Outline plan for the CS405 module 2011-12
- Weeks 1-3: EM ideas, principles, concepts and tools
- Weeks 4-5: EM and learning
- Weeks 6-7: EM and concurrent systems modelling
- Weeks 8-9: EM and software development
- Week 10: EM implications and prospects
The correspondence with weeks will be only approximate, but the above agenda gives a broad indication of the weight to be given to different topics.
Preliminary submission (theme and abstract only, unassessed) for WEB-EM-08 is due by Wednesday 23rd November (week 8): final submission by Tuesday 31st January (week 3, term 2). Examination content will be closely matched to the five topics listed above: one compulsory question on basic elements of EM, two questions out of four covering the application areas and matters arising. Past exam papers are available via the university website: please note that each past examination relates to a different variant of the module, and you will be given a lot of guidance regarding the topics to which the questions relate.
The additional 10% assignment for MSc students only will be specified and organised as soon as possible, but requires knowledge of tools and models to be introduced in weeks 3-5 and accordingly is provisionally planned for Week 6.
Useful links and resources
The CS405 module webpage, from which you can access previous years' module webpages and handouts etc as well as the module forum. You can also inspect previous submissions to WEB-EM by past CS405 students. (NB To date, most online examples of EM relate to EDEN models: some CADENCE models based on WEB-EM-6 and WEB-EM-7 submissions will be made available later.)
The basis for the ten-minute introduction to Introduction to EM at the 4th year module fair was this presentation.
All other material is linked via the EM webpage at http://www.dcs.warwick.ac.uk/modelling/
Publications and projects are accessible from the LHS of EM webpage: EM papers are indexed by codes (e.g. #107) and projects are identified by name: e.g.
racingGardner1999. Also have many PhD and MSc-by-research theses: especially helpful as an overview of EM is Karl King's MSc thesis Uncovering Empirical Modelling. Some of the more recent papers have associated links to Web EDEN models.
Can access projects in department at
/dcs/emp/empublic/projects (also accessible as
~empublic/projects). In most cases, these projects can be run by entering this directory and typing
tkeden Run.e. It may be necessary to use older variants of tkeden - the most relevant are tkeden-1.46 (not available on the Linux 64-bit workstations, but may be downloadable for Windows), tkeden-1.66 and tkeden-1.67. Some models have introductory activities associated with them - see the 'Tour' tab on the projects archive page.
Follow the Software link for tkeden-1.73 (currently installed as default at DCS, and available on Mac and Windows) and for the Web EDEN tool. There is currently little documentation for Cadence apart from what has been covered in lab sessions over the last 3 years - this includes information about adding your own C++ extensions and making distributed models (some of the older documentation may need adapting or updating as Cadence has been under development). For the latest download, you should consult https://github.com/knicos and the module forum. There is helpful documentation for EDEN in the form of Notes on definitive notations, under the
Help tab in tkeden and in The EDEN Handbook online. Other tools worth being aware of are the Dependency Modelling Tool and the Empirical Modelling Presentation Environment, both of which can be found in the Projects archive.
When developing / running EDEN models, you can save the history of your interaction. the entire history of interaction is also saved by default. The last few histories of your interactions are preserved in your home directory (on Linux/Mac) or in the directory in which the tkeden executable lives on Windows - these are called .tkeden-history, .tkeden-history-1, .tkeden-history-2, etc. These are useful for development, debugging and communicating about model-building with others, but it is generally better to use an independent text editor to record the model-building as a "rational reconstruction" of the exploratory interactions with the interpreter. Beware of text editors that use proprietory text formats.
There are very few publications about EM other than those on the EM website (note that Empirical Modelling is not the same as what is otherwise known as 'empirical modelling'!), but you can find some of the external papers relevant to EM referenced in CS405 in the lecture notes for previous years.