The Ultimate FISH Guide

Advanced methods and approaches for solving Sudoku puzzles

Postby tarek » Sun Feb 10, 2008 5:41 pm

the idea of indroducing sashimi as a different adjective was not to introduce the "sashimi unfinned fish" from the backdoor. it obviously brought it up but that wasn't my intention.

my idea was simply a way of using the labelling system correctly.

the 4 combinations (of finned, unfinned, sashimi, non-sashimi) do include "sashimi unfinned". discussing that issue is a seperate matter.

i am glad that in the brief few posts previous to this, the sashiminess adjective has been in active use next to the Finning adjective. That is all that I want for now.

Mike,

technically speaking, "without the fins, doesn't mean that the fish had or needs to have fins to start with".

The issue regarding the "sashimi unfinned creature" seems to be:
Code: Select all
when you encounter degenerate patterns, how would you label them?
1. in their pre-degeneration staus without referring to their degenerativeness OR
2. in their pre-degeneration staus but referring to their degenerativeness OR
3. in their Post-degeneration staus ?


ronk (I think) uses point 3 & there is merit in what he is saying. I would probably take a more lenient approach & allow point 2.

tarek
User avatar
tarek
 
Posts: 2624
Joined: 05 January 2006

Postby daj95376 » Mon Feb 11, 2008 10:20 pm

I do not like adding a Sashiminess adjective to describe a fish. It seems to me that we're adding unnecessary complexity.

With the added adjective, everyone seems to agree that: 1) non-sashimi unfinned, 2) non-sashimi finned, or 3) sashimi finned are the only real possibilities.

However, these three alternatives can be abbreviated as per ronk's preference: 1) unfinned, 2) finned, or 3) sashimi.

This leaves a compact format of:

Code: Select all
Qualifier + Shape + Size

Qualifier = {unfinned, finned, sashimi}
Shape     = {basic, franken, mutant, kraken}
Size      = {1-Fish, X-Wing, Swordfish, Jellyfish, Starfish, Whale, Leviathan}

unfinned can be assumed if Qualifier is missing
basic    can be assumed if Shape     is missing

Thus, one fish may be described as Jellyfish, and another fish may be described as sashimi mutant Starfish.

Note: Since writing my GFF program, I've yet to encounter a Leviathan.
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

Postby tarek » Mon Feb 11, 2008 11:27 pm

daj95376 wrote:I do not like adding a Sashiminess adjective to describe a fish. It seems to me that we're adding unnecessary complexity.

With the added adjective, everyone seems to agree that: 1) non-sashimi unfinned, 2) non-sashimi finned, or 3) sashimi finned are the only real possibilities.

the "everyone" in your statement DOES include me. however, it is me who is pushing to seperate sashiminess from finning.

You may find it unnecessary, redundant & more complex. I see it as better description & less complex. I can't put unfinned, finned & sashimi in one class. that is EXACTLY what I'm against. i am after a system that hopefully would stand the test of time.

tarek
User avatar
tarek
 
Posts: 2624
Joined: 05 January 2006

Postby hobiwan » Tue Feb 12, 2008 10:37 am

tarek wrote:You may find it unnecessary, redundant & more complex. I see it as better description & less complex. I can't put unfinned, finned & sashimi in one class. that is EXACTLY what I'm against. i am after a system that hopefully would stand the test of time.

I agree with you. The real important difference is unfinned/finned. Sashimi adds additional but not so important (for me at least) information.
hobiwan
2012 Supporter
 
Posts: 321
Joined: 16 January 2008
Location: Klagenfurt

Postby daj95376 » Fri Feb 29, 2008 1:46 am

While tarek rewrites the head post to include Sashiminess and ronk's latest Exemplars ...

Code: Select all
 *-----------------------------------------------------------*
 | 7     8     19    | 5     136   136   | 2     4     139   |
 | 5     2     3     | 89    189   4     | 7     19    6     |
 | 19    6     4     | 39    2     7     | 139   8     5     |
 |-------------------+-------------------+-------------------|
 | 6     7     5     | 1     39    39    | 4     2     8     |
 | 4     39    2     | 6     7     8     | 5     13-9  139   |
 | 39    1     8     | 2     4     5     | 39    6     7     |
 |-------------------+-------------------+-------------------|
 | 1389  4     19    | 389   5     139   | 6     7     2     |
 | 2     39    7     | 4     1369  1369  | 8     5     19    |
 | 189   5     6     | 7     89    2     | 139   139   4     |
 *-----------------------------------------------------------*

Here is a Sashimi finned mutant Jellyfish that seems legal for the elimination. (Yes, there are other Jellyfish for the elimination, but I want to discuss this one.)

Code: Select all
 4-Fish r6c29b1\r18c18           Sfm 121\220  <> 9  [r5c8]
 +-----------------------------------+
 |  *  /  X  |  *  *  *  |  *  *  X  |
 |  X  /  /  |  .  .  .  |  .  *  /  |
 |  X  /  /  |  .  .  .  |  .  *  /  |
 |-----------+-----------+-----------|
 |  *  /  .  |  .  .  .  |  .  *  /  |
 |  *  #  .  |  .  .  .  |  . **  #  |
 |  X  /  /  |  /  /  /  |  #  X  /  |
 |-----------+-----------+-----------|
 |  *  /  .  |  .  .  .  |  .  *  /  |
 |  *  X  *  |  *  *  *  |  *  *  X  |
 |  *  /  .  |  .  .  .  |  .  *  /  |
 +-----------------------------------+

When the candidate grid is examined, all of the candidates in [c8] would be eliminated if it weren't for the fin cells. This is because none of the Base sectors actually intersect an occupied cell in Cover sector [c8].

Code: Select all
 +-----------------------------------+
 |  .  .  9  |  .  .  .  |  .  .  9  |
 |  .  .  .  |  9  9  .  |  . -9  .  |
 |  9  .  .  |  9  .  .  |  9  .  .  |
 |-----------+-----------+-----------|
 |  .  .  .  |  .  9  9  |  .  .  .  |
 |  . #9  .  |  .  .  .  |  . -9 #9  |
 |  9  .  .  |  .  .  .  | #9  .  .  |
 |-----------+-----------+-----------|
 |  9  .  9  |  9  .  9  |  .  .  .  |
 |  .  9  .  |  .  9  9  |  .  .  9  |
 |  9  .  .  |  .  9  .  |  9 -9  .  |
 +-----------------------------------+

Comments? (Have I presented a scenario similar to this already? The gray matter is slipping fast!)
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

Postby ronk » Fri Feb 29, 2008 2:35 am

daj95376 wrote:Here is a Sashimi finned mutant Jellyfish that seems legal for the elimination.
Code: Select all
 4-Fish r6c29b1\r18c18           Sfm 121\220  <> 9  [r5c8]
 +-----------------------------------+
 |  *  /  X  |  *  *  *  |  *  *  X  |
 |  X  /  /  |  .  .  .  |  .  *  /  |
 |  X  /  /  |  .  .  .  |  .  *  /  |
 |-----------+-----------+-----------|
 |  *  /  .  |  .  .  .  |  .  *  /  |
 |  *  #  .  |  .  .  .  |  . **  #  |
 |  X  /  /  |  /  /  /  |  #  X  /  |
 |-----------+-----------+-----------|
 |  *  /  .  |  .  .  .  |  .  *  /  |
 |  *  X  *  |  *  *  *  |  *  *  X  |
 |  *  /  .  |  .  .  .  |  .  *  /  |
 +-----------------------------------+

If, as in the pencilmarks, there is not a candidate at r6c8, then c8 is not a valid cover unit (sector).
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

underlying (unfinned) fish would exclude all in unit

Postby Pat » Mon Mar 03, 2008 2:15 pm

daj95376 wrote:
Code: Select all
 *-----------------------------------------------------------*
 | 7     8     19    | 5     136   136   | 2     4     139   |
 | 5     2     3     | 89    189   4     | 7     19    6     |
 | 19    6     4     | 39    2     7     | 139   8     5     |
 |-------------------+-------------------+-------------------|
 | 6     7     5     | 1     39    39    | 4     2     8     |
 | 4     39    2     | 6     7     8     | 5     13-9  139   |
 | 39    1     8     | 2     4     5     | 39    6     7     |
 |-------------------+-------------------+-------------------|
 | 1389  4     19    | 389   5     139   | 6     7     2     |
 | 2     39    7     | 4     1369  1369  | 8     5     19    |
 | 189   5     6     | 7     89    2     | 139   139   4     |
 *-----------------------------------------------------------*



Here is a Sashimi finned mutant Jellyfish that seems legal for the elimination.
( Yes, there are other Jellyfish for the elimination, but I want to discuss this one. )


4-Fish 121\220 r6c29b1\r18c18 ==> r5c8 <> 9

Code: Select all
 +-----------------------------------+
 |  *  /  X  |  *  *  *  |  *  *  X  |
 |  X  /  /  |  .  .  .  |  .  *  /  |
 |  X  /  /  |  .  .  .  |  .  *  /  |
 |-----------+-----------+-----------|
 |  *  /  .  |  .  .  .  |  .  *  /  |
 |  *  #  .  |  .  .  .  |  . **  #  |
 |  X  /  /  |  /  /  /  |  #  X  /  |
 |-----------+-----------+-----------|
 |  *  /  .  |  .  .  .  |  .  *  /  |
 |  *  X  *  |  *  *  *  |  *  *  X  |
 |  *  /  .  |  .  .  .  |  .  *  /  |
 +-----------------------------------+



When the candidate grid is examined, all of the candidates in c8 would be eliminated if it weren't for the fin-cells. This is because none of the base-sectors intersect an occupied cell in cover-sector c8.

Code: Select all
 +-----------------------------------+
 |  .  .  9  |  .  .  .  |  .  .  9  |
 |  .  .  .  |  9  9  .  |  . -9  .  |
 |  9  .  .  |  9  .  .  |  9  .  .  |
 |-----------+-----------+-----------|
 |  .  .  .  |  .  9  9  |  .  .  .  |
 |  . #9  .  |  .  .  .  |  . -9 #9  |
 |  9  .  .  |  .  .  .  | #9  .  .  |
 |-----------+-----------+-----------|
 |  9  .  9  |  9  .  9  |  .  .  .  |
 |  .  9  .  |  .  9  9  |  .  .  9  |
 |  9  .  .  |  .  9  .  |  9 -9  .  |
 +-----------------------------------+



Comments?
( Have I presented a scenario similar to this already? )




yes you have -- in the other Forum --

daj95376 (2007.Dec.21) wrote:Below is a finned fish -- sans 'X' -- for an elimination.

5-Fish r6c3567\r578c4b6 ==> r2c4 <> X

Code: Select all
 +-----------------------------------------------+
 |   .   .   /   |   *   /   /   |   /   .   .   |
 |   .   .   #   |   $   /   /   |   #   .   .   |
 |   .   .   /   |   *   #   /   |   /   .   .   |
 |---------------+---------------+---------------|
 |   .   .   /   |   *   /   /   |   .   *   *   |
 |   *   *   .   |   *   .   .   |   *   *   *   |
 |   /   /   /   |   .   /   /   |   /   .   .   |
 |---------------+---------------+---------------|
 |   *   *   .   |   *   .   .   |   .   *   *   |
 |   *   *   .   |   *   .   .   |   .   *   *   |
 |   .   .   /   |   *   /   /   |   /   .   .   |
 +-----------------------------------------------+



Now, take a close look at the underlying, unfinned fish. It wants to eliminate all non-empty cells in b2.

Code: Select all
 +-----------------------------------------------+
 |   .   .   /   |   *   /   /   |   /   .   .   |
 |   .   .   /   |   *   /   /   |   /   .   .   |
 |   .   .   /   |   *   /   /   |   /   .   .   |
 |---------------+---------------+---------------|
 |   .   .   /   |   *   /   /   |   .   *   *   |
 |   *   *   .   |   *   .   .   |   *   *   *   |
 |   /   /   /   |   .   /   /   |   /   .   .   |
 |---------------+---------------+---------------|
 |   *   *   .   |   *   .   .   |   .   *   *   |
 |   *   *   .   |   *   .   .   |   .   *   *   |
 |   .   .   /   |   *   /   /   |   /   .   .   |
 +-----------------------------------------------+


User avatar
Pat
 
Posts: 3448
Joined: 18 July 2005

Re: underlying (unfinned) fish would exclude all in unit

Postby ronk » Mon Mar 03, 2008 2:46 pm

Pat wrote:
daj95376 wrote:Have I presented a scenario similar to this already?

yes you have -- in the other Forum --

daj95376 (2007.Dec.21)"

Not the same scenario IMO, since the former was about the validity of a sashimi should all its fin cells be false.

(BTW it's usually not necessary to quote an entire post.)
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Postby Pat » Mon Mar 03, 2008 2:55 pm

ronk wrote:Not the same scenario IMO,
since the former was about the validity of a sashimi should all its fin-cells be false.


as far as i can tell,
both are about the underlying (unfinned) fish
( i.e. what happens when the digit is not in any of the fin-cells )
User avatar
Pat
 
Posts: 3448
Joined: 18 July 2005

Postby ronk » Mon Mar 03, 2008 5:06 pm

Pat wrote:
ronk wrote:Not the same scenario IMO,
since the former was about the validity of a sashimi should all its fin-cells be false.

as far as i can tell,
both are about the underlying (unfinned) fish
( i.e. what happens when the digit is not in any of the fin-cells )

It might seem that way. However, in the later case where ...
daj95376 wrote:
Code: Select all
 4-Fish r6c29b1\r18c18           Sfm 121\220  <> 9  [r5c8]
 +-----------------------------------+
 |  *  /  X  |  *  *  *  |  *  *  X  |
 |  X  /  /  |  .  .  .  |  .  *  /  |
 |  X  /  /  |  .  .  .  |  .  *  /  |
 |-----------+-----------+-----------|
 |  *  /  .  |  .  .  .  |  .  *  /  |
 |  *  #  .  |  .  .  .  |  . **  #  |
 |  X  /  /  |  /  /  /  |  #  X  /  |
 |-----------+-----------+-----------|
 |  *  /  .  |  .  .  .  |  .  *  /  |
 |  *  X  *  |  *  *  *  |  *  *  X  |
 |  *  /  .  |  .  .  .  |  .  *  /  |
 +-----------------------------------+

... the "exemplar" is correct, but only as a general case. Unfortunately, it is based on these "pencilmarks" ...
daj95376 wrote:
Code: Select all
+-----------------------------------+
 |  .  .  9  |  .  .  .  |  .  .  9  |
 |  .  .  .  |  9  9  .  |  . -9  .  |
 |  9  .  .  |  9  .  .  |  9  .  .  |
 |-----------+-----------+-----------|
 |  .  .  .  |  .  9  9  |  .  .  .  |
 |  . #9  .  |  .  .  .  |  . -9 #9  |
 |  9  .  .  |  .  .  .  | #9  .  .  |
 |-----------+-----------+-----------|
 |  9  .  9  |  9  .  9  |  .  .  .  |
 |  .  9  .  |  .  9  9  |  .  .  9  |
 |  9  .  .  |  .  9  .  |  9 -9  .  |
 +-----------------------------------+

... which does not have a candidate in r6c8. IOW cover unit c8 is not covering any candidate of any base unit, which means c8 is not a valid cover unit. Therefore the exemplar is invalid for these pencilmarks, and what might happen if all the fin cells were false is irrelevant.

The same might be true for the other scenario, but we can't tell because no pencilmarks were provided.
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Postby daj95376 » Tue Mar 04, 2008 1:44 am

As I've learned the very hard way, there are two important tests that a finned/unfinned fish must pass before it can be used.

Test #1: The (underlying unfinned) fish must be valid in the first place. This involves checking every unit in its exemplar to verify that at least one cell is characterized as '.' or 'X'. If not, then the fish can not be used. (The 5-Fish Pat cites from the Programmers Forum.)

Test #2: Every cover sector in the candidate grid must contain at least one occupied '.' or 'X' cell from the exemplar for the (underlying unfinned) fish used. If not, then the fish can not be used. (The 4-Fish I listed above.)
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

Postby ronk » Tue Mar 04, 2008 3:14 am

daj95376 wrote:Test #1: The (underlying unfinned) fish must be valid in the first place. This involves checking every unit in its exemplar to verify that at least one cell is characterized as '.' or 'X'. If not, then the fish can not be used. (The 5-Fish Pat cites from the Programmers Forum.)

I don't believe this test is necessary. If the "underlying unfinned" fish is "invalid", it merely means that one of the fin cells must ultimately be true.

Have you ever seen an elimination based on such a fish to be invalid?
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Postby daj95376 » Tue Mar 04, 2008 7:40 am

ronk wrote:Have you ever seen an elimination based on such a fish to be invalid?

I use Templates to reduce my search space to known eliminations, so there isn't any chance that an elimination would be invalid. Be it Test #1 or Test #2 that fails, the finned fish does correctly identify the elimination cell in all of my results. This only leaves a variant of your question.

Q: Even if Test #1 and/or Test #2 fail for a finned fish, is the elimination ever invalid?

I suspect the answer is no.

Note: Originally, I only performed Test #2. That lead to my post in the Programmers Forum. I then (mistakenly) dropped Test #2 and went with Test #1 after discussions there. This eventually let to my most recent post and the discover of my mistake in dropping Test #2.

Addendum: ===== ===== ===== =====

I tested 1,515 of my puzzles. Here are the results for Franken/mutant fish found in their solutions.
(note: multiple fish were reported possible for most Franken/mutant fish eliminations.)

Code: Select all
T_none:   Results when      Test #1 and Test #2 are not applied
T_1only:  Results when only Test #1                     applied
T_2only:  Results when only             Test #2         applied
T_12:     Results when both Test #1 and Test #2         applied

Code: Select all
T_none   = 4,843 fish = T_12 + 1,322
T_1only  = 3,895 fish = T_12 +   374
T_2only  = 3,531 fish = T_12 +    10
T_12     = 3,521 fish

Because T_none is significantly larger than ( T_1only + T_2only ), I'm guessing that various permutations of the only failed fish led to the T_none results.

Note: I noticed one case where 5 of 6 possible fish for an elimination disappeared between T_none and T_12.
Last edited by daj95376 on Wed Mar 05, 2008 1:41 pm, edited 2 times in total.
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

Postby ronk » Wed Mar 05, 2008 1:50 pm

daj95376 wrote:I tested 1,515 of my puzzles. Here are the results for Franken/mutant fish found in their solutions.
(note: multiple fish were reported possible for most Franken/mutant fish eliminations.)
Code: Select all
T_none   = 4,843 fish = T_12 + 1,322
T_1only  = 3,895 fish = T_12 +   374
T_2only  = 3,531 fisl = T_12 +    10
T_12     = 3,521 fish

Because T_none is significantly larger than ( T_1only + T_2only ), I'm guessing that various permutations of the latter failed fish led to the T_none results.

First of all, I'm surprised by the small difference between (the counts for) T_2only and T12.

However, the approximate 3:2 ratio for T_none:T_2only seems to make sense. I think you were counting 3 fish instead of 2 for every finned pattern**, whether 1-finned or 2-finned. The difference between T_none and T_2only*2/3 implies there were about 300 unfinned fish. Did you happen to count those also?

** For example, using your recent example:
Code: Select all
 4-Fish r6c29b1\r18c18           Sfm 121\220  <> 9  [r5c8]
 +-----------------------------------+
 |  *  /  X  |  *  *  *  |  *  *  X  |
 |  X  /  /  |  .  .  .  |  .  *  /  |
 |  X  /  /  |  .  .  .  |  .  *  /  |
 |-----------+-----------+-----------|
 |  *  /  .  |  .  .  .  |  .  *  /  |
 |  *  #  .  |  .  .  .  |  . **  #  |
 |  X  /  /  |  /  /  /  |  #  /  /  |
 |-----------+-----------+-----------|
 |  *  /  .  |  .  .  .  |  .  *  /  |
 |  *  X  *  |  *  *  *  |  *  *  X  |
 |  *  /  .  |  .  .  .  |  .  *  /  |
 +-----------------------------------+

I think T_none would count:
r6c29b1\r18c18 + r5b6
r6c29b1\r158c1 + c8b6
r6c29b1\r18c1b6 + r5c8

c8 is superfluous for the last two, and makes the first invalid. (Note r6c8 is empty per the pencilmarks.)
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Postby daj95376 » Wed Mar 05, 2008 6:43 pm

To me, the results for T_1only and T_2only seem to imply that the fish getting past Test #1 are often caught by Test #2, but the converse is not true. Test #2 is the more dominant filter!

The 3:2 ratio doesn't hold up as a generality for most eliminations. Sometimes a high percentage of the possible fish are invalid, and other times it is a very small percentage. However, I will agree that there are a number of cases where an Nx(N-1) sub-fish could only be combined legitimately with two of the three possible sectors passing through the elimination cell -- to form two of three possible NxN fish.

There are only 54 unfinned Franken/mutant fish in my results. They appear in all of the T_ files; i.e., they are all legitimate fish.

r6c29b1\r18c18 appears in T_none and T_1only, which implies that it's caught by Test #2. r6c29b1\r158c1 and r6c29b1\r18c1b6 appear in T_12, which implies that they're legitimate fish.
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

PreviousNext

Return to Advanced solving techniques