Installation Notes for Matlab Functions on Unix/Linus Systems Jim Ramsay, McGill University To install the Matlab functions and sample analyses in a Unix system, using Matlab Version 6, follow these instructions, which will install -- the Matlab functions -- the sample analyses and the data that are analyzed A. Installing the Matlab functions: 1. Create a directory to contain the Matlab functions. For example, on my system the directory is ~/fdaM 2. Put the file "functions.tar.Z" into this directory, and uncompress it with the command uncompress functions.tar.Z 3. Extract the function files, each with extension ".m", from this .tar file using the command tar -xvf functions.tar Each of these function files will contain a function with the same name as the file, and possibly some supporting functions only used by this function. Documentation on the use of the functions is found in the leading lines of the file. 4. Within this directory, create a subdirectory with the name "@fd". That is, on my system, this would have the path ~/fdaM/@fd This subdirectory will contain functions that are process objects of the "fd" class. Move the file "@fd.tar.Z" into this directory, and extract the function files as you did in steps 2 and 3. 5. Repeat step 4 twice, with subdirectory names "@basis", "@bifd", and "@Lfd", and corresponding files "@basis.tar.Z", "@bifd.tar.Z" and "@Lfd.tar.Z", respectively. B. Installing the examples: There are currently six sample analyses bundled with the data that are analyzed: -- the gait data, in file gait.tar.Z -- the growth data, in file growth.tar.Z -- the handwriting data, in file handwrit.tar.Z -- the lip movement data, in file lip.tar.Z -- the pinch force data, in file pinch.tar.Z -- the monthly and daily weather data, in file weather.tar.Z You might consider setting up a separate subdirectory for each of these analyses, perhaps within a directory "examples" in the directory containing the functions set up above. For each of these analyses and data, move the .tar.Z file with appropriate name into the appropriate subdirectory. Then extract the files in this file using uncompress and tar commands as above. To run a sample analysis, start Matlab. At the top of each sample analysis file, with the extension ".m", you will find two "addpath" commands that attach, respectively, the "functions" directory, and the directory containing the sample data. The paths in these commands are what I use in my system, and you may have to change them to what is appropriate for your system. For example, at the top of the "monthly.m" file, you will find the two commands addpath('~/fdaM') addpath('~/fdaM/examples/weather') that add the needed paths on my system. Last modified on 10 March 2003