SudokuSolver Forum
http://www.rcbroughton.co.uk/sudoku/forum/

Some Gattai-3 puzzle
http://www.rcbroughton.co.uk/sudoku/forum/viewtopic.php?f=4&t=861
Page 1 of 1

Author:  dyitto [ Sun Jan 09, 2011 6:52 pm ]
Post subject:  Some Gattai-3 puzzle

Image

http://www.scrybqj.com/images_diversen/ ... i_3_23.pdf

Code lines:
Code:
095017000000000500700000000000004000000008320001000070006800000020060000400000000
002001050000000000700006800000009000080000604000000000000030000000000005000400320
000000000000000000000000000000000000000000000000000689000001000048000000000090507


Extra constraint is: the yellow 2x2-regions have a killer sum of 23.

I'd appreciate your opinion on this puzzle.

Ultimately I want to explore new ways of connecting multiple sudoku grids in one puzzle - apart from overlapping regions.
But if this puzzle is fine then I know my program is already basically working.

Author:  enxio27 [ Mon Jan 10, 2011 11:03 pm ]
Post subject:  Re: Some Gattai-3 puzzle

Darn! I was ready to go until I saw the killer aspect. I know most users here seem to favor the killer variety, but unfortunately, I'm not one of them.

Author:  dyitto [ Tue Jan 11, 2011 6:46 am ]
Post subject:  Re: Some Gattai-3 puzzle

Then here's one without the killer aspect:

Image

http://www.scrybqj.com/images_diversen/ ... killer.pdf

Code:
000540070000083000009600002002000004906000000100800000000010500000900000000000003
000074006000000502500000100004300000000000000000106400000000000000000060000040010
006000000000000000004000000000000000000000740300005020000003000000500000500004600

Author:  Ed [ Mon Jan 17, 2011 9:41 pm ]
Post subject:  Re: Some Gattai-3 puzzle

dyitto wrote:
I'd appreciate your opinion on this puzzle...Ultimately I want to explore new ways of connecting multiple sudoku grids in one puzzle - apart from overlapping regions.
But if this puzzle is fine then I know my program is already basically working.
The killer version of this variant starts well. Suspect I made a slip-up so will start again (doing this one on paper). Very easy to get tricked with those not-completely-overlapping grids.

Here are some killer steps that are available from the beginning. Will let you know how I went with the rest after a holiday.
2 steps:
1. 4 in G3n1 only in r23c3 in 23(4) cage
1a. ->23(4) must have 4 = {2489} (only valid combo)
1b. ->G3 r2c2 = 2, G3 r3c2 = 8
1c. G3 r23c3 = {49}: both locked G1n9 & c8, G3n1 & c3

2. 2 and 3 for G2n7 must be at r8c123
2a. -> 23(4) cage at G2r7c2 must have 2 or 3
2b. = {2579/3479/3569}(no 1)
2c. must have 9: 9 locked for G2n7; G3n3
2d. 23(4) can't have both 2 & 3 -> G2 r8c23 can have one of 2,3 -> G2 r8c1 can only be from (23)

Great to have another variant maker here.
Ed

Author:  Ed [ Mon Jan 24, 2011 8:27 pm ]
Post subject:  Re: Some Gattai-3 puzzle

Finished off the killer one. Just as well I started again because had made an error. It was basically singles for the rest but was still slow going. The twist of those weird half over-lapping nonets made it interesting. Thanks for the puzzle.

This is the solution...hard to convey clearly.
Hidden Text:
+-------+-------+-------+---+-------+-------+-------+
| 2 9 5 | 3 1 7 | 4 8 6 | | 9 6 2 | 3 8 1 | 4 5 7 |
| 1 6 4 | 9 8 2 | 5 3 7 | | 1 3 8 | 7 4 5 | 9 6 2 |
| 7 8 3 | 4 5 6 | 9 1 2 | | 7 5 4 | 9 2 6 | 8 3 1 |
+-------+-------+-------+---+-------+-------+-------+
| 9 3 2 | 6 7 4 | 1 5 8 | | 4 7 6 | 1 5 9 | 2 8 3 |
| 6 5 7 | 1 9 8 | 3 2 4 |---| 5 8 9 | 2 7 3 | 6 1 4 |
| 8 4 1 | 2 3-5 6 7-9 4 3-2 1 8-6 4 | 5 7 9 |
+-------+----- ----+-------+
| 5 7 6 | 8 4 3 2 9 1 8 6 4 7 5 3 2 | 1 9 8 |
| 3 2 9 | 7 6 1 8 4 5 7 2 9 3 6 1 8 | 7 4 5 |
| 4 1 8 | 5 2-9 7 6-3 2 8-1 5 4-9 7 | 3 2 6 |
+-------+----- ----+-------+
| 8 5 1 4 6 9 7 2 3 |
| 4 3 2 7 1 5 6 8 9 |
+--------+---------+--------+
| 7 9 5 6 3 1 8 4 2 |
| 6 4 8 2 5 7 3 9 1 |
| 2 1 3 8 9 4 5 6 7 |
+--------+---------+--------+
Cheers
Ed

Author:  dyitto [ Tue Jan 25, 2011 9:05 pm ]
Post subject:  Re: Some Gattai-3 puzzle

Thanks for the solution! It matches mine. :)

I must admit that I'm not much of a killer solver myself and ruined this puzzle twice with pen and paper.
I didn't do a 3rd attempt because of some unfinished business elsewhere.

Hope I'll get some more ideas on how to connect multiple sudokus into one puzzle.

Author:  enxio27 [ Mon Mar 26, 2012 11:49 pm ]
Post subject:  Re: Some Gattai-3 puzzle

Richard, is there any way SudokuSolver could be made to handle these puzzles? Do I understand correctly that you have/had planned to enable SudokuSolver to handle user-defined multi-grid configurations? If so, what is the status on that?

Page 1 of 1 All times are UTC
Powered by phpBB® Forum Software © phpBB Group
https://www.phpbb.com/