The New Sudoku Players' Forum

[StrmCkr]

bug:
als- w - wings when double linked rule applies the result display contains errors

Code: Select all

.------------------.-----------------.--------------.
| 4    18     5    | 3   178     178 | 9   6    2   |
| 27   6      3    | 4   28      9   | 1   78   5   |
| 279  1289   289  | 56  56      128 | 3   78   4   |
:------------------+-----------------+--------------:
| 259  2589   2489 | 19  1249    6   | 7   159  3   |
| 239  239    6    | 7   12349   5   | 24  19   8   |
| 1    23579  2479 | 8   2349    234 | 24  59   6   |
:------------------+-----------------+--------------:
| 8    35     1    | 2   3579    37  | 6   4    79  |
| 6    279    279  | 19  48      48  | 5   3    179 |
| 35   4      79   | 56  135679  137 | 8   2    179 |
'------------------'-----------------'--------------'

Grouped Almost Locked Set W-Wing: A=r28c5-{248}, B=r6c7-{24}, connect by 2b5 4b5 Double linked, so Rank0 logic =>
r1c5<>8 r6c2<>2 r6c3<>2 r6c3<>4 r1c1<> r1c1<>

this happens frequently : 2.0.0.625 is the version i have
its also missing r6c6 from the overlapping strong link in the chain part in r6c6 <> 3

[yzfwsf]

bug:
als- w - wings when double linked rule applies the result display contains errors

Code: Select all

.------------------.-----------------.--------------.
| 4    18     5    | 3   178     178 | 9   6    2   |
| 27   6      3    | 4   28      9   | 1   78   5   |
| 279  1289   289  | 56  56      128 | 3   78   4   |
:------------------+-----------------+--------------:
| 259  2589   2489 | 19  1249    6   | 7   159  3   |
| 239  239    6    | 7   12349   5   | 24  19   8   |
| 1    23579  2479 | 8   2349    234 | 24  59   6   |
:------------------+-----------------+--------------:
| 8    35     1    | 2   3579    37  | 6   4    79  |
| 6    279    279  | 19  48      48  | 5   3    179 |
| 35   4      79   | 56  135679  137 | 8   2    179 |
'------------------'-----------------'--------------'

Grouped Almost Locked Set W-Wing: A=r28c5-{248}, B=r6c7-{24}, connect by 2b5 4b5 Double linked, so Rank0 logic =>
r1c5<>8 r6c2<>2 r6c3<>2 r6c3<>4 r1c1<> r1c1<>

this happens frequently : 2.0.0.625 is the version i have
its also missing r6c6 from the overlapping strong link in the chain part in r6c6 <> 3

Thanks for the bug feedback on the wrong text explanation, but I don't know how to explain r6c6<>3. The former is easy to fix since it was forgotten to change when reusing the code, and the latter doesn't know where the logic is implemented.

[StrmCkr]

als a - strong link (x) - als b - strong link (c)
- (x = X) - ( c = C)
C is peer of A
then its a loop and all weak links are now strong links & strong links are now weak {same as chain rules}
the not x and not c end up being the only choice for the cell which is r6c6 <> 3

[yzfwsf]

als a - strong link (x) - als b - strong link (c)
- (x = X) - ( c = C)
C is peer of A
then its a loop and all weak links are now strong links & strong links are now weak {same as chain rules}
the not x and not c end up being the only choice for the cell which is r6c6 <> 3

Chain expression of Grouped Almost Locked Set W-Wing: r6c7(2=4) - r6c56 = r45c5 - r28c5(4=82) - r456c5 = r6c6 - 2r6c7;loop
For r6c6<>3: 4r6c6 = r456 - r28c5(4=82) - r456c5 = 2r6c6 ==>r6c6<>3
I don't think r6c6<>3 can be directly derived from ALS-W-Wing.

[eleven]

... how to explain r6c6<>3.

One of r28c5 has to be 2 or 4, and must go to r6c6 => -3r6c6, -8r1c5.

[yzfwsf]

... how to explain r6c6<>3.

One of r28c5 has to be 2 or 4, and must go to r6c6 => -3r6c6.

I know this, but this is not the elimination of ALS-W-Wing, but the elimination of its contained substructures： 4 Truths = {28N5 24B5}; 4 Links = {248c5 6n6}

[StrmCkr]

It's due to the nature of the Overlaping links( eri) so that r6c6 is 2 or 4 exclusively. ( It's one of the chain elimination rules)

It's not just from the rank zero construction which it also is.

This type of elimination wasn't something I was looking outfor even with eri based w rings

This one was a intresting accidental discovery that when using xsudoku for checking logic on a als w ring validated its inclusion.

my functions primarily deal with the als to als eliminations and what happens if als a and b have double linked rules applied.
Currently it cannot identify this elimination either.

[shye]

there are two deductions here, the ALS w-wing and an almost locked triple of sorts
each is rank0, but if you combine both deductions they no longer are


ALS w-wing:
you can follow this up with a hidden 24 pair in r6

Code: Select all

5 Truths = {28N5 6N7 24B5}
5 Links = {24r6 248c5}
6 Eliminations --> r6c235<>2, r6c35<>4, r1c5<>8

almost locked triple:
you can follow this up with a naked 24 pair in r6

Code: Select all

4 Truths = {28N5 24B5}
4 Links = {248c5 6n6}
2 Eliminations --> r1c5<>8, r6c6<>3

i agree that it feels a little weird to bundle these together

[StrmCkr]

Code: Select all

+---------------+------------------------+----------------+
| .    .    .   | .  -24    .            | .     .    .   |
| .    .    .   | .  -24    .            | .     .    .   |
| .    .    .   | .  -24    .            | .     .    .   |
+---------------+------------------------+----------------+
| .    .    .   | .  (24)   .            | .     .    .   |
| .    .    .   | .  (24)   .            | .     .    .   |
| -24  -24  -24 | .  (-24)  -1356789(24) | (24)  -24  -24 |
+---------------+------------------------+----------------+
| .    .    .   | .  -24    .            | .     .    .   |
| .    .    .   | .  (24)   .            | .     .    .   |
| .    .    .   | .  -24    .            | .     .    .   |
+---------------+------------------------+----------------+

simplified the version into a w- wing doing the same thing
chain form:
(2) 51 = 51(4) - (4)50 = 31,40,49 - (4)67 = 67(2) - (2)31,40,49 = 50(2) ->> loop

flip the links and 50 = 2 or 4

(2) 51 - 51(4) = (4)50 - 31,40,49 = (4)67 - 67(2) = (2)31,40,49 - 50(2) ->> loop

its in the aic chain elimination rules
i cross checked this one with the aic code and it performer's the elimination

the question is do we add chain rules to als W- Rings & W -Rings that have overlaps in the chain parts or just leave it as it adds complexity to elimination rules.

it is interesting as the als can be a locked set 2 ways

als a -> als b with 1 RC + an external strong link
als a -> als B with 2 external strong links ****

currently my model applies the als elimination and doesn't account for the chains being overlapped.

my als +aic chain rules finds the eliminations just fine and the als eliminations on top of it.

where the **** has 2 chains that could potential have an overlap.

[StrmCkr]

Code: Select all

.--------------------.--------------------.---------------.
| 567     478  5678  | 4567    9    456   | 1    3    2   |
| 125679  479  25679 | 234567  17   13456 | 79   679  8   |
| 12679   3    2679  | 267     8    16    | 5    679  4   |
:--------------------+--------------------+---------------:
| 379     5    4     | 39      6    8     | 2    79   1   |
| 39      2    1     | 349     345  7     | 6    8    359 |
| 8       6    379   | 1       35   2     | 379  4    359 |
:--------------------+--------------------+---------------:
| 3579    789  35789 | 3457    17   13459 | 349  2    6   |
| 2369    1    2369  | 8       34   3469  | 349  5    7   |
| 4       79   35679 | 3567    2    3569  | 8    1    39  |
'--------------------'--------------------'---------------'

WXYZ-Wing: 3579 in r6c579,r4c8,Pivot Cell Is r6c9 => r6c3<>3
is missing the double linked Rule -> as it also tags r5c9 <> 9
RCC [7,9]
and it has pivot cells [r6c79]

[yzfwsf]

If the pivot cells are r6c79, the structure will be a "Sue de Coq". In fact the solver also gives this "Sue de Coq".

[StrmCkr]

It's also a wxyz wing as per my post[http://forum.enjoysudoku.com/wxyz-wings-t30012.html

Sue de coqs overlap with als functions as their componet's of Aals and bigger can be combined into smaller als

XY, xyz, wxyz are localized 2 sector als functions.

Unless you wish to use the orginal alinged pair deffintion of these 4.
Ie the 1 pivot needs to have all 4 canddiates and only has 3 bivavles connected to it.

Which basically reduces wxyz to showing up once every 10k puzzles.

Ape evolved into als, so did these 4.

Cheers.

[yzfwsf]

Release V626，I can't log into my Google Drive for now, so I'm using Baidu Drive.
Change chain name, Memery chain to whip
https://pan.baidu.com/s/1UNdY9h1gVDEaJRKUz9KrNQ
Extracted Code：ogns


https://toi.teracloud.jp/share/11c1553ceaa847f3
Extracted Code：enjoysudoku

[denis_berthier]

.
Thanks for the effort, but this is not enough for me. The plagiarism remains on Google.
And who's mad enough to download anything from a website controlled by the CCP?
.

[ghfick]

Thank you to YZF. I was able to download 626 from the teracloud site.

Your visuals and chains of the recently added techniques will be a real help to the Sudoku community. I appreciate, too, that you have added a feature to check if a puzzle is minimal. Very nice! I have only just begun to explore this version. Congratulations! Your work is absolutely first rate.

btw I continue to use Wine to use YZF_Sudoku with the Linux Mint OS. I encourage all members of Sudoku community to try YZF_Sudoku. I am continuing to reach out to others concerning the use of Wine [ and other comparable tools ] with current Mac OS's.

The absurd and harassing note from Dennis The Menace is outrageous. I have reported his abuse and harassment to the forum moderator. He has reached yet another new low by bringing politics into the forum. Ridiculous.