Techniques spanning more than one grid in overlapping puzzle

For fans of Killer Sudoku, Samurai Sudoku and other variants

Techniques spanning more than one grid in overlapping puzzle

Postby tarek » Tue Jul 29, 2008 6:17 am

This issue has been brought forward recently & is quite interesting.

As I am finalising my overlappping puuzle solver, a programmers perspective has also surfaced.

An overlapping puzzle such as the "Samurai" variant can be viewed as 5 overlapping puzzles or as 1 big puzzle.

If you see it as 5 overlapping puzzles then you would then see that twhat happens in the overlapping section in 1 grid has to be carried to that section in the other grid (or grids)

If you look at it as 1 big puzzle, then the overlappingsections are only special cells that can SEE more cells.

This will have an impact on our solving tecniques.

My interest would be mainly in FISH & to see if some fruitful cover sectors can be used from more than one grid.

Examples here would shed a light on this matter

No need to limit things to FISH only.

I will post an example as soon as I come across one.

tarek
User avatar
tarek
 
Posts: 3762
Joined: 05 January 2006

Postby Glyn » Tue Jul 29, 2008 9:36 am

tarek Perhaps the most suitable Gattai to start fishing in for an early catch would be DoubleDoku, Gattai-3, Butterfly, Flower and Kazagurama (Windmill) as these have a high proportion of overlap cells. In some cases a cell can be a member of 3 or even 4 grids. This might help check out any code.
The narrowness of intersection zones of in the more popular Samurai and Clueless (9 cells only) may restrict the types of fish permissible.
Glyn
 
Posts: 357
Joined: 26 April 2007

Postby Jean-Christophe » Tue Jul 29, 2008 7:18 pm

In my soft JSudoku, I opted for "1 big puzzle" since this was a natural extension to the variants I already supproted.

Here is a samurai which requires patterns spanning several sub-grids. JSudoku used some XY-chains and ALS-XZ.

Code: Select all
..1..6.4.......98.7.6..2.......8...1...4..8.52.3.......1..4....42..........27....
..9....8.....7.2.95.1.3....9.....4.8..5..9......3.4..6......86....7.2..........7.
...3.8...............9.7...2.1............1249.6..........3........2........5....
..93..............6.7.5.........1.5..12..........46.3..7.9..5..8.....9.7.2.7.4...
...8..........1.9....7...622.4..7.........329.9.5....7....7...1.29.5......5.639..
Jean-Christophe
 
Posts: 149
Joined: 22 January 2006

Postby Pat » Wed Jul 30, 2008 6:11 am

hey tarek,

the following frog
is probably not what you had in mind--

Code: Select all
 . . . | . . . | . . .
 . . . | . . . | . . .
 . . . | . . . | . . .
-------+-------+------
 . . . | . . . | . . .
 . . . | . . . | . . .
 . . . | . . . | . . .
-------+-------+-------
 . . . | . . . | . . . | . . .
 / / / | . . . | . . . | * * *
 . . . | . . . | . . . | . . .
        -------+-------+------
         . . . | . . . | . . .
         . . . | . . . | . . .
         . . . | . . . | . . .
        -------+-------+------
         . . . | . . . | . . .
         . . . | . . . | . . .
         . . . | . . . | . . .

User avatar
Pat
 
Posts: 4056
Joined: 18 July 2005

Postby tarek » Wed Jul 30, 2008 1:33 pm

Pat wrote:hey tarek,

the following frog
is probably not what you had in mind--
I was thinking of something like this:
Code: Select all
 . . . | . * . | . . .
 . . . | . * . | . . .
 . . . | . * . | . . .
-------+-------+------
 . . . | . * . | . . .
 . . . | . * . | . . .
 . . . | . * . | . . .
-------+-------+-------
 . . . | . X . | X / X | / / / | / / /
 . . . | . / . | * * * | . . . | . . .
 / / / | / X / | X / X | . . . | . . .
                -------+-------+------
                 . . . | . . . | . . .
                 . . . | . . . | . . .
                 . . . | . . . | . . .
                -------+-------+------
                 . . . | . . . | . . .
                 . . . | . . . | . . .
                 . . . | . . . | . . .
g1r9g2r1/g1b9g1c5 ==> [g1r123456c5][g1r8c789]<>X


(For he example, ignore the g2r1/g2b1 intersetion)

The frog you've posted has a LOL relationship where *** should be the same as ///.

It is easier to demonstrate it with a bigger fish

tarek
User avatar
tarek
 
Posts: 3762
Joined: 05 January 2006

Postby Jean-Christophe » Wed Jul 30, 2008 4:04 pm

tarek wrote:I was thinking of something like this:
Code: Select all
 . . . | . * . | . . .
 . . . | . * . | . . .
 . . . | . * . | . . .
-------+-------+------
 . . . | . * . | . . .
 . . . | . * . | . . .
 . . . | . * . | . . .
-------+-------+-------
 . . . | . X . | X / X | / / / | / / /
 . . . | . / . | * * * | . . . | . . .
 / / / | / X / | X / X | . . . | . . .
                -------+-------+------
                 . . . | . . . | . . .
                 . . . | . . . | . . .
                 . . . | . . . | . . .
                -------+-------+------
                 . . . | . . . | . . .
                 . . . | . . . | . . .
                 . . . | . . . | . . .
g1r9g2r1/g1b9g1c5 ==> [g1r123456c5][g1r8c789]<>X


No need of any fish here. Simple intersection gives:
Code: Select all
X for g2r1 locked @ g2r1c13 -> locked for g2b1, g1r7 -> g1r9c5 = X (HS)


tarek wrote:The frog you've posted has a LOL relationship where *** should be the same as ///.

Yes, it's called Twin nonets
Jean-Christophe
 
Posts: 149
Joined: 22 January 2006

Postby tarek » Thu Jul 31, 2008 6:06 am

Jean-Christophe wrote:No need of any fish here. Simple intersection gives:[code]X for g2r1 locked @ g2r1c13 -> locked for g2b1, g1r7 -> g1r9c5 = X (HS
You obviously missed that I asked you to ignore this fact for the purpose of demonstration the x-wing in the lines that followed the diagram.

The fact is, we need a better example:D

The ALS are great & will force you to look at the puzzle as 1 entity.

BTW, intersections are 1-fish & should follow the same logic, a bit more difficult though for them to span different grids & remain fruitful.

tarek
User avatar
tarek
 
Posts: 3762
Joined: 05 January 2006

Postby Pat » Fri Aug 01, 2008 7:07 am

Jean-Christophe, thanks for providing the SuDoPedia reference

tarek, the twin nonets -- my un-exciting frog -- is the closest i could come to anything which might resemble fish. what i'm trying to say is, i really have no expectation of finding any fruitful new type of fish here. ( i say nothing about ALS or any other methods. )
User avatar
Pat
 
Posts: 4056
Joined: 18 July 2005

Postby tarek » Fri Aug 01, 2008 8:25 am

Pat wrote:what i'm trying to say is, i really have no expectation of finding any fruitful new type of fish here. ( i say nothing about ALS or any other methods. )
I also harbour no such expectations. The exciting thing to me was to find mutant fish with (base/cover) sectors coming form different grids.

tarek
User avatar
tarek
 
Posts: 3762
Joined: 05 January 2006


Return to Sudoku variants