Sports League Scheduling with GALib

[James Cavanaugh]

I am using the work of Jason Leonard to write an optimization program for Little League Scheduling.  I am using the  GA1DArrayAlleleGenome as he does.  My program works, however the algorithm I use (Jason's) does not handle one "hard"constraint.  I use the SwapMutator and PartialMatchCrossover as suggested.

This constraint is that a team cannot play two days in a row.  I have tried using a GA2DArrayAlleleGenome but each "row" of the array exchanges genetic information.  Is there a way to prevent this?  I think if I use each row in the array to be a "round" I can get better optimization of this constraint.  The current algorithm builds a genome, then manipulates it (predictably) to create a "fixture" or schedule.  For my hard constraint, this secondary algorithm needs to be part of the GA I believe.  Right now it really is not.

I am a newbie to C++ and GALib.  (I know Eiffel better than C++) however C++ does not seem to be my problem.  My fear is I must create my own genome class, or mutator and crossover operators.  I don't know if I understand the examples enough to attempt this.

Can anyone help?  I would appreciate any help on this matter.

Thanks,

Jim

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/pipermail/galib/attachments/20020211/e5e78536/attachment.htm