Calculating number of parents?

For scripts to aid with computation or simulation in cellular automata.
Post Reply
FlamingObsidian
Posts: 1
Joined: September 8th, 2017, 12:14 pm

Calculating number of parents?

Post by FlamingObsidian » September 8th, 2017, 12:17 pm

How should I use scripts to calculate a number of parents? I am looking to use an algorithm similar to https://oeis.org/A196447 to find a solution to the unique father problem. There's no mention of how the number of parents was calculated, so I don't know what to do.

User avatar
dvgrn
Moderator
Posts: 10670
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Calculating number of parents?

Post by dvgrn » September 8th, 2017, 4:18 pm

FlamingObsidian wrote:How should I use scripts to calculate a number of parents? I am looking to use an algorithm similar to https://oeis.org/A196447 to find a solution to the unique father problem. There's no mention of how the number of parents was calculated, so I don't know what to do.
The number of parents can be found with WLS, JLS, or lifesrc. EDIT: It may be difficult to match the numbers in A196447, though -- have to find out what definition of "parent" Nicolay was using. Clearly he wasn't counting trivial parents, since there are always either none at all or an infinite number. I wonder if it's just parents in the pattern's radius-1 zone of influence, or something like that?

I haven't tried this lately, but it should be possible to set up WLS/JLS to do something like this, with "don't-care" cells outside the zone of influence. The rumor is that the count goes a lot faster if you start the search at the center of the pattern and expand outwards. But you'd have to set up each search manually, where you probably really need some kind of completely automatic countparents() function. That would require some serious adjustments to lifesrc source code, I think...!

It also looks like Nicolay Beluchenko's program can still be downloaded -- if not from his site, then from the Wayback Machine. I believe this is what he originally used to generate the A196447 sequence. Unfortunately the source code doesn't seem to be available, so that's probably a dead end unless you can get in touch with him directly.

It doesn't seem as if it would be easy to modify the "lifeeden" program to limit the search to stable configurations, in any case -- that's a fairly significant change, and I'm not sure how it would be done exactly.

Post Reply