exocet pattern in hardest puzzles

Advanced methods and approaches for solving Sudoku puzzles

Re: exocet pattern in hardest puzzles

Postby ronk » Sat Jul 28, 2012 12:12 pm

David P Bird wrote:I don't see any other possibility that having a candidate locked in the object pair to form an AHS
Look at ronk's extreme example < here >. There is an Almost Hidden Set (578+x)r2358c7. The object cell pair r34c7 must contain a base digit (1234) and a digit locked in the hidden set (578) so (69) can be eliminated from these cells.

With my illustration as it stands, I have to agree with champagne. It should be OK after removing base candidates in the added AAHS cells (r58c5 and r5c8 of the exemplar). This will at least guarantee a base candidate exists in each "object pair."

I don't recall examples for either of the two following illustrations. Examples would be appreciated. [edit: daj95376 and champagne, that means each of you for your own illustration.]

daj95376 wrote:Note: there is a rare scenario that DPB's scenarios don't cover. (the distinction between QExocet and JExocet.)

Code: Select all
 "Q" is forced to contain the candidate
 *-------*-------*-------*
 | B B - | - - - | - - - |
 | - - - | Q / / | R . . |
 | - - - | Q / / | R . . |
 *-------*-------*-------*
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 *-------*-------*-------*
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 *-------*-------*-------*

champagne wrote: ... a last tentative contribution to your jexocet definition.
Code: Select all
*-------*-------*-------*
 | B B - | - - - | - - - |
 | - - - | Q . . | R . . |
 | - - - | Q . . | R . . |
 *-------*-------*-------*
 | . . 0 | . . . | . . . |
 | . . . | 0 . . | 0 . . |
 | . . . | . . . |. . . |
 *-------*-------*-------*
 | . . . | . . . | . . . |
 | . . 0 |. . . | . . . |
 | . . 0 | . . . | . . . |
 *-------*-------*-------*

in such a situation, the jexocet is established.

champagne wrote:
David P Bird wrote:
Code: Select all
                                    v           v                         v   
r4   . . \ | \ . . | \ . .      . . O | \ . . | O . .     . . \ | \ . . | O . .   
r5   . . O | O . . | O . . <    . . O | \ . . | O . .     . . O | O . . | O . . <   
r6   . . \ | \ . . | \ . .      . . O | \ . . | O . .     . . \ | \ . . | O . .   
r7   . . \ | \ . . | \ . .      . . O | \ . . | O . .     . . \ | \ . . | O . .   
r8   . . O | O . . | O . . <    . . O | \ . . | O . .     . . \ | \ . . | O . .   
r9   . . \ | \ . . | \ . .      . . O | \ . . | O . .     . . \ | \ . . | O . .
      2 Parallel Lines(I)        2 Parallel Lines(II)      2 Orthogonal Lines
         rows 5 & 8                 columns 3 & 7             row 5 & column 7


I have difficulties with these diagrams.
First one is ok,
In the second and third one, I can't see the possibility to have so many occurrences of the digit.
I don't see clearly the case where the cross line in stack 1 is empty and Q;R is forced in stacks 2 and 3.

Good luck with that! I brought up the same point in mid-April and it met the fate of a proverbial lead balloon. Perhaps illustrations with a separate "don't care" symbol, say "x", would have flown better.

Code: Select all
                                    v           v                         v   
r4   . . \ | \ . . | \ . .      . . x | \ . . | x . .     . . \ | \ . . | x . .   
r5   . . O | O . . | O . . <    . . x | \ . . | x . .     . . O | O . . | x . . <   
r6   . . \ | \ . . | \ . .      . . x | \ . . | x . .     . . \ | \ . . | x . .   
r7   . . \ | \ . . | \ . .      . . x | \ . . | x . .     . . \ | \ . . | x . .   
r8   . . O | O . . | O . . <    . . x | \ . . | x . .     . . \ | \ . . | x . .   
r9   . . \ | \ . . | \ . .      . . x | \ . . | x . .     . . \ | \ . . | x . .
 'x' <-> presence or absence of candidate 'O' is irrelevant
Last edited by ronk on Sat Jul 28, 2012 3:00 pm, edited 1 time in total.
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Re: exocet pattern in hardest puzzles

Postby champagne » Sat Jul 28, 2012 1:38 pm

ronk wrote:I don't recall examples for either of the two following illustrations. Examples would be appreciated.


XWing induced pattern is discussed head of page 9 in that thread.

I have examples of the second pattern, but with quads in a box.
Dannny shows a pattern with 5 free cells.

In my case, this does not come as Jexocet, but as simple Exocet and I intended to adjust the code to get free of that exocet having no added value.
I have to look in the files to find an example

EDIT one example

..3..6...4.67...3..7....5...6...597..3...7.4.9............2......8.....1.4...3.5.;6519;elev;3594;6;7;

Code: Select all
A     B     C   |D      E     F     |G     H     I     
1258  12589 3   |124589 14589 6     |12478 1289  24789 
4     12589 6   |7      1589  1289  |128   3     289   
128   7     129 |123489 13489 12489 |5     12689 24689 
-------------------------------------------------------
128   6     4   |1238   138   5     |9     7     238   
1258  3     125 |12689  1689  7     |1268  4     2568   
9     1258  7   |123468 13468 1248  |12368 1268  23568 
-------------------------------------------------------
13567 159   159 |145689 2     1489  |34678 689   346789
23567 259   8   |4569   45679 49    |23467 269   1     
1267  4     129 |1689   16789 3     |2678  5     26789 


r1c1 r3c1 r5c3 r6c2 1258;
champagne
2017 Supporter
 
Posts: 5626
Joined: 02 August 2007
Location: France Brittany

Re: exocet pattern in hardest puzzles

Postby daj95376 » Sat Jul 28, 2012 3:01 pm

champagne wrote:..3..6...4.67...3..7....5...6...597..3...7.4.9............2......8.....1.4...3.5.;6519;elev;3594;6;7;

r1c1 r3c1 r5c3 r6c2 1258;

Nice example champagne!!! I wish that I had it in my QExocet analyzer.
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

Re: exocet pattern in hardest puzzles

Postby ronk » Sat Jul 28, 2012 3:23 pm

daj95376 wrote:
champagne wrote:..3..6...4.67...3..7....5...6...597..3...7.4.9............2......8.....1.4...3.5.;6519;elev;3594;6;7;
r1c1 r3c1 r5c3 r6c2 1258;
Nice example champagne!!! I wish that I had it in my QExocet analyzer.
It would be nicer if there was at least one exclusion. ;)
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Re: exocet pattern in hardest puzzles

Postby daj95376 » Sat Jul 28, 2012 6:17 pm

ronk wrote:I don't recall examples for either of the two following illustrations. Examples would be appreciated. [edit: daj95376 and champagne, that means each of you for your own illustration.]

daj95376 wrote:Note: there is a rare scenario that DPB's scenarios don't cover. (the distinction between QExocet and JExocet.)

Code: Select all
 "Q" is forced to contain the candidate
 *-------*-------*-------*
 | B B - | - - - | - - - |
 | - - - | Q / / | R . . |
 | - - - | Q / / | R . . |
 *-------*-------*-------*
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 *-------*-------*-------*
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 *-------*-------*-------*


I don't have an example handy, but I'm working on it. I'm hoping to find an example by comparing my QExocet results for file "02 index.txt" to champagne's "04b Jexocet.txt". That's still being resolved.

I recall DPB and I getting into a "warm discussion" about a puzzle that wasn't a JExocet but was a QExocet. Unfortunately, JExocet puzzles are discussed in at least three threads, and I'm not going to scan them looking for our discussion.

Hopefully, I'll locate an example before I lose track of your message in this thread. In the meantime, I would like to explain my pattern further. First off, it's important to realize that it's a minimal pattern for several possibilities. The important feature is that r23c56 are void for the candidate under consideration.

Code: Select all
Legend:

(G):   a clue/given/solved cell     for the candidate under consideration
(g):   a clue/given/solved cell not for the candidate under consideration

Example Pattern #1: four (g) cells and no (G) cells

Code: Select all
 "Q" is forced to contain the candidate
 *-------*-------*-------*
 | B B - | - - - | - - - |
 | - - - | Q g g | R . . |
 | - - - | Q g g | R . . |
 *-------*-------*-------*
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 *-------*-------*-------*
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 | . . S | S . . | S . . |
 *-------*-------*-------*

Example Pattern #2a and #2b: two (g) cells and one (G) cell

Code: Select all
 "Q" is forced to contain the candidate
 *-------*-------*-------*
 | B B - | - - - | / / / |
 | / / / | / / / | / G / |
 | - - - | Q g g | / / / |
 *-------*-------*-------*
 | . . S | S . . | S . / |
 | . . S | S . . | S . / |
 | . . S | S . . | S . / |
 *-------*-------*-------*
 | . . S | S . . | S . / |
 | . . S | S . . | S . / |
 | . . S | S . . | S . / |
 *-------*-------*-------*

Code: Select all
 "Q" is forced to contain the candidate
 *-------*-------*-------*
 | B B - | - - / | - - - |
 | - - - | Q g / | R . . |
 | - - - | Q g / | R . . |
 *-------*-------*-------*
 | . . S | S . / | S . . |
 | . . S | S . / | S . . |
 | . . S | S . / | S . . |
 *-------*-------*-------*
 | . . S | / / / | S . . |
 | / / / | / / G | / / / |
 | . . S | / / / | S . . |
 *-------*-------*-------*

Example #3: one (g) cell and two (G) cells

Code: Select all
 "Q" is forced to contain the candidate
 *-------*-------*-------*
 | B B - | - - / | / / / |
 | / / / | / / / | / G / |
 | - - - | Q g / | / / / |
 *-------*-------*-------*
 | . . S | S . / | S . / |
 | . . S | S . / | S . / |
 | . . S | S . / | S . / |
 *-------*-------*-------*
 | . . S | / / / | S . / |
 | / / / | / / G | / / / |
 | . . S | / / / | S . / |
 *-------*-------*-------*

Example #4: no (g) cells and two (G) cells -- think JExocet's second pattern

Code: Select all
 "Q" is forced to contain the candidate
 *-------*-------*-------*
 | B B - | - / / | - - - |
 | - - - | Q / / | R . . |
 | - - - | Q / / | R . . |
 *-------*-------*-------*
 | . . S | / / / | S . . |
 | / / / | / G / | / / / |
 | . . S | / / / | S . . |
 *-------*-------*-------*
 | . . S | / / / | S . . |
 | / / / | / / G | / / / |
 | . . S | / / / | S . . |
 *-------*-------*-------*
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

Re: exocet pattern in hardest puzzles

Postby David P Bird » Sat Jul 28, 2012 10:35 pm

daj Your "rare scenario that DPB's scenarios don't cover" is very interesting. It has my prerequisites for a pattern, but I'm reluctant to expand the JE definition to include it firstly because is will be a rather rare oddball and secondly there are a lot of twists and turns needed to describe its different forms. It's probably better suited to templating,

It will help when you find a few examples, but I have a feeling that many of them in the wild will turn out to be resolvable using simpler methods, so won't make it into the hardest collections.

Finally, for your "g" cells, as we had before, isn't it only required that the subject digit doesn't occur in them as a candidate?
David P Bird
2010 Supporter
 
Posts: 957
Joined: 16 September 2008
Location: Middle England

Re: exocet pattern in hardest puzzles

Postby David P Bird » Sat Jul 28, 2012 10:41 pm

ronk wrote: With my illustration as it stands, I have to agree with champagne.

So you agree with this statement from him
I don't see any other possibility that having a candidate locked in the object pair to form an AHS

ronk wrote: It should be OK after removing base candidates in the added AAHS cells

Now what's that got to do with what he wrote? Anyway, I see no AAHS.
Making a derisory point about a failure to pick up a blunder of your own making is some going!

ronk wrote:Good luck with that! I brought up the same point in mid-April and it met the fate of a proverbial lead balloon.

Well depending on Xsudo logic when it isn't needed weighs down the balloon and requires unnecessarily complicating the example diagrams as was explained in the ensuing discussion.

People in grass houses shouldn't stow thrones – or something like that.

BTW I consider it bad practice to make unannounced edits to posts once they have received a response, as it is probable that the changes will go unnoticed.
David P Bird
2010 Supporter
 
Posts: 957
Joined: 16 September 2008
Location: Middle England

Re: exocet pattern in hardest puzzles

Postby daj95376 » Sat Jul 28, 2012 10:49 pm

David P Bird wrote:Finally, for your "g" cells, as we had before, isn't it only required that the subject digit doesn't occur in them as a candidate?

I didn't want to pursue eliminating the candidate value from the (g) cells, so I cheated. _ :) _

As for examples, see my next post.
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

Re: exocet pattern in hardest puzzles

Postby daj95376 » Sat Jul 28, 2012 10:55 pm

Ron,

The first of 653 QExocet puzzles that champagne's solver doesn't consider to be a JExocet.

Note: I had to canonicalize champagne's puzzles in "02 index.txt" and "04b Jexocet.txt" in order to find the 653 puzzles.

Code: Select all
.........4.7..92...8..3...1.9.3....85...78........14....56...9..6...3..7..2.1..6.

 +--------------------------------------------------------------------------------+
 |  12369   1235    1369    |  124578  24568   24567   |  356789  34578   34569   |
 |  4       135     7       |  158     568     9       |  2       358     356     |
 |  269     8       69      |  2457   Q3      Q24567   |  5679    457     1       |
 |--------------------------+--------------------------+--------------------------|
 |  1267    9       146     |  3       2456    2456    |  1567    1257    8       |
 |  5       1234    1346    | B249     7       8       |  1369    123     2369    |
 |  23678   237     368     | B259     2569    1       |  4       2357    23569   |
 |--------------------------+--------------------------+--------------------------|
 |  1378    1347    5       |  6       248     247     |  138     9       234     |
 |  189     6       1489    |  24589  R24589  R3       |  158     12458   7       |
 |  3789    347     2       |  45789   1       457     |  358     6       345     |
 +--------------------------------------------------------------------------------+
 # 162 eliminations remain

 ### -2459- QExocet   Base = r56c4   Target = r3c6==r7c5,r8c5

Code: Select all
 one (g) and two (G) -- Example #3 from above
 +-----------------------------------+
 |  9  .  9  |  .  .  .  |  9  .  9  |
 |  .  .  .  |  .  .  9  |  .  .  .  |
 |  9  .  9  |  .  3  ~  |  9  .  .  |
 |-----------+-----------+-----------|
 |  .  9  .  |  .  .  .  |  .  .  .  |
 |  .  .  .  | B9  .  .  |  9  .  9  |
 |  .  .  .  | B9 -9  .  |  .  .  9  |
 |-----------+-----------+-----------|
 |  .  .  .  |  .  /  /  |  .  9  .  |
 |  9  .  9  | -9  9  3  |  .  .  .  |
 |  9  .  .  | -9  1  /  |  .  .  .  |
 +-----------------------------------+
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

Re: exocet pattern in hardest puzzles

Postby ronk » Sun Jul 29, 2012 2:04 am

daj95376 wrote:The first of 653 QExocet puzzles that champagne's solver doesn't consider to be a JExocet.

Code: Select all
.........4.7..92...8..3...1.9.3....85...78........14....56...9..6...3..7..2.1..6.

 +--------------------------------------------------------------------------------+
 |  12369   1235    1369    |  124578  24568   24567   |  356789  34578   34569   |
 |  4       135     7       |  158     568     9       |  2       358     356     |
 |  269     8       69      |  2457   Q3      Q24567   |  5679    457     1       |
 |--------------------------+--------------------------+--------------------------|
 |  1267    9       146     |  3       2456    2456    |  1567    1257    8       |
 |  5       1234    1346    | B249     7       8       |  1369    123     2369    |
 |  23678   237     368     | B259     2569    1       |  4       2357    23569   |
 |--------------------------+--------------------------+--------------------------|
 |  1378    1347    5       |  6       248     247     |  138     9       234     |
 |  189     6       1489    |  24589  R24589  R3       |  158     12458   7       |
 |  3789    347     2       |  45789   1       457     |  358     6       345     |
 +--------------------------------------------------------------------------------+

 ### -2459- QExocet   Base = r56c4   Target = r3c6==r7c5,r8c5

daj95376, thanks for that example. It appears to have taken a lot more effort to find than I thought it would, so thanks again.

I recognize your 9b8 (empty rectangle) strong inference ("SI"), but also see SI 9c5. Either one forces r8c5=9 when r56c4=9. Seeing the SI in c5 reminds me of a prior example with an SI in a row, but also perpendicular to the "cross-lines."

This said, I also see why your pattern is not considered a "JExocet", for which all strong inferences are defined to be in two base cells and three cross-lines. It's not stated this way, but that's what it is IMO. Lastly, I didn't try to follow your "dependent exocet."
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Re: exocet pattern in hardest puzzles

Postby ronk » Sun Jul 29, 2012 2:51 am

David P Bird wrote:So you agree with this statement
champagne wrote:I don't see any other possibility that having a candidate locked in the object pair to form an AHS

ronk wrote: It should be OK after removing base candidates in the added AAHS cells

Now what's that got to do with what he wrote? Anyway, I see no AAHS.

I think champagne meant ... I don't see any other possibility than having a base candidate locked in the object pair to form an AHS ... with which I now agree.

As to "AHS vs AAHS", I'm betting this is similar to the "hp vs ahp" type arguments I remember seeing on Eureka! No thanks!

As to the rest of your post, I recommend you stick to the topic of sudoku.
ronk
2012 Supporter
 
Posts: 4764
Joined: 02 November 2005
Location: Southeastern USA

Re: exocet pattern in hardest puzzles

Postby champagne » Sun Jul 29, 2012 6:01 am

daj95376 wrote:Note: I had to canonicalize champagne's puzzles in "02 index.txt" and "04b Jexocet.txt" in order to find the 653 puzzles.


Danny,

you used the only file where the puzzle is given in canonical form (one must be there).
I did my best in the last update to use in all other files the "native" form of the puzzle.

The reference file to find missing data should be the file "01 file 1"

Any deviation to that rule would be a handling error of the files on my side.

But this also means that using the file 02 index as input creates trouble
I can add in the zip file a "02 native_index"
champagne
2017 Supporter
 
Posts: 5626
Joined: 02 August 2007
Location: France Brittany

Re: exocet pattern in hardest puzzles

Postby champagne » Sun Jul 29, 2012 8:28 am

daj95376 wrote:The first of 653 QExocet puzzles that champagne's solver doesn't consider to be a JExocet.


I checked that one.

your Qexocet is seen without any problem as a double exocet by my solver.
It's not recognised as a Jexocet due to the "4" pattern.

It is easy, but not direct to see that "4" in the base leads to "4" in the target.

4r56c4 => 4r4c3 => <4>r8c3 => pattern recognised by the Jexocet process

This reflects my understanding of the limits of the Jexocet concept
champagne
2017 Supporter
 
Posts: 5626
Joined: 02 August 2007
Location: France Brittany

Re: exocet pattern in hardest puzzles

Postby daj95376 » Sun Jul 29, 2012 6:27 pm

champagne wrote:your Qexocet is seen without any problem as a double exocet by my solver.
It's not recognised as a Jexocet due to the "4" pattern.

I thought a "double" Exocet had two distinct sets of base cells. Is this what was called a "twin" Exocet? I call it a single QExocet with a single secondary dependancy.

Hmmm, <4> huh!!! Really? When I examine the following grid, I get two columns (outside the BQR chute) containing <4> for the three cross-rows. I thought that met the JExocet definition.

Code: Select all
 +-----------------------------------+
 |  .  .  .  | -4  4  4  |  .  4  4  |
 |  4  .  .  |  .  .  .  |  .  .  .  |
 |  .  .  .  | -4  . Q4  |  . *4  .  | <- cross-row #1
 |-----------+-----------+-----------|
 |  .  . *4  |  . -4 -4  |  .  .  .  | <- cross-row #2
 |  . -4 -4  | B4  .  .  |  .  .  .  |
 |  .  .  .  |  .  .  .  |  4  .  .  |
 |-----------+-----------+-----------|
 |  .  4  .  |  .  4  4  |  .  .  4  |
 |  .  . *4  | -4 R4  .  |  . *4  .  | <- cross-row #3
 |  .  4  .  | -4  .  4  |  .  .  4  |
 +-----------------------------------+
          ^                    ^
          |- column #1         |- column #2

This probably means that some/many of the remaining 652 QExocet puzzles, where you don't find a JExocet present, may actually be over this characteristic. _ :( _
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

Re: exocet pattern in hardest puzzles

Postby champagne » Sun Jul 29, 2012 7:49 pm

daj95376 wrote:
champagne wrote:your Qexocet is seen without any problem as a double exocet by my solver.
It's not recognised as a Jexocet due to the "4" pattern.

I thought a "double" Exocet had two distinct sets of base cells. Is this what was called a "twin" Exocet? I call it a single QExocet with a single secondary dependancy.


The terminology is still informal in that field, but I would agree with you that the term "double exocet" has first been used for 2 conjugated exocet in the same band having 2 different bases and the same digits.

Generally, it is a killer;

Here, my solver sees 2 different exocets sharing the same base and one target cell.


I think the term "twin exocet" is clear. It relates to a specific exocet as in "platinum blonde" where one cell of a classical target is replaced by an AHS.





daj95376 wrote:
champagne wrote:your Qexocet is seen without any problem as a double exocet by my solver.
It's not recognised as a Jexocet due to the "4" pattern.



Hmmm, <4> huh!!! Really? ...

This probably means that some/many of the remaining 652 QExocet puzzles, where you don't find a JExocet present, may actually be over this characteristic. _ :( _


Here I have to be very cautious.

The concept of "J exocet" is not mine. I can be wrong about the limits of that subset of the exocets family.

So far, I have not seen a clear definition of the spec for downgraded forms, so there is room for interpretation.

Anyway, I keep active a "band" exhaustive search covering all cases valid in a band and not complying to my interpretation of the Jexocet characteristics.

The search in that process is a permutation analysis assuming the target false.
champagne
2017 Supporter
 
Posts: 5626
Joined: 02 August 2007
Location: France Brittany

PreviousNext

Return to Advanced solving techniques