The New Sudoku Players' Forum

[StrmCkr]

removed irrelevant

[ghfick]

Here is an example using YZF 609

[img]ERI%20Pair.png[/img]

[StrmCkr]

ah that version:
remote pair with dual Eri link

Code: Select all

+----------------+-------------------+---------------+
| .    .     .   | .     -12    .    | .    .    .   |
| .    -12   .   | .     (12)   .    | .    .    .   |
| .    .     .   | .     -12    .    | .    .    .   |
+----------------+-------------------+---------------+
| .    .     .   | /     (12)   /    | .    .    .   |
| -12  (12)  -12 | (12)  (-12)  (12) | -12  -12  -12 |
| .    .     .   | /     (12)   /    | .    .    .   |
+----------------+-------------------+---------------+
| .    .     .   | .     -12    .    | .    .    .   |
| .    .     .   | .     -12    .    | .    .    .   |
| .    .     .   | .     -12    .    | .    .    .   |
+----------------+-------------------+---------------+


which came out of this conversation from udosuku who pointed out how a pattern i found functioned internally with out the need of a locked set.
http://forum.enjoysudoku.com/post62938.html#p62938

which also falls under the W-wing category as moves for 1, and 2 individually but only hits the R2C2 eliminations.
the W-ring Type e version hits all the eliminations as marked however the exemplar was never added as only 2 cases where shown for rings and there is many others that work including this one for

Capture.PNG (39.87 KiB) Viewed 1041 times

i will note

there is some overlap with remote pairs when the a/b group lands on the same sector and are visible both as strong links or reduced to a naked pair from other moves.

Code: Select all

+----------------+----------------+----------------+
| .    .     .   | .    .     .   | .    //    .   |
| -12  (12)  -12 | -12  -12   -12 | -12  (12)  -12 |
| .    .     .   | .    .     .   | .    //    .   |
+----------------+----------------+----------------+
| -12  -12   -12 | -12  (12)  -12 | -12  (12)  -12 |
| .    .     .   | .    .     .   | .    //    .   |
| .    .     .   | .    .     .   | .    //    .   |
+----------------+----------------+----------------+
| .    .     .   | .    .     .   | .    //    .   |
| .    .     .   | .    .     .   | .    //    .   |
| .    .     .   | .    .     .   | .    //    .   |
+----------------+----------------+----------------+

http://forum.enjoysudoku.com/two-definitions-of-w-wing-terminology-question-t6452.html
i added this E type wring as another exemplar for posterity.

edit: re did the attachment noticed a screen typo from my code.

[yzfwsf]

Emm, It's similar to your idea, but I don't think it's exactly the same. I used rank 0 theory to build this pattern. Here I asked all members how to name the pattern.
2*cell Truths+2*box Truths \ 2*row cover + 2* col cover
http://forum.enjoysudoku.com/how-to-name-this-pattern-t37438.html

[StrmCkr]

interesting must have missed that thread:
i would have linked to the thread i just posted in here as its fundamentals are identical as described by udosuku: which lead me to check other techniques specifically w-wing/rings

cross checking overlaps with W-Rings added to previous post

[yzfwsf]

http://forum.enjoysudoku.com/two-definitions-of-w-wing-terminology-question-t6452.html
i added this E type wring as another exemplar for posterity.

Traditionally, W-Wing is a pattern which two bi-value cells connected with single digit (grouped) strong link

[StrmCkr]

If you read the thread it was also expanded into ring varations using the 2nd digit back to the oringal bivavles using a 2nd strong link or grouoed link (these links can overlap)
Just after the inital name was created w wing for non loops and rings for the loops.
ive had both coded in mine for the same age as that thread, and its also in the als w wing/ring code i recently expanded on as mentioned by elven in the thread you linked me to.

[yzfwsf]

If you expand the definition of W-ring, it does contain ERI-Pair, but I don't think it conflicts with giving a name to this special structure, just like 4-node X-Chain can be named by many other names.

[StrmCkr]

Ita Not an expansion
Its
Prexisiting named and defined type.

Adding another named technique for something already defined just muddles the water more.

pre exsiting techniques overlap eachother in speciifc senarios but not in full cases..

Your example case type in all moprhizm match w-rings 1 for 1
. If their was anything that was specific unique to it then sure id go with a new name.

It ia a nice for a re discovery by the way since it was mentioned by a few as something new and others pointed it back to being a w ring.
Either way the final choice of how to use it is your own.

[rjamil]

Hi StrmCkr,

i will note

there is some overlap with remote pairs when the a/b group lands on the same sector and are visible both as strong links or reduced to a naked pair from other moves.

Code: Select all

+----------------+----------------+----------------+
| .    .     .   | .    .     .   | .    .     .   |
| -12  (12)  -12 | -12  -12   -12 | -12  (12)  -12 |
| .    .     .   | .    .     .   | .    .     .   |
+----------------+----------------+----------------+
| -12  -12   -12 | -12  (12)  -12 | -12  (12)  -12 |
| .    .     .   | .    .     .   | .    .     .   |
| .    .     .   | .    .     .   | .    .     .   |
+----------------+----------------+----------------+
| .    .     .   | .    .     .   | .    .     .   |
| .    .     .   | .    .     .   | .    .     .   |
| .    .     .   | .    .     .   | .    .     .   |
+----------------+----------------+----------------+

The above mentioned exemplar, in my humble opinion, falls under degenerated case; and should be detected before Naked Pair move. It must also exclude 12 from r1356789c8 (just like, -12 @ ERI cell as cannibalistic, insisted by tarek here). Otherwise, two row wise Naked Pair moves, i.e., r2c28 and r4c58, may also perform before W-Ring with Naked Pair as strong link.

R. Jamil

[StrmCkr]

Raj, col 8 would have zero eliminations as the formation requires 2 strong links on digit 1&2 on that col. what is missing would be the strong link non grouped cell removing all other non 1,2 digits. From the column cells (ie hidden pair)
Then naked sets on 2 rows applies.

Making this version practically obsolete. Then again 3 moves vs. 1.
Which is. Not the same as
The center cells on the eri are removed as they see. Both 1 and 2 solved and cause errors if they solve as eithet
Which Same principle with the w wing type e, and occurs in extended empty rectangle fishes

[rjamil]

Hi StrmCkr,

Raj, col 8 would have zero eliminations as the formation requires 2 strong links on digit 1&2 on that col. what is missing would be the strong link non grouped cell removing all other non 1,2 digits. From the column cells (ie hidden pair)
Then naked sets on 2 rows applies.

Making this version practically obsolete. Then again 3 moves vs. 1.

Well, then its just a theoretical exemplar that will never exist in real life. Similarly, if the formation requires 2 strong links on digit 1&2 on column 8 then it must also be excluded from b3 other cells too from below exemplar:

Code: Select all

+----------------+----------------+----------------+
| -12  (12)  -12 | -12  -12   -12 | -12  (12)  -12 |
| .    .     .   | .    .     .   | -12  /     -12 |
| -12  -12   -12 | -12  (12)  -12 | -12  (12)  -12 |
+----------------+----------------+----------------+
| .    .     .   | .    .     .   | .    /     .   |
| .    .     .   | .    .     .   | .    /     .   |
| .    .     .   | .    .     .   | .    /     .   |
+----------------+----------------+----------------+
| .    .     .   | .    .     .   | .    /     .   |
| .    .     .   | .    .     .   | .    /     .   |
| .    .     .   | .    .     .   | .    /     .   |
+----------------+----------------+----------------+

Similarly, if the formation requires two strong links on digit 1&2 on box 3 then it must also be excluded from r456789c8:

Code: Select all

+----------------+----------------+----------------+
| -12  (12)  -12 | -12  -12   -12 | /    (12)  /   |
| .    .     .   | .    .     .   | /    /     /   |
| -12  -12   -12 | -12  (12)  -12 | /    (12)  /   |
+----------------+----------------+----------------+
| .    .     .   | .    .     .   | .    -12   .   |
| .    .     .   | .    .     .   | .    -12   .   |
| .    .     .   | .    .     .   | .    -12   .   |
+----------------+----------------+----------------+
| .    .     .   | .    .     .   | .    -12   .   |
| .    .     .   | .    .     .   | .    -12   .   |
| .    .     .   | .    .     .   | .    -12   .   |
+----------------+----------------+----------------+

R. Jamil

[StrmCkr]

Well, then its just a theoretical exemplar that will never exist in real life

They do happen in real life basics remove them before they show up in other techniques.

Those 2 exemplars are correct and the elimination are also already included in the w-ring code.

2 more examples for posterity please note the name tagged by xsudoku as this also happens to the ones I've posted above.{plus all the other formations not documented}

Capture2.PNG (75.41 KiB) Viewed 995 times

Capture.PNG (78.6 KiB) Viewed 995 times

[yzfwsf]

http://forum.enjoysudoku.com/extreme-puzzle-no-5-t37866.html
After discussion here, I check my source code and find that the primary filter is wrong in the puzzle generator. After correction, the puzzle of se9.0 ~ 9.3 can be generated quickly.

Please download the two files at the same time and put them in the same directory, remove ". 7z" at the end of each file name, and then decompress them.

Release 609V3:speed up for generating SE9.0~9.3 puzzles
Please download from next post.

[yzfwsf]

Fix the "bug + 1" text hint bug mentioned by rjamil.
http://forum.enjoysudoku.com/2-skyscrapers-in-sequence-not-so-good-idea-t37887.html#p290231

Please download the two files at the same time and put them in the same directory, remove ". 7z" at the end of each file name, and then decompress them.

YZF_SUDOKU_V4.7z.001.7z

(256 KiB) Downloaded 179 times

YZF_SUDOKU_V4.7z.002.7z

(255.53 KiB) Downloaded 169 times