Growth in lattice v. cont. space

[JAN KREFT]

Hi Swarmers,

having been a lurker and swarm-support reader for some time I could 
finally start a new life as a Swarm user some days ago and feel excited 
about it.

The first problem I have come across is implementing space. I want to
model social interactions of bacteria in biofilms. As a start and also a
basic unit of the biofilm model, I want to model the growth of a single
(immotile) bacterial cell into a cell cluster. The mother cell divides
into two daughter cells of equal size that then grow until their size has
doubled, thus starting the division cyle over again. 

I first considered a Grid2d world like in Heatbugs, but implementing
growth in a Grid2d means avoiding overlap of cells after division. This
would need a major reshuffling of cells in an outward direction after
every division cycle (the cells should not do that themselves). With a
continuous space, overlap would not arise upon division but during growth
(seems to be more natural) and must be avoided also. Maybe a
representation of pressure and a method to minimize pressure could do the
job. Does anyone have experience with a similar problem? Also, I would
appreciate comments on advantages/disadvantages of implementing growth in
a lattice versus continuous world, both from the modelling and the
programming perspective. 

Thanks,

Jan.

Jan Kreft                       Phone  +44 (0)1222 874000 ext 6806
Biology                         Fax    +44 (0)1222 874305
University of Wales Cardiff     E-mail address@hidden
PO Box 915, Cardiff CF1 3TL, UK