Using Math Software Under UNIX
with material from Dave Hart
- Introduction
- UNIX - The absolute least you need to know
- Using the Math Software
- Using graphics over the network
- Using batch processing
- Further Reading
Introduction
Purpose
Distributed computing puts into practice a favorite Hoosier proverb: "Use the right tool for the job." As distributed computing continues to grow, the network is becoming one vast virtual machine, with different parts performing different functions. UITS provides a variety of hardware platforms, from large shared computers to desktop workstations in Student Technology classrooms and labs. Workstations typically provide superior graphics and a more convenient user interface, and offer powerful logical and arithmetic processing capabilities, but generally lack the large memory that shared computers provide. On a shared computer, you can use Maple to manipulate formulas thousands of lines long, or use Matlab to perform billions of computations -- all within minutes. Using workstations in conjunction with multiuser systems, via the network, can combine the best functionality of both.Who this document will help
This document is intended for users of Maple, Matlab, and Mathematica who need the computing power of very large computers. This is not a tutorial for any of these software packages; instead see Getting Started with Maple, Matlab, or Mathematica. We assume that you are fairly familiar with one of these packages. We also assume you are comfortable using the basic functionality of network utilities such as FTP and Telnet (or SSH). Our goal is to ease the transition from a window-style desktop environment to a command-line based UNIX environment.The standard math tools at IUB -- Maple, Mathematica, and Matlab -- have versions that can run on many different platforms. The microcomputer version offers a relatively friendly interface, while the mainframe version offers raw power. For everyday problems, you'll find the microcomputer version fast and easy to use. Even when you need the power of the mainframe, you may still want to develop algorithms and debug code on your desktop, then turn to shared systems for production runs that require large storage or extended number-crunching.
This document is designed for users at Indiana University (Bloomington or IUPUI campuses), who have accounts on either the Research SP system, Steel, Nations, DaVinci, or Ships. The information contained herein may not be appropriate for your specific circumstances (especially when refering to characteristics of specific computers).
How to use this document
Everyone should read the UNIX - The Least You Need to Know and Starting the Math Software sections, unless you already know this information. You may then pick among the remaining sections to fit your needs.Please send errors, omissions, comments, etc. to statmath@indiana.edu, or post to iu.statmath.
