MAD Installation of Mad
This page is part of the MAD Manual. See MAD Manual.
In this section we detail the steps needed to install Mad on a UNIX or PC platform.
Mad requires Matlab version 6.5 or higher.
Mad is available from Tomlab Optimization.
Installation Instructions (UNIX)
Mad is supplied in the gzipped TOMLAB tar file *.tar.gz. Please follow these instructions.
- Place *.tar.gz in an appropriate directory dir. TOMLAB and Mad will be placed in its own sub-directory so you don't need to create a specific TOMLAB or MAD directory.
- Now, from within directory dir, gunzip, and tar extract. Mad will be installed in tomlab/mad gunzip *.tar.gz tar xvf *.tar. If you have any difficulties using gunzip or tar please check with your system's administrator that they are installed and accessible to you before contacting the author.
- To be able to use Mad it must be in your Matlab path and must have several other paths and Matlab global variables initialized. To ensure this occurs correctly on starting TOMLAB you must edit your personal TOMLAB startup.m in /tomlab. MAD succesfully installed should appear in your Matlab command window when starting TOMLAB.
Installation Instructions (Windows)
Follow the directions in the TOMLAB installer. Edit the TOMLAB startup.m file to make sure that MAD will be initialized when starting TOMLAB. MAD succesfully installed should appear in your Matlab command window when starting TOMLAB.
Online Help Facilities
After successful installation you can get help on many aspects of using Mad from the Matlab command line. For example,
- help MAD gives a list of web-pages, directories and classes associated with Mad.
- help fmad/Contents lists all the functions of the fmad class used for forward mode automatic differentiation.
- help madutil lists all Mad's utility functions designed to ease use of automatic differentiation.
- help MADEXAMPLES lists the folders contained example scripts and source files, many of which we meet in this user guide.
- help MADbasic lists the example scripts and function files for all the basic examples. All the example scripts, which start with MADEX (e.g. MADEXForward1), have been written in cell mode allowing us to generate the LATEX source code for most examples featured in this user guide.