Scientific Computing
Getting started
Spring, 2003

Back to the main course home page.

This should help you get started using C or C++ and Matlab together for computing and visualization.

Environment: You need access to C/C++ and Matlab. All registered students may use the math department network of SUN workstations that run a version of the UNIX operating system and have installed both the SUN and GNU C/C++ compilers complete with performance tools and a window based debugger. If you choose to work elsewhere, a personal computer or computer at work are common choices, you must have a C/C++ compiler and soem visualization system such as Matlab (recommended) or Excel or gnuplot. You can purchase a compiler and or Matlab at most computer stores, including the NYU computer store.

Working with C/C++ and Matlab: To use C/C++ for computing and Matlab for visualization, you need to get results from the C/C++ into the Matlab environment. One simple way to do this is to make your C/C++ program create a file that Matlab can read. The files lis.C and lp.m give one way to do this. In my UNIX environment, I type the command "CC lis.C -o xlis" to compile the C++ program. Then I type "xlis" to execute the compiled executable. First it asks the user to type a number, then it creates a data file called "lis.m", which is really a few Matlab commands. Look at the file to see what this means. In Matlab, I just type "lp" and all the commands in that file are executed. The result should be a plot of a Lisajous figure.

Learning C: If you are not comfortable with C or C++, you need to get comfortable right away. I recommend The C Programming Language by Brian Kernighan and Dennis Ritchie. The first seven chapters suffice for most of what we do in this course. To continue with C++, try C++ Primer by Stanley Lippman. There are good Schaums Outline books on C and C++. Don't panic, the first few assignments require very easy programs and the course materials contain several program templates to work from.

Learning Matlab: The Matlab environment includes a primer for beginners. Once you understand the basics of Matlab commands, you should start experimenting with m-files such as "lis.m" and "lp.m". This avoids tedious typing and editing of commands.