| View previous topic :: View next topic |
| Author |
Message |
| elroy
| | Joined: 16 Mar 2006 | | Posts: 6 | | : | | Items |
|
Posted: Thu Mar 16, 2006 2:57 pm Post subject: What is SimpleSoduko really telling me? |
|
|
SimpleSoduko is neat although admittedly the only one I have --
I used it to (auto)crack following solveable puzzle but it still required GUESSWORK, got lots of "NO HINTS AVAILABLE" messages until things got rolling (it does have at least ONE solution)..
>
I know SS's prime purpose isn't solving imported puzzles, but what is SimpleSoduko telling me about such a puzzle -- that it's perhaps invalid, maybe with more than one solution, or simply that it's a hard one to solve
>
8__4__6__
_3__6__7_
__7_____2
1__6__3__
_7__4__8_
__3__9__6
___8__4__
_9__5__2_
__1_____5 (courtesy of tso at sudoku.com forum demonstating puzzles with (perhaps none) bi-valued candidates (this has just one))
TIA -- elroy (don't know where else to post such product-specific question) |
|
| Back to top |
|
 |
| Henk
 | | Joined: 13 Nov 2005 | | Posts: 105 | | : | | Items |
|
Posted: Thu Mar 16, 2006 3:42 pm Post subject: |
|
|
For puzzles that need more advanced techniques, Simple Sudoku probably isn't the best choice. You might want to look at one of these programs:
Into Sudoku (My own program, supports most techniques until triple implication chains, and it contains an quick overview of whats in the puzzle)
Sudoku Cue (I didn't test the latest version because I do not have the .Net framework, but it has some nice techniques in it)
Sudoku Susser (Pretty good, but it uses a lot of guessing techniques like nishio and tabling)
Just google on the name, and you will find the program...
But neither of these programs can break the puzzle you posted with pure logic only. This requires an extremely advanced and complete arsenal of techniques. _________________ Generate and solve Sudoku puzzles with Into Sudoku! |
|
| Back to top |
|
 |
| elroy
| | Joined: 16 Mar 2006 | | Posts: 6 | | : | | Items |
|
Posted: Thu Mar 16, 2006 4:46 pm Post subject: |
|
|
Many many thanks, Henk, I'll certainly follow up your input.
BTW, I've subsequently discovered author (Vidar) of that puzzle (it's called Monster#2) from another forum&thread (if interested): http://www.sudoku.com/forums/viewtopic.php?t=3499
It all making sense NOW, thank you very much again. -- elroy  |
|
| Back to top |
|
 |
| tarek
| | Joined: 31 Dec 2005 | | Posts: 153 | | : | | Location: London, UK | Items |
|
Posted: Thu Mar 16, 2006 4:59 pm Post subject: |
|
|
I remeber this one from vidarino,
It is an example of a NO nonesense puzzle, straight into business....
this is the only step that my solver struggled with (one contradiction elimination was required)........
| Code: | *-----------------------------------------------------------------*
| 8 125 259 | 4 12379 12357 | 6 1359 139 |
| 2459 3 2459 | 1259 6 1258 | 1589 7 1489 |
| 4569 1456 7 | 1359 1389 1358 | 1589 13459 2 |
|---------------------+---------------------+---------------------|
| 1 2458 24589 | 6 278 2578 | 3 459 479 |
| 2569 7 2569 | 135 4 135 | 1259 8 19 |
| 245 2458 3 | 1257 1278 9 | 1257 145 6 |
|---------------------+---------------------+---------------------|
| 23567 256 256 | 8 12379 12367 | 4 1369 1379 |
| 3467 9 468 | 137 5 13467 | 178 2 1378 |
| 23467 2468 1 | 2379 2379 23467 | 789 369 5 |
*-----------------------------------------------------------------* |
This my final Analysis log | Code: | Difficulty Score: 137300 (Ultimate)
Non single moves: 39
Advanced moves: 25
Difficulty Index: 0.64
Non Simple xy Chains: 8
Non Simple chain Index: 0.32
Guesses: 1
Guessing Index: 0.04 |
Interesting to see what's the programmers' take on this (including you Vidar )
Tarek |
|
| Back to top |
|
 |
| Ruud Site Admin
 | | Joined: 17 Sep 2005 | | Posts: 708 | | : | | Location: Netherlands | Items |
|
Posted: Fri Mar 17, 2006 3:31 pm Post subject: |
|
|
There are 5 backdoors (magic cells) in this one:
R8C3, R8C6, R8C7, R9C2, R9C7
I cannot find a single step that isolates any one of them. 12 ugly chains required by my solver...
Ruud. |
|
| Back to top |
|
 |
| vidarino
| | Joined: 10 Feb 2006 | | Posts: 38 | | : | | Location: Haugesund, Norway | Items |
|
Posted: Fri Mar 17, 2006 6:07 pm Post subject: |
|
|
Glad you like my puzzle. It's a tricky bastard, allright...
My solver opens with a few nice loops with groups, the first two being;
| Code: |
R2C9-1-R5C9-9-[R5C13]=9=R4C3=8=R8C3-8-R8C9=8=R2C9 -> R2C9 <> 1
R6C7-1-R5C9-9-[R5C13]=9=R4C3=8=R8C3-8-R8C9=8=R2C9=4=R4C9=7=R6C7 -> R6C7 <> 1
|
... but it gets ugly really fast afterwards.
It gets through it without blind trial and error, but some of the forcing chain/net steps are as close to trial and error as you can get. One step is a forcing net(?) where each of four candidates in one cell eliminate a candidate in another cell, and all of the implication chains were long and quite unwieldy. (My solver doesn't output them yet, so I had to work them out by hand afterwards. One word; oof.)
Oh, and by the way, I have just set free another monster in this thread on the Players' Forum.
Vidar |
|
| Back to top |
|
 |
|