network simulations

[Claudia Pahl]

Hi folks,

here a note concerning the simulation of networks with SWARM. We are
interested in network models both for ecological and socio-economic
systems. Therefore we have implemented a preliminary generic network
simulation tool. First applications are a model for consumer behaviour
(network elements are consumers which differ in decision rules, links are
given by information exchange) and a model for energy flows in a size based
ecosystem model (elements are given by species which differ in their
predation window, links are given by resource flows). To test the
reliability of the models we start with simple network configurations and
make them progressively more complicated. In case of the ecosystem model we
can further test for the conservation of material exchanges. At the moment
we neglect space and focus simply on the complexity of the network
configurations. So far this short notice about our progress with SWARM. We
are interested to exchange information about model concepts -
implementations with other people working on similar topics.

More generally I consider that generic tools for network simulations would
be very helpful. What would be required?

- abstract network classes (network, elements, links..)
- network generation routines
- network analysis tools
- network graphics
- testing procedures to check for accuracy.....

It might be useful to start from the beginning by considering spatial
networks. Further one should conceive of the possibility of hierarchical
networks with different levels operating on different time scales. This is
of particular interest when we try to simulate the relationship among
different levels of organization as is for example the case in questions
regarding biological effects in ecotoxicology or regarding the dynamics of
institutional variables in mutual exchange with the behaviour of
individuals in socio-economic systems.

It definitely will not be easy to come up here with generic solutions. We
have already made this experience when we compare the requirements for
socio-economic and ecological systems. Nevertheless I am convinced that one
should try to coordinate efforts in this direction to facilitate exchange,
to avoid redundancies, to optimize performance and to guarantee
reliability. Perhaps it would be worthwhile to consider having a workshop
on this topic.

A final comment. Life would be easier (at least for us) if the SWARM
developers provided a more systematic documentation about their class
structure in general, specific classes, methods etc. I can fully understand
that programming has priority to writing documentations, in particular
since users are expected to be able to use SWARM without having to know all
the details of the underlying structure. Nevertheless we definitely would
appreciate a more systmatic listing and explanations.

All the best

Claudia


____________________________________________________________________________
Claudia Pahl-Wostl                                  e-mail: address@hidden
EAWAG, Switzerland                                  phone:    +41-1-823-5542
Environmental Physics                               fax:      +41-1-823-5210