View previous topic :: View next topic |
Author |
Message |
| MCondron
| Joined: 17 Jul 2006 | Posts: 35 | : | Location: Houston, TX | Items |
|
Posted: Fri Dec 29, 2006 12:29 am Post subject: Concrete example of franken swordfish |
|
|
I don't think this is ground-breaking news, but here is an example of a franken swordfish. I have generated a few hundred of these, but so far this is the only one that can be fully solved using only more basic patterns (X-wing, swordfish, and jellyfish, along with finned and sashimi variations). I've spent some time making these to test my solving code. If anyone's interested, I can post more. I haven't seen many concrete examples of things like this or larger ones; are there many?
Here's the closest starting point:
Code: |
17...6..5.....7.......3.67...784...3....59...9...63.....4.7.28.7.9.......123..4.7
|
Some hidden singles, a line-box interaction, a finned swordfish, and a sashimi swordfish then lead to the franken swordfish. It looks like this.
(Cols 5,7,8 form the base/root set; rows 2,8 and box 6 form the cover/victim set, digit 1):
Code: |
..3|.2. .2. |..3 .23
1 7 ...|4.. ... 6 |... 4.. 5
.8.|..9 .89 |.89 ..9
| |
.23 .23 ..3|12. 12. |1.3 123 12.
456 456 .56|45. ... 7 |... 4.. 4..
.8. .89 .8.|..9 .8. |.89 ..9 .89
| |
.2. .2. ...|12. 12.| 12.
45. 45. .5.|45. 3 45.| 6 7 4..
.8. .89 .8.|..9 .8.| .89
===========+===========+===========
.2. .2. | 12.|1.. 12.
.56 .56 7 | 8 4 ...|.5. .56 3
... ... | ...|..9 ..9
| |
.23 .23 1.3|12. |1.. 12. .2.
4.. 4.6 ..6|... 5 9 |... 4.6 4.6
.8. .8. .8.|7.. |78. ... .8.
| |
.2. 1..|12. |1.. 12. .2.
9 45. .5.|... 6 3 |.5. 45. 4..
.8. .8.|7.. |78. ... .8.
===========+===========+===========
..3 ..3 |1.. 1..| 1..
.56 .56 4 |.56 7 .5.| 2 8 ..6
... ... |..9 ...| ..9
| |
... |12. 12. 12.|1.3 1.3 1..
7 .56 9 |456 ... 45.|.5. .56 ..6
.8. |... .8. .8.|... ... ...
| |
... | ... ...| ...
.56 1 2 | 3 ... .5.| 4 .56 7
.8. | .89 .8.| ..9
|
or, in the format that I think most people are used to seeing here (this was generated by hand from the above and may have a typo or two)
Code: |
1 7 38 | 249 289 6 | 389 2349 5
234568 2345689 3568 | -12459 *128 7 | *1389 *12349 -12489
2458 24589 58 | 12459 3 12458 | 6 7 12489
----------------------------------------------------------------
256 256 7 | 8 4 12 | *159 *12569 3
2348 23468 1368 | 127 5 9 | *178 *1246 2468
9 2458 158 | 127 6 3 | *1578 *1245 248
----------------------------------------------------------------
356 356 4 | 1569 7 15 | 2 8 169
7 568 9 | -12456 *128 -12458 | *135 *1356 -16
568 1 2 | 3 89 58 | 4 569 7
|
|
|
Back to top |
|
|
| daj95376
| Joined: 05 Feb 2006 | Posts: 349 | : | | Items |
|
Posted: Fri Dec 29, 2006 1:39 am Post subject: |
|
|
Interesting puzzle. Another perspective.
Code: | # your closest starting point:
# before finned X-Wing eliminations [r56c9]<>1
#
# Jellyfish c3578\r2568 -or-
# kraken cell [r4c7]=1 => [r56c9]<>1 and X-Wing c58\r28
# kraken cell [r4c8]=1 => [r56c9]<>1 and X-Wing c57\r28
*--------------------------------------------------------------------------------------*
| 1 7 38 | 249 289 6 | 389 2349 5 |
| 234568 2345689 3568 | 2459-1 *1289 7 |*1389 *12349 2489-1 |
| 2458 24589 58 | 12459 3 12458 | 6 7 12489 |
|----------------------------+----------------------------+----------------------------|
| 256 256 7 | 8 4 12 |#159 #12569 3 |
| 23468 23468 *1368 | 127 5 9 |*178 *1246 2468-1 |
| 9 2458 *158 | 127 6 3 |*1578 *1245 248-1 |
|----------------------------+----------------------------+----------------------------|
| 356 356 4 | 1569 7 15 | 2 8 169 |
| 7 3568 9 | 2456-1 *128 2458-1 |*135 *1356 6-1 |
| 568 1 2 | 3 89 58 | 4 569 7 |
*--------------------------------------------------------------------------------------*
|
To answer your question, there are a number of people in the Players' Forums who would probably be interested in your fish collection. |
|
Back to top |
|
|
| MCondron
| Joined: 17 Jul 2006 | Posts: 35 | : | Location: Houston, TX | Items |
|
Posted: Fri Dec 29, 2006 11:21 am Post subject: |
|
|
daj95376 wrote: | Interesting puzzle. Another perspective.
Code: | # your closest starting point:
# before finned X-Wing eliminations [r56c9]<>1
#
# Jellyfish c3578\r2568 -or-
# kraken cell [r4c7]=1 => [r56c9]<>1 and X-Wing c58\r28
# kraken cell [r4c8]=1 => [r56c9]<>1 and X-Wing c57\r28
|
|
Hmm...my solver should have spotted the finned jellyfish. Thanks for the feedback. I will investigate.
Quote: |
To answer your question, there are a number of people in the Players' Forums who would probably be interested in your fish collection. |
Yes, I was debating whether to post this here or there, but since it was part of an effort to build a test library for a solver, this seemed slightly more appropriate. This also has relevance in the "Setting Sudoku" forum. |
|
Back to top |
|
|
| MCondron
| Joined: 17 Jul 2006 | Posts: 35 | : | Location: Houston, TX | Items |
|
Posted: Fri Dec 29, 2006 11:45 am Post subject: |
|
|
I see why my solver didn't see the jellyfish.
This is how my solver proceeds:
Code: |
1....6..5
.........
....3.67.
..784...3
....59...
9...6....
..4.7.28.
7.9......
.123..4..
Step 1: Hidden single in box 5 digit 3
Step 2: Hidden single in col 6 digit 7
Step 3: Hidden single in row 1 digit 7
Step 4: Hidden single in row 9 digit 7
Step 5: Line-box interaction: Row 7 Box 7 Digit 3
Step 6: Finned X-Wing: Rows 49 Cols 18 Digit 6
Step 7: Finned Swordfish: Rows 149 Cols 578 Digit 9
Step 8: Sashimi Swordfish: Rows 347 Cols 469 Digit 1
Step 9: Franken Swordfish: C5C7C8 R2R8B6 Digit 1
|
The sashimi swordfish in step 8 clears the 1's from box 6 (col 9) that would have been the victims of the finned jellyfish.
My solver at the time this was generated ran in this order:
- Naked/hidden singles - quads; line-box interactions
- X-wing
- Swordfish
- Jellyfish
- Finned x-wing
- Sashimi x-wing
- Finned swordfish
- Sashimi swordfish
- Finned jellyfish
- Sashimi jellyfish
- Franken swordfish
Yet another example of how critically the order of solution methods influences our (or our software's) perception of a puzzle.
DAJ, your code seems to evaluate and report on all possibilities for each step, is this right? I might try doing something like that as well.
Mike |
|
Back to top |
|
|
| daj95376
| Joined: 05 Feb 2006 | Posts: 349 | : | | Items |
|
Posted: Fri Dec 29, 2006 5:24 pm Post subject: |
|
|
MCondron,
My post was for a Kraken Jellyfish -- the least desirable type of fish. Since your solver was able to crack the puzzle with finned, Sashimi, and Franken fish, then that's the preferred way to go. Nice catch! |
|
Back to top |
|
|
| MCondron
| Joined: 17 Jul 2006 | Posts: 35 | : | Location: Houston, TX | Items |
|
Posted: Fri Dec 29, 2006 5:45 pm Post subject: |
|
|
daj95376 wrote: | My post was for a Kraken Jellyfish |
Hmm. Looking at the pattern as a jellyfish, I would call those cells (R4C78) "fins", but then I'm still unclear about some details of the terminology that's used here. They seem to behave exactly as fins do. Is there a difference?
Quote: | Since your solver was able to crack the puzzle with finned, Sashimi, and Franken fish, then that's the preferred way to go. Nice catch! |
Thanks. I've definitely learned a lot from this forum and the player's forum. |
|
Back to top |
|
|
| daj95376
| Joined: 05 Feb 2006 | Posts: 349 | : | | Items |
|
Posted: Fri Dec 29, 2006 9:53 pm Post subject: |
|
|
MCondron wrote: | daj95376 wrote: | My post was for a Kraken Jellyfish |
Hmm. Looking at the pattern as a jellyfish, I would call those cells (R4C78) "fins", but then I'm still unclear about some details of the terminology that's used here. They seem to behave exactly as fins do. Is there a difference?
Quote: | Since your solver was able to crack the puzzle with finned, Sashimi, and Franken fish, then that's the preferred way to go. Nice catch! |
Thanks. I've definitely learned a lot from this forum and the player's forum. |
In the Players' Forums, there is a slew of information on fish and the various types. All of the extraneous cells are called 'fin' cells. How they relate to the basic fish pattern determines how they're categorized.
Also, Ruud is managing a Sudoku Wiki site called Sudopedia. It has examples and/or definitions for many fish types. There is a lot of great information there in general!!!
Note: My Kraken Jellyfish response to your puzzle does not meet the standard definition for kraken because I extended the scope of the kraken (fin) cells to include secondary effects from X-Wings. |
|
Back to top |
|
|
| Pat
| Joined: 06 Sep 2006 | Posts: 128 | : | | Items |
|
Posted: Sun Feb 25, 2007 11:51 am Post subject: re: Franken Swordfish example |
|
|
MCondron wrote: |
so far this is the only one that can be fully solved using only more basic patterns (X-wing, swordfish, and jellyfish, along with finned and sashimi variations)
[ 24 clues ]
Code: |
1 . . | . . 6 | . . 5
. . . | . . . | . . .
. . . | . 3 . | 6 7 .
-------+-------+------
. . 7 | 8 4 . | . . 3
. . . | . 5 9 | . . .
9 . . | . 6 . | . . .
-------+-------+------
. . 4 | . 7 . | 2 8 .
7 . 9 | . . . | . . .
. 1 2 | 3 . . | 4 . .
|
This is how my solver proceeds:
- Hidden single in box 5 digit 3
- Hidden single in col 6 digit 7
- Hidden single in row 1 digit 7
- Hidden single in row 9 digit 7
- Line-box interaction: Row 7 Box 7 Digit 3
- Finned X-Wing: Rows 49 Cols 18 Digit 6
- Finned Swordfish: Rows 149 Cols 578 Digit 9
- Sashimi Swordfish: Rows 347 Cols 469 Digit 1
- Franken Swordfish: C5C7C8 R2R8B6 Digit 1
|
hey MCondron, your puzzle is even better than you claim — after the 4 "hidden singles", steps 5-8 are not needed — the Franken Swordfish c578\r28b6 excludes 1 in r2c49 + r8c469 + b6c9 — the significant exclusion is r8c9 — then "singles" to the end —
this is the first and only example i've seen where a Franken solves the puzzle!!
|
|
Back to top |
|
|
|