| View previous topic :: View next topic |
| Author |
Message |
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
 Posted: Wed Apr 14, 2010 9:17 pm Post subject: Puzzle 10/04/14 (B) |
 |
|
| Code: | +-----------------------+
| 2 . . | . . 9 | . . . |
| . 6 . | . . . | . . . |
| . . 7 | . 8 . | . . 2 |
|-------+-------+-------|
| . . . | 4 . . | . . 1 |
| . . 6 | . 7 . | . 8 5 |
| 4 . . | . . 3 | . 7 6 |
|-------+-------+-------|
| . . . | . . . | . . 8 |
| . . . | . 1 6 | . 9 . |
| . . 1 | 2 9 8 | 6 . 7 |
+-----------------------+
|
Play this puzzle online at the Daily Sudoku site |
|
| Back to top |
|
 |
tlanglet
Joined: 17 Oct 2007
Posts: 2468
Location: Northern California Foothills
|
| Posted: Thu Apr 15, 2010 4:25 am Post subject: |
|
|
Another move of questionable validity. The AUR78 in r48c12 forces r4c2=59 or r8c2=34.Looking at these two requirements separately, we find:
(59)r4c23 - (59=8)r4c3,
and
(34)r89c2 - (34=5)r3c2 - (5)r1c2
||
(34)r89c2 - (34=8)r1c2.
Thus, r1c3<>8. Is this valid, or do I have to treat each of the 4 digits, 3459, individually?
Chain found chasing a potential xy=wing:(4=3)r1c9 - r8c9 = r9c8 - (3=2)r4c8 - (2=9)r6c7 - (9=5)r6c3 - r1c3 = r1c5 - (5=4)r3c6; r1c5|r3c8<>4
x-wing 5 r16c35
type 1UR16 deletes 16 from r3c8,
xy-wing 3-45 with vertex in r1c5; rr2c5|r7c6<>4,
This brought me to another step of chaos!
| Code: |
*-----------------------------------------------------------*
| 2 8 #345 | 1 35 9 | 7 6 #34 |
| 1 6 #34 | 57 23 2457 | 8 #345 9 |
| 9 345 7 | 6 8 45 | 135 135 2 |
|-------------------+-------------------+-------------------|
| 78 579 89 | 4 6 25 | 239 23 1 |
| 3 2 6 | 9 7 1 | 4 8 5 |
| 4 1 59 | 8 25 3 | 29 7 6 |
|-------------------+-------------------+-------------------|
| 6 379 239 | 357 4 57 | 125 125 8 |
| 78 #347 2348 | 357 1 6 | 25 9 #34 |
| 5 #34 1 | 2 9 8 | 6 #34 7 |
*-----------------------------------------------------------* |
Notice the (hopefully valid) 8-cell DP34, marked # in r1c39, r2c38, r8c29 & r9c28. To prevent the deadly pattern, r1c3=5, r2c8=5 or r8c2=7.
(5)r1c3 - (5=9)r6c3 - r7c3 = (9-7)r7c2 = (7)r7c46; r8c4<>7
||
(5)r2c8 - (5=7)r2c4; r8c4<>7
||
(7)r8c1; r8c4<>7.
Another of my "funny" xyz-wings 345 in r1c3 (See code above for this move also.)
(3)r1c3; r3c2<>3
||
(4)r1c3 - (4=3)r2c3; r3c2<>3
||
(5)r1c3 - (5=3)r1c5 - r1c9 = r8c9 - r9c8 = (3)c9c2; r3c2<>3
Danny, you did not flag this puzzle as a BBDB, but I made it one for myself. Really just lots of fun.
Ted |
|
| Back to top |
|
|
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Thu Apr 15, 2010 5:41 am Post subject: |
|
|
| tlanglet wrote: | Another move of questionable validity. The AUR78 in r48c12 forces r4c2=59 or r8c2=34.Looking at these two requirements separately, we find:
(59)r4c23 - (59=8)r4c3,
and
(34)r89c2 - (34=5)r3c2 - (5)r1c2
||
(34)r89c2 - (34=8)r1c2.
Thus, r1c3<>8. Is this valid, or do I have to treat each of the 4 digits, 3459, individually?
Danny, you did not flag this puzzle as a BBDB, but I made it one for myself. Really just lots of fun.
|
My limited thoughts. I'm sure the notation can be expanded, but I'm happy with it as shown.
| Code: | (8=345)r139c2 - UR[(34)r8c2 = (59)r4c2] - (59=8)r46c3 => r1c3<>8
+-----------------------------------------------------------------------+
| 2 3458 3458 | 16 345 9 | 7 16 34 |
| 1 6 345 | 357 2345 2457 | 8 345 9 |
| 9 345 7 | 16 8 45 | 135 13456 2 |
|-----------------------+-----------------------+-----------------------|
| 78 78+59 589 | 4 6 25 | 239 23 1 |
| 3 2 6 | 9 7 1 | 4 8 5 |
| 4 1 59 | 8 25 3 | 29 7 6 |
|-----------------------+-----------------------+-----------------------|
| 6 379 239 | 357 345 457 | 125 125 8 |
| 78 78+34 2348 | 357 1 6 | 25 9 34 |
| 5 34 1 | 2 9 8 | 6 34 7 |
+-----------------------------------------------------------------------+
# 73 eliminations remain
|
Ted: "lots of fun" is the objective for every one of my puzzles. Thanks! |
|
| Back to top |
|
|
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Thu Apr 15, 2010 1:14 pm Post subject: |
|
|
I once wrote a (poorly received) post where I compared the AIC to forcing chain logic. Consider my chain above.
| Code: | (8=345)r139c2 - UR[(34)r8c2 = (59)r4c2] - (59=8)r46c3 => r1c3<>8
|
I split the AIC into two lines with one strong inference term duplicated. I'm also going to remove the UR qualifier and the conclusion (for now).
| Code: | (8=345)r139c2 - (34)r8c2
(34)r8c2 = (59)r4c2 - (59=8)r46c3
|
I reverse the direction of the top line.
| Code: | (34)r8c2 - (345=8)r139c2
(34)r8c2 = (59)r4c2 - (59 =8)r46c3
|
I add the UR and conclusion back into the notation.
| Code: | (34)r8c2 - (345=8)r139c2 => r1c3<>8
UR[(34)r8c2 = (59)r4c2] - (59 =8)r46c3 => r1c3<>8
|
I now have two forcing chain streams based on r8c2=34 or r8c2<>34. Does the placement of the weak inferences (-) remind anyone of Ted's original approach of r8c2=34 or r4c2=59 ? |
|
| Back to top |
|
|
arkietech
Joined: 31 Jul 2008
Posts: 1834
Location: Northwest Arkansas USA
|
| Posted: Thu Apr 15, 2010 1:16 pm Post subject: |
|
|
| tlanglet wrote: | | Danny, you did not flag this puzzle as a BBDB, but I made it one for myself. Really just lots of fun. |
There is a fish swimming about that helps. Good puzzle Danny!  |
|
| Back to top |
|
|
tlanglet
Joined: 17 Oct 2007
Posts: 2468
Location: Northern California Foothills
|
| Posted: Thu Apr 15, 2010 2:15 pm Post subject: |
|
|
Danny, thanks for your perspective; your AIC is much simpler and easier to understand.
Also, I read your post on forcing chains vs AICs but need to ponder at length on the issues.
Ted |
|
| Back to top |
|
|
Luke451
Joined: 20 Apr 2008
Posts: 310
Location: Southern Northern California
|
| Posted: Thu Apr 15, 2010 9:36 pm Post subject: |
|
|
| tlanglet wrote: | Another move of questionable validity. The AUR78 in r48c12 forces r4c2=59 or r8c2=34.Looking at these two requirements separately, we find:
(59)r4c23 - (59=8)r4c3,
and
(34)r89c2 - (34=5)r3c2 - (5)r1c2
||
(34)r89c2 - (34=8)r1c2.
Thus, r1c3<>8. Is this valid, or do I have to treat each of the 4 digits, 3459, individually?
Chain found chasing a potential xy=wing:(4=3)r1c9 - r8c9 = r9c8 - (3=2)r4c8 - (2=9)r6c7 - (9=5)r6c3 - r1c3 = r1c5 - (5=4)r3c6; r1c5|r3c8<>4
x-wing 5 r16c35
type 1UR16 deletes 16 from r3c8,
xy-wing 3-45 with vertex in r1c5; rr2c5|r7c6<>4,
This brought me to another step of chaos!
| Code: |
*-----------------------------------------------------------*
| 2 8 #345 | 1 35 9 | 7 6 #34 |
| 1 6 #34 | 57 23 2457 | 8 #345 9 |
| 9 345 7 | 6 8 45 | 135 135 2 |
|-------------------+-------------------+-------------------|
| 78 579 89 | 4 6 25 | 239 23 1 |
| 3 2 6 | 9 7 1 | 4 8 5 |
| 4 1 59 | 8 25 3 | 29 7 6 |
|-------------------+-------------------+-------------------|
| 6 379 239 | 357 4 57 | 125 125 8 |
| 78 #347 2348 | 357 1 6 | 25 9 #34 |
| 5 #34 1 | 2 9 8 | 6 #34 7 |
*-----------------------------------------------------------* |
Notice the (hopefully valid) 8-cell DP34, marked # in r1c39, r2c38, r8c29 & r9c28. To prevent the deadly pattern, r1c3=5, r2c8=5 or r8c2=7.
(5)r1c3 - (5=9)r6c3 - r7c3 = (9-7)r7c2 = (7)r7c46; r8c4<>7
||
(5)r2c8 - (5=7)r2c4; r8c4<>7
||
(7)r8c1; r8c4<>7.
Another of my "funny" xyz-wings 345 in r1c3 (See code above for this move also.)
(3)r1c3; r3c2<>3
||
(4)r1c3 - (4=3)r2c3; r3c2<>3
||
(5)r1c3 - (5=3)r1c5 - r1c9 = r8c9 - r9c8 = (3)c9c2; r3c2<>3
Danny, you did not flag this puzzle as a BBDB, but I made it one for myself. Really just lots of fun.
Ted |
Man, that's some creative stuff! It's good to think outside the box.
I'm buying it all, except..
I'm concerned about the (34) DP. What doesn't seem right is the back to back diagonal cells in columns 89. I've never seen that done, but then I sure haven't seen it all.
I do remember Myth Jellies once trying two diagonals in a six cell pattern, but he soon returned to retract it as a DP. I'd post the link, but the world has ended over at Players (for now.)
Maybe someone can confirm or deny this one with some proof. As for me, it seems that (5)r2c8 breaks up any DP possibilities, and it's been well over 10 minutes since I was last mistaken . |
|
| Back to top |
|
|
tlanglet
Joined: 17 Oct 2007
Posts: 2468
Location: Northern California Foothills
|
| Posted: Fri Apr 16, 2010 12:45 am Post subject: |
|
|
Thanks for the comments Luke. As you know, I never got to read the thread about valid DPs on the Players forum before it crashed. For the present, if the pattern involves N cells, and N/2 rows, columns and houses, then I assume it is valid.
Maybe I should/could send this code to Myth if I do not get any other input soon.
Ted |
|
| Back to top |
|
|
daj95376
Joined: 23 Aug 2008
Posts: 3854
|
| Posted: Fri Apr 16, 2010 2:46 am Post subject: |
|
|
I'm a novice when it comes to DPs. I can seldom spot them, but I test those with two candidates using XY-Chain loops.
| Code: | *-----------------------------------------------------------*
| 2 8 34+5 | 1 35 9 | 7 6 34 |
| 1 6 34 | 57 23 2457 | 8 34+5 9 |
| 9 345 7 | 6 8 45 | 135 135 2 |
|-------------------+-------------------+-------------------|
| 78 579 89 | 4 6 25 | 239 23 1 |
| 3 2 6 | 9 7 1 | 4 8 5 |
| 4 1 59 | 8 25 3 | 29 7 6 |
|-------------------+-------------------+-------------------|
| 6 379 239 | 357 4 57 | 125 125 8 |
| 78 34+7 2348 | 357 1 6 | 25 9 34 |
| 5 34 1 | 2 9 8 | 6 34 7 |
*-----------------------------------------------------------*
|
If r1c3,r2c8<>5 and r8c2<>7:
(4=3)r1c3 - (3=4)r1c9 - (4=3)r8c9 - (3=4)r8c2 - (4=3)r9c2 - (3=4)r9c8 - (4=3)r2c8 - (3=4)r2c3 - loop
-or-
(3=4)r1c3 - (4=3)r1c9 - (3=4)r8c9 - (4=3)r8c2 - (3=4)r9c2 - (4=3)r9c8 - (3=4)r2c8 - (4=3)r2c3 - loop
This can also be viewed as coloring on the DP cells:
(blue)r1c3 - (green)r1c9 - (blue)r8c9 - (green)r8c2 - (blue)r9c2 - (green)r9c8 - (blue)r2c8 - (green)r2c3 - loop
There isn't any coloring conflict, and each cell is colored only once.
I think Ted has a DP. |
|
| Back to top |
|
|
Marty R.
Joined: 12 Feb 2006
Posts: 5770
Location: Rochester, NY, USA
|
| Posted: Fri Apr 16, 2010 3:47 am Post subject: |
|
|
Type 4 UR (16)
Remote pairs (34)
DP (25-35-23) in boxes 789. There's an 89 pseudo cell in c3 which forms a pair to finish it off. |
|
| Back to top |
|
|
Luke451
Joined: 20 Apr 2008
Posts: 310
Location: Southern Northern California
|
| Posted: Fri Apr 16, 2010 4:55 am Post subject: |
|
|
| daj95376 wrote: | I'm a novice when it comes to DPs. I can seldom spot them, but I test those with two candidates using XY-Chain loops.
| Code: | *-----------------------------------------------------------*
| 2 8 34+5 | 1 35 9 | 7 6 34 |
| 1 6 34 | 57 23 2457 | 8 34+5 9 |
| 9 345 7 | 6 8 45 | 135 135 2 |
|-------------------+-------------------+-------------------|
| 78 579 89 | 4 6 25 | 239 23 1 |
| 3 2 6 | 9 7 1 | 4 8 5 |
| 4 1 59 | 8 25 3 | 29 7 6 |
|-------------------+-------------------+-------------------|
| 6 379 239 | 357 4 57 | 125 125 8 |
| 78 34+7 2348 | 357 1 6 | 25 9 34 |
| 5 34 1 | 2 9 8 | 6 34 7 |
*-----------------------------------------------------------*
|
If r1c3,r2c8<>5 and r8c2<>7:
(4=3)r1c3 - (3=4)r1c9 - (4=3)r8c9 - (3=4)r8c2 - (4=3)r9c2 - (3=4)r9c8 - (4=3)r2c8 - (3=4)r2c3 - loop
-or-
(3=4)r1c3 - (4=3)r1c9 - (3=4)r8c9 - (4=3)r8c2 - (3=4)r9c2 - (4=3)r9c8 - (3=4)r2c8 - (4=3)r2c3 - loop
This can also be viewed as coloring on the DP cells:
(blue)r1c3 - (green)r1c9 - (blue)r8c9 - (green)r8c2 - (blue)r9c2 - (green)r9c8 - (blue)r2c8 - (green)r2c3 - loop
There isn't any coloring conflict, and each cell is colored only once.
I think Ted has a DP. |
This is very interesting 
Danny, your continuous loops do prove two solutions if the the tri-sis is false. The second set of diagonal cells serve to extend the pattern into a conventional DP. I think y'all are onto something!
Very well done, Ted, and thanks, Danny, for the perspective. |
|
| Back to top |
|
|
Luke451
Joined: 20 Apr 2008
Posts: 310
Location: Southern Northern California
|
| Posted: Sat Apr 17, 2010 2:35 pm Post subject: |
|
|
Just a quick follow up on this one for anyone interested.
Check out this Sudopedia article on Deadly Patterns. The fourth one cited under "8 cells" is the one Ted used above. |
|
| 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
