View previous topic :: View next topic |
Author |
Message |
David Bryant
Joined: 29 Jul 2005 Posts: 559 Location: Denver, Colorado
|
Posted: Thu Mar 09, 2006 4:18 am Post subject: A very tough minimal sudoku |
|
|
Here's a very difficult 17-clue sudoku puzzle. Someone_somewhere sent me this puzzle in a private message. I'm not yet sure how to reason my way through to the solution.
Code: | 6.2.5....
.....4.3.
.........
43...8...
.1....2..
......7..
5..27....
.......81
...6..... |
Oh -- s_s called this a "zen" sudoku, apparently implying that it can be solved if one thinks about it (becomes it?) hard enough. dcb |
|
Back to top |
|
![](templates/subSilver/images/spacer.gif) |
someone_somewhere
Joined: 07 Aug 2005 Posts: 275 Location: Munich
|
Posted: Sun Mar 12, 2006 2:12 pm Post subject: |
|
|
Yes, I called it Zen.
No single response in this forum is confirming it's name.
I know how to get to the solution, but it takes a long time, like in the Zen training. No brute force is needed.
If someone makes/made some progress on it, share it with us.
see u, |
|
Back to top |
|
![](templates/subSilver/images/spacer.gif) |
keith
Joined: 19 Sep 2005 Posts: 3355 Location: near Detroit, Michigan, USA
|
Posted: Sun Mar 12, 2006 10:09 pm Post subject: Position 1 |
|
|
This is about the toughest puzzle I have ever tried.
I found six pinned squares, and a pair in column 9. Then after another seven moves (intersections to remove possibilities, not to solve squares), I get to this:
Code: |
+----------------------+----------------------+----------------------+
| 6 4789 2 | 13789 5 379 | 1489 179 49 |
| 1789 5789 15789 | 1789 169 4 | 1689 3 2 |
| 13789 4789 134789 | 1789 169 2 | 14689 15679 4569 |
+----------------------+----------------------+----------------------+
| 4 3 5679 | 1579 2 8 | 169 1569 569 |
| 789 1 56789 | 34579 349 35679 | 2 4569 38 |
| 289 25689 5689 | 13459 1349 3569 | 7 14569 38 |
+----------------------+----------------------+----------------------+
| 5 4689 134689 | 2 7 139 | 3469 69 469 |
| 2379 2679 3679 | 3459 349 359 | 3569 8 1 |
| 139 49 1349 | 6 8 1359 | 3459 2 7 |
+----------------------+----------------------+----------------------+
|
I have played with this for a while, and I have no clue! The best I can suggest is to examine the squares that could be <4>.
Let's call this "Position 1". What next?
Keith |
|
Back to top |
|
![](templates/subSilver/images/spacer.gif) |
David Bryant
Joined: 29 Jul 2005 Posts: 559 Location: Denver, Colorado
|
Posted: Sun Mar 12, 2006 11:12 pm Post subject: There's a "coloring" move on the "7"s |
|
|
Hi, Keith!
I think we can eliminate one more possibility from the grid using "standard" methods. After that I think we'll have to start using "double-implication chains."
Code: | +----------------------+----------------------+----------------------+
| 6 4789 2 | 13789 5 379B | 1489 179A 49 |
| 1789 5789 15789 | 1789 169 4 | 1689 3 2 |
| 13789 4789 134789 | 1789 169 2 | 14689 15679a 4569 |
+----------------------+----------------------+----------------------+
| 4 3 5679c | 1579C 2 8 | 169 1569 569 |
| 789 1 56789 | 34579 349 35679b | 2 4569 38 |
| 289 25689 5689 | 13459 1349 3569 | 7 14569 38 |
+----------------------+----------------------+----------------------+
| 5 4689 134689 | 2 7 139 | 3469 69 469 |
| 2379 2679 3679 | 3459 349 359 | 3569 8 1 |
| 139 49 1349 | 6 8 1359 | 3459 2 7 |
+----------------------+----------------------+----------------------+ |
There are three binary chains in the "7"s, marked Aa, Bb, and Cc above.
1. r1c8 = 7 ==> r1c6 <> 7 ==> r5c6 = 7 ==> r4c4 <> 7 ==> r4c3 = 7 ==> r3c3 <> 7.
2. r1c8 <> 7 ==> r3c8 = 7 ==> r3c3 <> 7.
So we can eliminate the possible "7" at r3c3.
I'm still working on my next move. dcb |
|
Back to top |
|
![](templates/subSilver/images/spacer.gif) |
David Bryant
Joined: 29 Jul 2005 Posts: 559 Location: Denver, Colorado
|
Posted: Mon Mar 13, 2006 12:30 am Post subject: R1C9 = 9 |
|
|
I'm starting from the position previously posted, with the "7" removed from r3c3. We can demonstrate that r1c9 <> 4 as follows.
r1c9 = 4 ==> {6, 9} pair in r7c8&9 ==> {3, 4, 5, 9} quad in r8c456&7.
Making these eliminations the matrix looks like this.
Code: | 6 789 2 13789 5 379 189 179 4
1789 5789 15789 1789 169 4 1689 3 2
13789 4789 13489 1789 169 2 1689 15679 569
4 3 5679 1579 2 8 169 1569 569
789 1 56789 34579 349 35679 2 4569 38
289 25689 5689 13459 1349 3569 7 14569 38
5 48 1348 2 7 13 34 69 69
27 267 67 3459 349 359 35 8 1
139 49 1349 6 8 1359 345 2 7 |
But now we can spot another quad, {4, 7, 8, 9} in column 2. Making this elimination in the column also reveals that the "7" in column 2 must lie in the top left 3x3 box, so we can simplify the matrix like this.
Code: | 6 789 2 13789 5 379 189 179 4
189 5 189 1789 169 4 1689 3 2
1389 4789 13489 1789 169 2 1689 15679 569
4 3 5679 1579 2 8 169 1569 569
789 1 56789 34579 349 35679 2 4569 38
289 26 5689 13459 1349 3569 7 14569 38
5 48 1348 2 7 13 34 69 69
27 26 67 3459 349 359 35 8 1
139 49 1349 6 8 1359 345 2 7 |
And now we have a series of forced moves:
1. r2c4 = 7 (unique horizontal)
2. r5c6 = 7 (unique vertical)
3. r6c6 = 6 (unique vertical)
4. r6c2 = 2 (sole candidate)
5. r4c3 = 7 (unique horizontal -- because of the "7" at r2c4)
6. r8c3 = 6 (sole candidate)
7. r8c2 = 2 (sole candidate)
But we can't have two "2"s in column 2, so r1c9 <> 4, and r1c9 = 9. This leaves the matrix looking like this.
Code: | 6 478 2 1378 5 37 148 17 9
1789 5789 15789 1789 169 4 1689 3 2
13789 4789 13489 1789 169 2 14689 15679 456
4 3 5679 1579 2 8 169 1569 56
789 1 56789 34579 349 35679 2 4569 38
289 25689 5689 13459 1349 3569 7 14569 38
5 4689 134689 2 7 139 3469 69 46
2379 2679 3679 3459 349 359 3569 8 1
139 49 1349 6 8 1359 3459 2 7 |
From here we can see right away that r7c9 = 4, because if we put in the "6" it will be just like the case with the {6, 9} pair that we analyzed above. So here's where I'm going to leave it for now:
Code: | 6 478 2 1378 5 37 148 17 9
1789 5789 15789 1789 169 4 168 3 2
13789 4789 13489 1789 169 2 1468 1567 56
4 3 5679 1579 2 8 169 1569 56
789 1 56789 34579 349 35679 2 4569 38
289 25689 5689 13459 1349 3569 7 14569 38
5 689 13689 2 7 139 369 69 4
2379 2679 3679 3459 349 359 3569 8 1
139 49 1349 6 8 1359 359 2 7 |
|
|
Back to top |
|
![](templates/subSilver/images/spacer.gif) |
someone_somewhere
Joined: 07 Aug 2005 Posts: 275 Location: Munich
|
Posted: Mon Mar 13, 2006 8:59 am Post subject: |
|
|
Thank you, guys.
I used also double implication chains, but ... I had to travel to a couple of other galaxies, until I got to the solution.
It is ZEN.
Hope that you are finding a way, your way ...
Wake me up, when you have something for someone_somewhere
see u, |
|
Back to top |
|
![](templates/subSilver/images/spacer.gif) |
|
|
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
|