Re: Re: Re: colony splitting

Hi readbeard,
Please explain, how do you calculate the number of colonies in a given environment.
You mentioned in the first 'situtation':

Time t0 => 2 colonies
Time t1 => 3 colonies

This is only possible if you don't count single cell as one colony. Please verify that it is the truth.

From my perspective, in the first situation at t0 and t1, both have 3 colonies.

Time t0, we have 7 => 'a' , 1 => 'b' and 1 => 'b'
Time t1, we have 3 => 'a', 3 => 'b' and 3 => 'a'

If my perspective is correct, then your problems is that how many 2-D groups you can find for given 'environment' at different timings. There may be well known algorithms for this. What I recall is the TileFall perl Program from PerlPress. See how the initial groups are calculated and you may find your algorithm from there.

I also suggest to research Graphical algorithm index and to get some valuable hints.

artist

Comment on Re: Re: Re: colony splitting

Replies are listed 'Best First'.
Re: Re: Re: Re: colony splitting by redbeard (Beadle) on Mar 13, 2003 at 06:10 UTC
Yup - my bad - There are 3 colonies at t0 and 4 at t1, as both b's are seperate colonies. I may later thing about implementing colony merges, but that's a rather tricky business, as Grosberg et al. have shown that colony merging is largely determined by genetics (so, I may keep track of "original colony" later on) but that's a level of complexity that won't be necessary for what I want to do with this model. I'll take a look at the links, though - thanks!	[reply]

Replies are listed 'Best First'.

Re: Re: Re: Re: colony splitting
by redbeard (Beadle) on Mar 13, 2003 at 06:10 UTC

Yup - my bad - There are 3 colonies at t0 and 4 at t1, as both b's are seperate colonies. I may later thing about implementing colony merges, but that's a rather tricky business, as Grosberg et al. have shown that colony merging is largely determined by genetics (so, I may keep track of "original colony" later on) but that's a level of complexity that won't be necessary for what I want to do with this model. I'll take a look at the links, though - thanks!

[reply]