Peter Ross's home page

To get in touch with me:
e-mail to P.Ross@napier.ac.uk ,
phone 0131-455-2766 (use +44 131 455 2766 if phoning from outside the UK) ,
or write to:
    Prof. Peter Ross,
    School of Computing,
    Napier University,
    10 Colinton Road,
    Edinburgh EH10 5DT,
    Scotland

• Background information
• Research interests
• Stuff intended for people in the School of Computing
• A page of photos, mainly of Huangshan mountains, taken on a recent visit to China for a workshop on nature-inspired computing organised in conjunction with the University of Science and Technology of China in Hefei, in October 2004
• The EU-funded SIGNAL project, investigating how to make autonomous robots `grow up' with experience and time
• A report on EPSRC-funded evolutionary scheduling research
• Handy web links, which gives you some  more general idea of my interests.
• A page of useful timetabling-related stuff
• A very simple demo of a two-person zero-sum game written in Java, with source code

Here are one or two items of software you might like:

• A fast and simple robot simulator (version 1.8) in C++, using the Fast Light Tool Kit for the user interface. It runs under Linux or Windows. Think of it as a step forward from Olivier Michel's old Khepera simulator. This one has a separate world editor and allows circular and coloured objects; so, for example you can draw objects in the background colour and thus bite arcs and holes out of other objects. The robot has noisy IR sensors and also colour sensors, and can sense the colour it is standing on. Anything grey (that is, equal R, G and B) counts as carpet that the robot can cross. It can affect its world by dropping, detecting and picking up grey blobs, and the user can drag such blobs around too. On my 2Ghz PC the robot will do around 10,000 steps per second, where a step involves a movement, collision detection, any necessary skidding or changes due to collisions and then sensor updating.
• A program (version 1.5a) to let you explore some of the issues to do with cell matching that arise (or should arise) in artificial immune systems research. The program displays a two-dimensional "shape space"; a cell located at (X,Y) has a particular `shape' that depends continuously on X and Y. In keeping with Perelson and Oster's model, cells can be stimulated by approximate matching with other cells whose shape is close enough to being a perfect complement of their own shape. Interestingly, this so-called self-assertion model of cell interaction leads to the formation of semi-stable regions in shape-space. The software lets you explore several varieties of what approximate matching might mean, and play with the associated parameters.
• A GA package, pga version 4.0, which is intended for teaching and basic research in GAs. Source code in C, documentation included; can be used interactively, or non-interactively for massive batches of experiments. Loads of built-in problems and GA options: for example you can often vary the length of the chromosome, you can use an island GA, have the chromosomes laid out on a 2-d grid with local mating, use adaptive mutation rates, and so on. The built-in problem set includes some of the oldest test functions and also things like simplistic timetabling (not nearly as good as the package listed below), Royal Road functions, NK landscapes, Long Path problems and more.
• Old GA-based job-shop scheduling software: source code, in C. Separate gzipped Postscript documentation is available. No fancy user interface; it was built for research, in a Unix environment. Based on my pga, additions by Hsiao-lan Fang.
• A GA-based timetabling package, mainly useful for exam timetabling although it will handle some flavours of lecture timetabling too. Many universities have used this. Our state of the art has advanced somewhat beyond this now. Source code in C, documentation included. It has a limited, text-based user interface; it was built for research, in a Unix environment. Based originally on my pga, many additions by Emma Hart, Dave Corne and me.
• A bug-fixed version of random.c, a good random number generator originally created at Berkeley for BSD in 1988 and included (in buggy form) in various commercial operating systems. The bug is serious. For example, in Sun's Solaris 2.6 the man page for random suggests that all bits are usable and that random() & 01 generates an acceptably pseudo-random sequence of bits. Although this is true, the buggy version gives you access to only four such sequences!
• A program which finds cliques in graphs (just the largest, or all of them). It is deterministic, and can take a very long time if given a large graph. Source code in C, by me.
• A simple program of mine which searches quickly for true sentences of the form:

  This sentence contains three As, one B, three Cs, two Ds, thirty-three Es, five Fs, two Gs, seven Hs, eleven Is, one J, one K, two Ls, one M, twenty-one Ns, sixteen Os, one P, one Q, six Rs, twenty-five Ss, twenty-one Ts, two Us, six Vs, eight Ws, four Xs, five Ys and one Z.

• A fancier version of the sentence generator, with a nice GUI using the admirable, free, small, fast Fast Light Toolkit, a GUI toolkit for X and Windows that is well documented, has loads of examples, provides a neat and simple GUI builder, works well with OpenGL and is generally much more fun to use than heavyweight commercially-produced IDEs.