My code ignored this problem, I didn't know you were talking about merges. In my code I simply recomputed the liberty count when there was a merge.
I'm not convinced all of this is worthwhile, especially when you keep adding more data structure. Also, it seems like modern processors favor keeping things simple and focusing more on cache behavior and branching stuff. All the heavy logic needed to update things incrementally may have worked better 15 years ago. So I guess you just have to try to see what works for you. Code it several different ways to see what works the best. You can use the code of one implementation to debug then next - always checking to see that you get the same answer. - Don 2009/8/15 Jason House <jason.james.ho...@gmail.com> > On Aug 15, 2009, at 8:22 AM, Don Dailey <dailey....@gmail.com> wrote: > > > > 2009/8/15 Jason House < <jason.james.ho...@gmail.com> > jason.james.ho...@gmail.com> > >> On Aug 14, 2009, at 11:02 PM, "David Fotland" < <fotl...@smart-games.com> >> fotl...@smart-games.com> wrote: >> >> Moves often merge two groups. >> >> I count liberties incrementally as I make moves, so no need to search to >> count. >> >> >> How do you detect shared libreties to avoid double counting. Simple >> addition does not work for real liberties (and I think you've said >> previously that you track real liberties. >> > > I don't know how David does it or if there are special tricks, but you get > real updates without that much extra work - you just have to look at a few > points around the newly placed stone. > > > > It's not that simple. Shared liberties can occur far away with longer > changes. I've come up with a scheme that looks a few points around, but only > works for chains up to about seven stones. Consider two parallel, linear > chains of 5 stones each, separated by a space. A stone placed at one end to > connect them is likely miss the more distant shared liberties. > > XXXXX > + > XXXXX > > It's also possible to construct more complicated cases (with non-linear > chains) where there are no locally shared liberties, but some exist further > away. > > _______________________________________________ > computer-go mailing list > computer-go@computer-go.org > http://www.computer-go.org/mailman/listinfo/computer-go/ >
_______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
