imported into memory for specialized computational purposes and mathematical topics such as calculus, linear algebra, discrete mathematics, geometry, graphics, number theory, and statistics.

Matlab
Another general purpose mathematical software package with a graphical user interface.  It seems to be the most commonly used by students and researchers, probably because of the simplicity of its programming language, easy learning curve, and large user community sharing privately developed third party programs and toolboxes. There is a capacity for modular structure with functions having separate code files from their calling scripts, all capable of interacting and referencing each other. The package consists of a base kernel of general functionality and add-on packages called toolboxes that are collections of specialized functions useful for a particular discipline or narrow computational purpose.  The server is also licensed for these toolboxes:

Signal Processing Toolbox
This toolbox has functions for analyzing time series of periodic phenomena in the frequency domain, creating various frequency filters, working in the complex plane, finding poles and residues, and a user interface for creating wave forms and so forth.

Image Processing Toolbox
This toolbox has functions for manipulating images that are imported as matrices of pixel RGB values or of pixel intensities with color maps.  It is similar to Photoshop, but with the convenience of doing the manipulations within a single script in the Matlab environment.

System Identification Toolbox
This toolbox consists of functions and graphical interface tools useful in determining the suitability of various linear models for explaining observational time series data.  Values are generated from system states measured at previous times and models are validated or assessed by application to regions in the time series that were not used in model construction.

Control System Toolbox
This is a toolbox related to the System Identification Toolbox and includes many of the same functions.  These are utilities for control systems engineering such as response functions for various input forms.  Also there are graphical interface tools such as the LTI viewer for analyzing linear time invariant systems.

Optimization Toolbox
This toolbox consists of functions useful in finding optimal values for parameters or variables in a system trying to models observational data.  This includes functions based on various search algorithms.  These accommodate systems where variables have bounded ranges and linear and non-linear constraints in the relationships among underlying variables.

Statistics Toolbox
This is a toolbox newly being added to the WTS license for the Fall 2008 semester.  It consists of a set of functions and routines that go beyond the elementary descriptive statistical functions that are part of the base kernel.  These include procedures for various probability distributions, multivariate statistics, nonlinear modeling, ANOVA, hypothesis testing, statistical data visualization, and so forth.

Simulink
This is a computer aided design utility for constructing models to simulate dynamical systems. There are building blocks with inputs, outputs, or both that can be connected together in loops and circuits with GUI based input of parameter values.  Included in the functionality are modular blocks for sources, sinks, system operations, numerical integration and differentiation, gain, etc.  Variable values can be monitored in real time during simulations.

Stateflow
This is another simulation utility, much like Simulink but it is an event and logic driven simulation resource rather than dynamical systems simulation resource.