Solvers -- almost-locked sets

Programs which generate, solve, and analyze Sudoku puzzles

Solvers -- almost-locked sets

Postby Bob Hanson » Tue Dec 23, 2008 1:19 am

I'm wondering what other solvers, besides the Sudoku Assistant, <plug>http://www.stolaf.edu/people/hansonr/sudoku</plug type="shameless"> handle almost-locked sets.

If these exist, do they handle them in the X-direction (similar to X-Wings and Swordfish, but not locked)?

My recent checking of many puzzles suggests that these X-direction almost-locked sets are both very common and very easy to spot by hand.
Bob Hanson
 
Posts: 75
Joined: 04 December 2005

Re: Solvers -- almost-locked sets

Postby champagne » Sat Dec 27, 2008 8:41 pm

Bob Hanson wrote:I'm wondering what other solvers, besides the Sudoku Assistant, <plug>http://www.stolaf.edu/people/hansonr/sudoku</plug type="shameless"> handle almost-locked sets.

If these exist, do they handle them in the X-direction (similar to X-Wings and Swordfish, but not locked)?

My recent checking of many puzzles suggests that these X-direction almost-locked sets are both very common and very easy to spot by hand.


Hi bob,

When I started developpement of my program years ago, I was invited to look at your site and I found similar views in your 3D medusa. I took over at that time some of your findings.

Meantime, I developped the program in several directions. ALS are not directly handled by the program. Instead, I work on the complementary pattern, AHS (Almost cells for me).

This is described in details in my website. It is for me an old story.

I studied a while interactions between ALS, but the code was fully redundant with AIC's crossing AHS/ACs, so I discarded that code completely.

I can try to answer more precisely if I catch exactly what you are waiting for

champagne
champagne
2017 Supporter
 
Posts: 5675
Joined: 02 August 2007
Location: France Brittany

Postby tarek » Sun Dec 28, 2008 9:13 pm

As far as I'm aware ....

No one has implemented the x-ALS in the solvers ......

Are there any examples that can't be represented as fish (which are more elegant & probably better for solvers) ....

How many possible x-cycle eliminations does this method miss (which POM doesn't) ?

tarek
User avatar
tarek
 
Posts: 2622
Joined: 05 January 2006

Postby Bob Hanson » Tue Dec 30, 2008 6:57 am

It looks to me like 5x6, 6x7, and (gulp!) 7x8 X-almost-locked sets are not accounted for in these standard checks.

Take a look at http://www.stolaf.edu/people/hansonr/sudoku/ex_Aig.htm

A 7x8 example is

http://www.stolaf.edu/people/hansonr/sudoku/ex_A1g.htm?i520a.517.18
Bob Hanson
 
Posts: 75
Joined: 04 December 2005


Return to Software