The New Sudoku Players' Forum

[mith]

Code: Select all

+-------+-------+-------+
| . 1 . | 2 . . | . . . |
| . . 3 | . . 4 | 5 . . |
| . 6 . | . 7 . | 8 1 . |
+-------+-------+-------+
| 1 . . | . 6 . | . 8 . |
| . . 4 | 5 . 2 | 3 . . |
| . . . | . . . | . . . |
+-------+-------+-------+
| 7 . . | . . . | . 6 . |
| . . 5 | . . 3 | 2 . . |
| . . . | . . . | . . . |
+-------+-------+-------+
.1.2.......3..45...6..7.81.1...6..8...45.23...........7......6...5..32...........

[SteveG48]

Code: Select all

 *--------------------------------------------------------------------------------------*
 | 45       1        789      | 2        3589     5689     | 4679     3479     34679    |
 | 289      2789     3        |d1689    c189      4        | 5        279     d2679     |
 | 45       6        29       | 39       7        59       | 8        1        2349     |
 *----------------------------+----------------------------+----------------------------|
 | 1        23579    279      | 3479     6        79       | 479      8        24579    |
 | 689      789      4        | 5       b189      2        | 3        79      a179-6    |
 | 23689    235789   26789    | 134789   13489    1789     | 14679    24579    1245679  |
 *----------------------------+----------------------------+----------------------------|
 | 7        23489    1289     | 1489     124589   1589     | 149      6        134589   |
 | 689      489      5        | 146789   1489     3        | 2        479      14789    |
 | 23689    23489    12689    | 146789   124589   156789   | 1479     34579    1345789  |
 *--------------------------------------------------------------------------------------*


1r5c9 = r5c5 - 1r2c5 = (16)r2c49 => -6 r5c9 ; btte

[denis_berthier]

I find the same elimination r5c9 ≠ 6 than SteveG48, but with many simpler ones available before

Code: Select all

***********************************************************************************************
***  SudoRules 20.1.s based on CSP-Rules 2.1.s, config = W+SFin
***  Using CLIPS 6.32-r778
***********************************************************************************************
256 candidates, 1998 csp-links and 1998 links. Density = 6.12%
whip[1]: c2n4{r9 .} ==> r9c1 ≠ 4, r8c1 ≠ 4
whip[1]: c2n5{r6 .} ==> r6c1 ≠ 5
hidden-pairs-in-a-column: c1{n4 n5}{r1 r3} ==> r3c1 ≠ 9, r3c1 ≠ 2, r1c1 ≠ 9, r1c1 ≠ 8
finned-x-wing-in-rows: n1{r5 r8}{c9 c5} ==> r9c5 ≠ 1, r7c5 ≠ 1
finned-x-wing-in-rows: n2{r3 r4}{c9 c3} ==> r6c3 ≠ 2
swordfish-in-columns: n1{c3 c6 c7}{r9 r7 r6} ==> r9c9 ≠ 1, r9c4 ≠ 1, r7c9 ≠ 1, r7c4 ≠ 1, r6c9 ≠ 1, r6c5 ≠ 1, r6c4 ≠ 1
swordfish-in-columns: n3{c1 c5 c8}{r9 r6 r1} ==> r9c9 ≠ 3, r9c2 ≠ 3, r6c4 ≠ 3, r6c2 ≠ 3, r1c9 ≠ 3
swordfish-in-columns: n6{c3 c6 c7}{r6 r9 r1} ==> r9c4 ≠ 6, r9c1 ≠ 6, r6c9 ≠ 6, r6c1 ≠ 6, r1c9 ≠ 6
hidden-pairs-in-a-block: b6{r5c9 r6c7}{n1 n6} ==> r6c7 ≠ 9, r6c7 ≠ 7, r6c7 ≠ 4, r5c9 ≠ 9, r5c9 ≠ 7
hidden-pairs-in-a-column: c4{n1 n6}{r2 r8} ==> r8c4 ≠ 9, r8c4 ≠ 8, r8c4 ≠ 7, r8c4 ≠ 4, r2c4 ≠ 9, r2c4 ≠ 8
whip[1]: r8n7{c9 .} ==> r9c7 ≠ 7, r9c8 ≠ 7, r9c9 ≠ 7
finned-x-wing-in-columns: n7{c7 c3}{r1 r4} ==> r4c2 ≠ 7
finned-x-wing-in-rows: n7{r5 r2}{c2 c8} ==> r1c8 ≠ 7
swordfish-in-rows: n7{r2 r5 r8}{c9 c2 c8} ==> r6c9 ≠ 7, r6c8 ≠ 7, r6c2 ≠ 7, r4c9 ≠ 7, r1c9 ≠ 7
hidden-triplets-in-a-column: c9{n1 n6 n7}{r8 r5 r2} ==> r8c9 ≠ 9, r8c9 ≠ 8, r8c9 ≠ 4, r2c9 ≠ 9, r2c9 ≠ 2
hidden-triplets-in-a-block: b9{r7c9 r9c8 r9c9}{n3 n5 n8} ==> r9c9 ≠ 9, r9c9 ≠ 4, r9c8 ≠ 9, r9c8 ≠ 4, r7c9 ≠ 9, r7c9 ≠ 4
swordfish-in-rows: n8{r2 r5 r8}{c5 c2 c1} ==> r9c5 ≠ 8, r9c2 ≠ 8, r9c1 ≠ 8, r7c5 ≠ 8, r7c2 ≠ 8, r6c5 ≠ 8, r6c2 ≠ 8, r6c1 ≠ 8, r1c5 ≠ 8
naked-triplets-in-a-block: b2{r1c5 r3c4 r3c6}{n5 n3 n9} ==> r2c5 ≠ 9, r1c6 ≠ 9, r1c6 ≠ 5
hidden-triplets-in-a-row: r1{n6 n7 n8}{c6 c7 c3} ==> r1c7 ≠ 9, r1c7 ≠ 4, r1c3 ≠ 9
naked-pairs-in-a-block: b3{r1c7 r2c9}{n6 n7} ==> r2c8 ≠ 7
biv-chain-rn[3]: r2n6{c9 c4} - r2n1{c4 c5} - r5n1{c5 c9} ==> r5c9 ≠ 6
singles and whip[1] to the end

[DEFISE]

Hi Denis,
Amazing, because we could also do directly:
whip[1]: c2n4{r9 .} ==> r9c1 ≠ 4, r8c1 ≠ 4
whip[1]: c2n5{r6 .} ==> r6c1 ≠ 5
hidden-pairs-in-a-column: c1{n4 n5}{r1 r3} ==> r3c1 ≠ 9, r3c1 ≠ 2, r1c1 ≠ 9, r1c1 ≠ 8
biv-chain-rn[3]: r2n6{c9 c4} - r2n1{c4 c5} - r5n1{c5 c9} ==> r5c9 ≠ 6
singles and whip[1] to the end

But I understand your resolution which follows the principle of "simplest first".

[denis_berthier]

Hi Denis,
Amazing, because we could also do directly:
whip[1]: c2n4{r9 .} ==> r9c1 ≠ 4, r8c1 ≠ 4
whip[1]: c2n5{r6 .} ==> r6c1 ≠ 5
hidden-pairs-in-a-column: c1{n4 n5}{r1 r3} ==> r3c1 ≠ 9, r3c1 ≠ 2, r1c1 ≠ 9, r1c1 ≠ 8
biv-chain-rn[3]: r2n6{c9 c4} - r2n1{c4 c5} - r5n1{c5 c9} ==> r5c9 ≠ 6
singles and whip[1] to the end

But I understand your resolution which follows the principle of "simplest first".

As mith puzzles are designed to have many Subsets, I activated Subsets.

[mith]

There are actually even more subsets/fish if you prefer all of them first (as Hodoku does, for example). (There's also a thematic Y-Wing, though it's not available immediately, unlike the many possible eliminations on 6.)

[denis_berthier]

There are actually even more subsets/fish if you prefer all of them first (as Hodoku does, for example).

Probably there are more Subsets, but they are destroyed by other eliminations before being used (SudoRules doesn't mention a Subset if it allows no new elimination).

[mith]

Yes, what I mean is if you find hidden quads and jellyfish before biv-chains, they are present (and given new eliminations). Hodoku finds a jellyfish, 5 swordfish, 2 x-wings, a hidden quad, and a naked quad... before finally trying a W-Wing to finish off the puzzle.

[denis_berthier]

Yes, what I mean is if you find hidden quads and jellyfish before biv-chains, they are present (and given new eliminations). Hodoku finds a jellyfish, 5 swordfish, 2 x-wings, a hidden quad, and a naked quad... before finally trying a W-Wing to finish off the puzzle.

For any puzzle, there are generally many resolution paths.
Rules are ordered differently in SudoRules and in Hodoku. For SudoRules, any pattern of length 3 will come before patterns of size 4 (Jellyfish, Quads, ...). Here the bivalue-chain[3] eliminates the possibility of any subsequent Subsets, as Singles are enough to finish the puzzle.

[mith]

I understand that. I'm certainly not saying the ordering in SudoRules is wrong!