dailysudoku.com Forum Index dailysudoku.com
Discussion of Daily Sudoku puzzles
 
 FAQFAQ   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   RegisterRegister 
 ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

Locating swordfish patterns

 
Post new topic   Reply to topic    dailysudoku.com Forum Index -> Solving techniques, and terminology
View previous topic :: View next topic  
Author Message
kuskey



Joined: 10 Dec 2008
Posts: 127
Location: Pembroke, NH

PostPosted: Fri Aug 19, 2011 8:11 pm    Post subject: Locating swordfish patterns Reply with quote

I have never been very successfull locating swordfish. I think I have read most commonly available threads on this subject. For some reason, I have not been able to reduce the definitions to a visual search algorithm. I visit Brainbashers frequently and many of their superhard puzzles contain swordfish, forcing me to peak at the step by step solution. I would appreciate any help you folks can offer.

kuskey
Back to top
View user's profile Send private message
daj95376



Joined: 23 Aug 2008
Posts: 3855

PostPosted: Fri Aug 19, 2011 9:28 pm    Post subject: Reply with quote

Besides Size, fish are also classified by their Shape.

Code:
Size of Fish:

1. 1-Fish (aka Cyclopsfish): All elements are in 1 x 1 sector
2. X-Wing                  : All elements are in 2 x 2 sectors
3. Swordfish               : All elements are in 3 x 3 sectors
4. Jellyfish               : All elements are in 4 x 4 sectors
5. Squirmbag (aka Starfish): All elements are in 5 x 5 sectors
6. Whale                   : All elements are in 6 x 6 sectors
7. Leviathan               : All elements are in 7 x 7 sectors

Code:
Shape of Fish:

1.  basic         : N x N line. (no fin)
2a. Finned        : N x N lines (with fin, limited to "basic fish with extra candidates")
2b. Sashimi       : N x N lines (with fin, not limited to 2a)
3.  Franken       : N (rows+boxes) x N (columns+boxes) (no fins)   -or- vv for rows/columns
4.  Finned Franken: N (rows+boxes) x N (columns+boxes) (with fins) -or- vv for rows/columns
5.  Mutant        : N x N sectors (no fins)
6.  Finned Mutant : N x N sectors (with fins)
7.  Kraken        : Any fish (1-6) that requires information from outside the pattern
___________________________________________________________________________________________

The difficulty in finding a fish is often directly related to the shape.

What shape of Swordfish are you hoping to locate?
Back to top
View user's profile Send private message
kuskey



Joined: 10 Dec 2008
Posts: 127
Location: Pembroke, NH

PostPosted: Fri Aug 19, 2011 10:13 pm    Post subject: locating swordfish patterns Reply with quote

That is a good question because I never associated size and shape as a pair. Initially its basic swordfish growing to Franken, and others perhaps. I have come across the terms you presented but never gave them much consideration. Some examples would help me if you would be so kind and exactly how would you logically search rows and columns to identify the pattern. The one I have not heard of is the 1-fish.
Back to top
View user's profile Send private message
daj95376



Joined: 23 Aug 2008
Posts: 3855

PostPosted: Sat Aug 20, 2011 12:15 am    Post subject: Reply with quote

The 1-fish is there for "completeness". It's equivalent to a Hidden Single.

Describing how to find a "basic" Swordfish is relatively easy. However, they are, IMO, less common than "finned" and "Sashimi" Swordfish, which are more difficult to find. The remaining shapes get successively more difficult to find, IMO.

[Withdrawn: I'm not really prepared to provide examples.]


Last edited by daj95376 on Sat Aug 20, 2011 1:50 am; edited 1 time in total
Back to top
View user's profile Send private message
Marty R.



Joined: 12 Feb 2006
Posts: 5161
Location: Rochester, NY, USA

PostPosted: Sat Aug 20, 2011 12:21 am    Post subject: Reply with quote

This is probably not the kind of input you're looking for, but I don't look; too much work for too little reward.
Back to top
View user's profile Send private message
keith



Joined: 19 Sep 2005
Posts: 3174
Location: near Detroit, Michigan, USA

PostPosted: Sat Aug 20, 2011 9:20 am    Post subject: Reply with quote

kuskey,

A swordfish is three strong links in rows (columns) that line up in some way in columns (rows).

I'm with Marty - they are not worth looking for. If you understand X-wings and kites / skyscrapers / turbot fish, swordfish are unnecessary. (I am sure someone has a counterexample to that statement.)

Please post those BB puzzles that "require" a swordfish in the "Other Puzzles" forum. We'll see what other solution paths there are.

As for the "almost" swordfish, finned, kraken, sashimi, franken, ... they are basically useless (not findable) for those of us who solve puzzles manually, with pencil and paper.

Keith
Back to top
View user's profile Send private message
kuskey



Joined: 10 Dec 2008
Posts: 127
Location: Pembroke, NH

PostPosted: Sun Aug 21, 2011 3:51 pm    Post subject: locating swordfish patterns Reply with quote

Thanks for the help. I agree with you all that trolling for fish is probably more work than its worth. Even X-wings can require a lot of time when the solved cells are minimal.

I will pass along any grids advertized to require swordfish so you can look at them. So far other techniques have solved the grids.
Back to top
View user's profile Send private message
daj95376



Joined: 23 Aug 2008
Posts: 3855

PostPosted: Sun Aug 21, 2011 10:01 pm    Post subject: Reply with quote

In another forum, Tarek is accumulating what he terms, "A Pure Swordfish Collection". (basic) Swordfish are the only advanced steps needed in his puzzles. They may not be easy to circumvent.

Tarek Puzzle #2:

Code:
 +-----------------------+
 | 7 . . | . 4 . | . . . |
 | . 5 . | . . 1 | 3 . . |
 | . . 6 | 2 . . | . 7 . |
 |-------+-------+-------|
 | . . 7 | . . . | . 2 . |
 | 9 . . | . . . | . . 8 |
 | . 3 . | . . . | 5 . . |
 |-------+-------+-------|
 | . 6 . | . . 3 | 1 . . |
 | . . 2 | 7 . . | . 6 . |
 | . . . | . 5 . | . . 4 |
 +-----------------------+

Play this puzzle online at the Daily Sudoku site
Back to top
View user's profile Send private message
Marty R.



Joined: 12 Feb 2006
Posts: 5161
Location: Rochester, NY, USA

PostPosted: Sun Aug 21, 2011 11:45 pm    Post subject: Reply with quote

I tried monkeying around with this puzzle in Draw/Play and got this far with basics.

Code:

+----------------+------------------+--------------+
| 7    2    1389 | 3589   4    589  | 6   1589 159 |
| 48   5    489  | 6      7    1    | 3   489  2   |
| 1348 1489 6    | 2      389  589  | 489 7    159 |
+----------------+------------------+--------------+
| 6    148  7    | 134589 1389 4589 | 49  2    139 |
| 9    14   5    | 134    6    2    | 7   134  8   |
| 2    3    148  | 1489   189  7    | 5   149  6   |
+----------------+------------------+--------------+
| 458  6    489  | 489    2    3    | 1   589  7   |
| 35   1489 2    | 7      189  489  | 89  6    35  |
| 138  7    1389 | 189    5    6    | 2   389  4   |
+----------------+------------------+--------------+

Play this puzzle online at the Daily Sudoku site

Quote:
They may not be easy to circumvent.

I believe this so I'm not going to continue on pencil and paper for the reason I stated in my earlier post.
Back to top
View user's profile Send private message
kuskey



Joined: 10 Dec 2008
Posts: 127
Location: Pembroke, NH

PostPosted: Mon Aug 22, 2011 1:55 am    Post subject: swordfish Reply with quote

Marty, I agree with your after-basics code, which I found to be quite challenging. Will try to get the code into some type of solver that facilitates searching for fish one number at a time. Any suggestions here?
Back to top
View user's profile Send private message
keith



Joined: 19 Sep 2005
Posts: 3174
Location: near Detroit, Michigan, USA

PostPosted: Mon Aug 22, 2011 3:39 am    Post subject: Reply with quote

The Swordfish is on 9 in R279. It makes eliminations in C348.

To be quite honest, I would never recognize it, and I am not bent on learning how.

Then, there is a similar swordfish on 8 that solves the puzzle.

Sudoku Susser is very good for "fishy patterns".

Keith
Back to top
View user's profile Send private message
daj95376



Joined: 23 Aug 2008
Posts: 3855

PostPosted: Mon Aug 22, 2011 6:11 am    Post subject: Re: locating swordfish patterns Reply with quote

kuskey wrote:
Thanks for the help. I agree with you all that trolling for fish is probably more work than its worth. Even X-wings can require a lot of time when the solved cells are minimal.

I will pass along any grids advertized to require swordfish so you can look at them. So far other techniques have solved the grids.

Sometimes, hunting for a Swordfish may be preferred. Here's the most reasonable way that I found to get around the Swordfish in Tarek's puzzle.

Code:
 after basics
 +--------------------------------------------------------------------------------+
 |  7       2       1389    |  3589    4       589     |  6       1589    159     |
 |  48      5       489     |  6       7       1       |  3       489     2       |
 |  1348    1489    6       |  2       389     589     |  489     7       159     |
 |--------------------------+--------------------------+--------------------------|
 |  6       148     7       |  134589  1389    4589    |  49      2       139     |
 |  9       14      5       |  134     6       2       |  7       134     8       |
 |  2       3       148     |  1489    189     7       |  5       149     6       |
 |--------------------------+--------------------------+--------------------------|
 |  458     6       489     |  489     2       3       |  1       589     7       |
 |  35      1489    2       |  7       189     489     |  89      6       35      |
 |  138     7       1389    |  189     5       6       |  2       389     4       |
 +--------------------------------------------------------------------------------+
 # 89 eliminations remain

S-Wing:  (4)r2c8 = r3c7  - (4=9)r4c7 - r8c7 = (9)r79c8  =>  r2c8<>9
S-Wing:  (4)r2c8 = r56c8 - (4=9)r4c7 - r8c7 = (9)r79c8  =>  r2c8<>9
Back to top
View user's profile Send private message
Marty R.



Joined: 12 Feb 2006
Posts: 5161
Location: Rochester, NY, USA

PostPosted: Mon Aug 22, 2011 3:34 pm    Post subject: Reply with quote

What is the definition of an S-Wing? It looks like you used the same candidate to follow two different paths that reached the same conclusion.
Back to top
View user's profile Send private message
daj95376



Joined: 23 Aug 2008
Posts: 3855

PostPosted: Mon Aug 22, 2011 5:38 pm    Post subject: Reply with quote

Marty R. wrote:
What is the definition of an S-Wing? It looks like you used the same candidate to follow two different paths that reached the same conclusion.

There are two, almost identical, S-Wings for the same elimination. I listed both.

There are several common patterns that involve three strong links using only two values -- W-Wing, M-Wing, S-Wing, and L2-Wing. It adds content to an AIC when it's identified as following one of these patterns.

I believe there is a thread in the Techniques forum showing S-Wing pattern configurations.
Back to top
View user's profile Send private message
Pat



Joined: 23 Feb 2010
Posts: 165

PostPosted: Thu Aug 25, 2011 3:51 pm    Post subject: Reply with quote

daj95376 wrote:
kuskey wrote:
The one I have not heard of is the 1-fish.

The 1-fish is there for "completeness".
It's equivalent to a Hidden Single.

a fish of order 1 is any of these 3 --
  • box\row (or box\column), often called "pointing".
  • row\box (or column\box), often called "claiming".
  • row\column (or column\row), usually called "hidden single". (here tarek would use the term Mutant.)

daj95376 wrote:

tarek is accumulating "A Pure Swordfish Collection".

(basic) Swordfish are the only advanced steps needed in his puzzles. They may not be easy to circumvent.


here's the link --

daj95376 wrote:
"basic" Swordfish---are, IMO, less common than "finned" and "Sashimi" Swordfish

yes

every "finned" fish is more common than the corresponding un-finned fish

(the un-finned fish requires the absence of the digit from all cells of the pattern; whenever one of these cells fails this requirement, we have the potential for a "finned" fish)
Back to top
View user's profile Send private message
Pat



Joined: 23 Feb 2010
Posts: 165

PostPosted: Thu Aug 25, 2011 4:00 pm    Post subject: Reply with quote

kuskey wrote:
I agree with you all that trolling for fish is probably more work than its worth. Even X-wings can require a lot of time when the solved cells are minimal.

(off-topic)
is this puzzle too easy? (does not need Swordfish)

Pat wrote:


    ...1..4.....8.3.1.....26..968.....4...1...5...7.....867..36.....1.5.9.....2..7...

Code:


 . . . | 1 . . | 4 . .
 . . . | 8 . 3 | . 1 .
 . . . | . 2 6 | . . 9
-------+-------+------
 6 8 . | . . . | . 4 .
 . . 1 | . . . | 5 . .
 . 7 . | . . . | . 8 6
-------+-------+------
 7 . . | 3 6 . | . . .
 . 1 . | 5 . 9 | . . .
 . . 2 | . . 7 | . . .


Back to top
View user's profile Send private message
Pat



Joined: 23 Feb 2010
Posts: 165

PostPosted: Fri Aug 26, 2011 8:17 am    Post subject: Reply with quote

kuskey wrote:
I have never been very successfull locating swordfish

how do you find "hidden" trios?
    it's exactly the same problem

    jaap (2005.Jul.19) wrote:
    my solver treats things such as 'the possible placements of digit 7 in column 4'
    the same way as 'the possible values of cell (3,5)'.

    This means that matched pairs is treated the same as X-wings.
    For example, an X-wings in my solver gives the output:
    Code:
     Must have number 9 at (1,5) (and number 9 at (9,8))
     or number 9 at (9,5) (and number 9 at (1,8))
     reducing options for digit 9 in column 5
     Must have number 9 at (1,8) (and number 9 at (9,5))
     or number 9 at (9,8) (and number 9 at (1,5))
     reducing options for digit 9 in column 8
    and a matched pair gives the output:
    Code:
     Must have number 2 at (1,2) (and number 3 at (1,9))
     or number 2 at (1,9) (and number 3 at (1,2))
     reducing options for digit 2 in row 1
     Must have number 3 at (1,2) (and number 2 at (1,9))
     or number 3 at (1,9) (and number 2 at (1,2))
     reducing options for digit 3 in row 1


    In the same way (if I had implemented them)
    a matched triplet would be equivalent to Swordfish,
    a matched quadruplet equivalent to Jellyfish,
    etc.
Back to top
View user's profile Send private message
ffred



Joined: 29 Oct 2012
Posts: 19
Location: Kent, Egland

PostPosted: Sun Nov 04, 2012 10:18 pm    Post subject: Reply with quote

Marty & Keith, I'm glad I'm not the only person who dislikes complex swordfish patterns - I thought I was a bit wimpish. Soon after the sudoku craze started (04, 05?) I was given a book of puzzles but it was so riddled with swordfish (which I couldn't see without writing in candidates) that I just gave up on sudokus.

Still, any port in a storm as they say, and I do look for swordfish if all else fails. (But not variants, which I've never found.) I'm about to post a puzzle that I just couldn't unlock without a swordfish.
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic    dailysudoku.com Forum Index -> Solving techniques, and terminology All times are GMT
Page 1 of 1

 
Jump to:  
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