| View previous topic :: View next topic |
| Author |
Message |
keith
Joined: 19 Sep 2005
Posts: 3355
Location: near Detroit, Michigan, USA
|
 Posted: Thu Oct 27, 2011 11:55 pm Post subject: |
 |
|
| Quote: | | Ironically, if you take the grid and paste it into Simple Sudoku |
Danny,
I assume you mean "the candidate grid".
Right, how is SS to know the difference between a given cell and a solved cell or naked single?
This absolutely matters if you are subsequently going to make any uniqueness or Deadly Pattern argument.
Keith |
|
| Back to top |
|
 |
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Fri Oct 28, 2011 4:59 am Post subject: |
|
|
| keith wrote: | Right, how is SS to know the difference between a given cell and a solved cell or naked single?
This absolutely matters if you are subsequently going to make any uniqueness or Deadly Pattern argument.
|
Your question is at the heart of handling a candidate grid.
A given/clue cell often can't be distinguished from a solved cell in a candidate grid. In some cases, it's possible to determine that a puzzle would have multiple solutions if a cell was "solved" instead of a given/clue. But this relies on the original puzzle being "minimal" in the first place. Most of the puzzle that I'd previously posted weren't minimal. Thus, any partially solved candidate grid posted from them would never be able to return the original puzzle.
What's even more complicated is determining the status of a (possible) Naked Single. Is it a given/clue whose peer candidates haven't been eliminated, or is it just a Naked Single. Again, everything hinges on the possibility of multiple solutions and an original puzzle that was minimal.
HoDoKu only treats a single candidate in a cell as a given/clue if the candidate is missing from all of the peer cells. Otherwise, the cell is treated as if it contains a Naked Single, and the peer candidates are eliminated as a step. I previously updated my solver to use this same logic.
However, I ran into grids that, once loaded into my solver, would be flagged as having multiple solutions. Others pointed out that there was a given/clue present in a candidate cell that had not been resolved by the person posting the candidate grid. Now, I'm caught between a rock and a hard spot on how to treat a candidate cell with only one candidate _  _
Bottom Line: Always include the original puzzle along with any candidate grid. Some solvers -- like HoDoKu and mine -- can accept both an original puzzle and a partially solved candidate grid.
Regards, Danny |
|
| Back to top |
|
|
Pat
Joined: 23 Feb 2010
Posts: 207
|
| Posted: Fri Oct 28, 2011 8:04 am Post subject: re: "Meister Sudokus" puzzle 40 |
|
|
re: "Meister Sudokus" puzzle 40
24 cells given + 14 "singles" = 38:
.6...4.1...4.2..6.3.586.24765...719..4.6..72..7.1...364....69515..48..7....51..8.
[ play ]
| Code: | +-------+-------+-------+
| . 6 . | . . 4 | . 1 . |
| . . 4 | . 2 . | . 6 . |
| 3 . 5 | 8 6 . | 2 4 7 |
+-------+-------+-------+
| 6 5 . | . . 7 | 1 9 . |
| . 4 . | 6 . . | 7 2 . |
| . 7 . | 1 . . | . 3 6 |
+-------+-------+-------+
| 4 . . | . . 6 | 9 5 1 |
| 5 . . | 4 8 . | . 7 . |
| . . . | 5 1 . | . 8 . |
+-------+-------+-------+
|
glad to see you had some fun with this one
by arkietech's rating, this is Easy,
as it only needs "intersections" ( box-line interactions )
but it does need 3 moves
and what i liked especially is the pincers situation --
each move itself solves nothing,
we must combine them to solve r7c3 --
(2) c2\b7
(3) c2\b7 -- or if you prefer, (3) b4\c3
(7) r9\b7 -- or if you prefer, (7) b8\r7 |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
