The New Sudoku Players' Forum

by **urhegyi** » Mon Apr 05, 2021 10:00 am

: fish.png (14.74 KiB) Viewed 635 times

Code: Select all: 1..4..8.......76....6.1.......7.1.9.4..........2.....3.....35.9..516..4.38..5..1.

by **pjb** » Tue Apr 06, 2021 1:29 am

A lot of fish??
I started with two double finned mutant SFs:

Finned mutant swordfish of 7s (c37r3\r9b16), fin at r3c8, r5c3 => -7 r5c8

Finned mutant swordfish of 9s (c16r1\r8b15), fin at r1c5, r6c1 => -9 r6c5

and then came to a grinding halt, requiring messy chains to advance

Phil

Website · by **denis_berthier** » Tue Apr 06, 2021 4:36 am

SER = 9.0
This puzzle is a typical one in the SER 9.0 category. I've found nothing noticeable (in particular, not a lot of fish - fortunately, I didn't rely on it for lunch).
Generally, this type of puzzle is not proposed to human solvers; but it can be useful from time to time to recall that this is the reality of Sudoku solving when you don't restrict your puzzles to pre-digested ones.
Resolution state after Singles:

Code: Select all: +----------------------+----------------------+----------------------+ ! 1 23579 379 ! 4 239 6 ! 8 2357 257 ! ! 2589 23459 3489 ! 23589 2389 7 ! 6 235 1 ! ! 2578 2357 6 ! 2358 1 258 ! 9 2357 4 ! +----------------------+----------------------+----------------------+ ! 568 356 38 ! 7 2348 1 ! 24 9 258 ! ! 4 135679 13789 ! 235689 2389 2589 ! 127 25678 2578 ! ! 56789 15679 2 ! 5689 489 589 ! 147 5678 3 ! +----------------------+----------------------+----------------------+ ! 26 1246 14 ! 28 7 3 ! 5 28 9 ! ! 279 279 5 ! 1 6 289 ! 3 4 278 ! ! 3 8 79 ! 29 5 4 ! 27 1 6 ! +----------------------+----------------------+----------------------+ 181 candidates, 1002 csp-links and 1002 links. Density = 6.15%

whip[1]: r4n6{c2 .} ==> r6c2 ≠ 6, r5c2 ≠ 6, r6c1 ≠ 6
naked-pairs-in-a-row: r7{c4 c8}{n2 n8} ==> r7c2 ≠ 2, r7c1 ≠ 2
singles ==> r7c1 = 6, r4c2 = 6
whip[1]: b7n2{r8c2 .} ==> r8c6 ≠ 2, r8c9 ≠ 2
whip[1]: b8n2{r9c4 .} ==> r2c4 ≠ 2, r3c4 ≠ 2, r5c4 ≠ 2
z-chain[3]: b2n9{r2c5 r2c4} - c1n9{r2 r8} - c6n9{r8 .} ==> r6c5 ≠ 9
z-chain[3]: b3n7{r3c8 r1c9} - c3n7{r1 r9} - c7n7{r9 .} ==> r5c8 ≠ 7
t-whip[3]: r8c9{n7 n8} - r7c8{n8 n2} - b3n2{r3c8 .} ==> r1c9 ≠ 7
whip[1]: b3n7{r3c8 .} ==> r6c8 ≠ 7
finned-x-wing-in-rows: n7{r9 r6}{c7 c3} ==> r5c3 ≠ 7
z-chain[4]: b8n8{r7c4 r8c6} - c9n8{r8 r4} - c3n8{r4 r2} - c5n8{r2 .} ==> r5c4 ≠ 8
whip[6]: r9c3{n9 n7} - r1c3{n7 n3} - c2n3{r3 r5} - r5n1{c2 c7} - c7n7{r5 r6} - b4n7{r6c1 .} ==> r5c3 ≠ 9
t-whip[7]: r8n8{c6 c9} - r7n8{c8 c4} - r7n2{c4 c8} - b3n2{r3c8 r1c9} - r4c9{n2 n5} - r4c1{n5 n8} - r3n8{c1 .} ==> r5c6 ≠ 8, r6c6 ≠ 8
biv-chain[4]: r9c7{n2 n7} - r8c9{n7 n8} - c6n8{r8 r3} - c6n2{r3 r5} ==> r5c7 ≠ 2
biv-chain[4]: r6c5{n8 n4} - r4n4{c5 c7} - c7n2{r4 r9} - r7c8{n2 n8} ==> r6c8 ≠ 8
biv-chain[3]: r6c6{n9 n5} - r6c8{n5 n6} - b5n6{r6c4 r5c4} ==> r5c4 ≠ 9
z-chain[3]: b4n9{r6c2 r6c1} - c4n9{r6 r9} - c3n9{r9 .} ==> r2c2 ≠ 9
t-whip[4]: r9n7{c3 c7} - r5c7{n7 n1} - b4n1{r5c2 r6c2} - r6n7{c2 .} ==> r8c1 ≠ 7
z-chain[4]: c3n9{r2 r9} - c3n7{r9 r1} - c1n7{r3 r6} - b4n9{r6c1 .} ==> r1c2 ≠ 9
finned-x-wing-in-rows: n9{r9 r1}{c3 c4} ==> r2c4 ≠ 9
whip[1]: b2n9{r2c5 .} ==> r5c5 ≠ 9
z-chain[6]: r7c4{n8 n2} - r9n2{c4 c7} - r4c7{n2 n4} - r6n4{c7 c5} - r6n8{c5 c1} - r3n8{c1 .} ==> r2c4 ≠ 8
z-chain[4]: r2c4{n5 n3} - r2c8{n3 n2} - r1c9{n2 n5} - r4n5{c9 .} ==> r2c1 ≠ 5
z-chain[6]: c8n7{r3 r1} - c8n3{r1 r2} - r2c4{n3 n5} - r3c6{n5 n8} - b8n8{r8c6 r7c4} - r7n2{c4 .} ==> r3c8 ≠ 2
whip[6]: r6c6{n9 n5} - r6c8{n5 n6} - r6c4{n6 n8} - b8n8{r7c4 r8c6} - r3n8{c6 c1} - c1n7{r3 .} ==> r6c1 ≠ 9
whip[1]: b4n9{r6c2 .} ==> r8c2 ≠ 9
biv-chain[3]: c2n9{r5 r6} - r6n1{c2 c7} - r5c7{n1 n7} ==> r5c2 ≠ 7
whip[1]: r5n7{c9 .} ==> r6c7 ≠ 7
z-chain[4]: r3n2{c2 c6} - c6n8{r3 r8} - r8n9{c6 c1} - r8n2{c1 .} ==> r1c2 ≠ 2
z-chain[4]: r3n2{c2 c6} - c6n8{r3 r8} - r8n9{c6 c1} - r8n2{c1 .} ==> r2c2 ≠ 2
whip[4]: r2c4{n3 n5} - r2c8{n5 n2} - r1n2{c9 c5} - c5n9{r1 .} ==> r2c5 ≠ 3
whip[6]: r2c4{n3 n5} - r2c8{n5 n2} - r1n2{c9 c5} - r3c6{n2 n8} - b8n8{r8c6 r7c4} - r7n2{c4 .} ==> r2c2 ≠ 3
whip[6]: r2c4{n3 n5} - r2c8{n5 n2} - r1n2{c9 c5} - b5n2{r4c5 r5c6} - r5n9{c6 c2} - c2n3{r5 .} ==> r2c3 ≠ 3
z-chain[7]: r6c7{n1 n4} - c5n4{r6 r4} - r4n3{c5 c3} - r5c3{n3 n8} - c8n8{r5 r7} - r8c9{n8 n7} - r5n7{c9 .} ==> r5c7 ≠ 1
singles ==> r5c7 = 7, r9c7 = 2, r4c7 = 4, r6c7 = 1, r7c8 = 8
naked-single ==> r7c4 = 2, r8c9 = 7, r8c2 = 2, r8c1 = 9, r8c6 = 8, r9c3 = 7, r9c4 = 9, r6c5 = 4
x-wing-in-rows: n8{r3 r6}{c1 c4} ==> r4c1 ≠ 8, r2c1 ≠ 8
singles ==> r2c1 = 2, r4c1 = 5, r3c6 = 2
whip[1]: c6n5{r6 .} ==> r5c4 ≠ 5, r6c4 ≠ 5
naked-pairs-in-a-row: r1{c3 c5}{n3 n9} ==> r1c8 ≠ 3, r1c2 ≠ 3
naked-pairs-in-a-row: r2{c4 c8}{n3 n5} ==> r2c2 ≠ 5
singles ==> r2c2 = 4, r7c2 = 1, r7c3 = 4, r5c3 = 1
x-wing-in-rows: n3{r1 r4}{c3 c5} ==> r5c5 ≠ 3
biv-chain[3]: r6c8{n5 n6} - r6c4{n6 n8} - r5n8{c5 c9} ==> r5c9 ≠ 5
stte

by **DEFISE** » Tue Apr 06, 2021 12:08 pm

0 fish but an "optimized" T&E(singles only):

8 singles
T&E(8r4c9) => error => -8r4c9
T&E(8r8c9) => error => -8r8c9 => 12 singles
T&E(8r6c1) => error => -8r6c1 => STTE.

by **AnotherLife** » Tue Apr 06, 2021 12:45 pm

I tried to solve this puzzle manually, and after the lcls steps I saw an ALS-chain eliminating the 7 from r1c9.

Code: Select all: .--------------------.-------------------.-------------------. | 1 23579 379 | 4 239 6 | 8 a2357 25-7 | | 2589 23459 3489 | 3589 2389 7 | 6 a235 1 | | 2578 2357 6 | 358 1 258 | 9 a2357 4 | :--------------------+-------------------+-------------------: | 58 6 38 | 7 2348 1 | 24 9 258 | | 4 13579 13789 | 35689 2389 2589 | 127 25678 2578 | | 5789 1579 2 | 5689 489 589 | 147 5678 3 | :--------------------+-------------------+------------------: | 6 14 14 | 28 7 3 | 5 b28 9 | | 279 279 5 | 1 6 89 | 3 4 c78 | | 3 8 79 | 29 5 4 | 27 1 6 | '--------------------'-------------------'-------------------'

Then I eliminated the 7's from r56c8 by the locked candidates rule, and the 7 from r5c3 by a kite. I got stuck here.

Code: Select all: .-------------------.-------------------.-----------------. | 1 23579 379 | 4 239 6 | 8 2357 25 | | 2589 23459 3489 | 3589 2389 7 | 6 235 1 | | 2578 2357 6 | 358 1 258 | 9 2357 4 | :-------------------+-------------------+-----------------: | 58 6 38 | 7 2348 1 | 24 9 258 | | 4 13579 1389 | 35689 2389 2589 | 127 2568 2578 | | 5789 1579 2 | 5689 489 589 | 147 568 3 | :-------------------+-------------------+-----------------: | 6 14 14 | 28 7 3 | 5 28 9 | | 279 279 5 | 1 6 89 | 3 4 78 | | 3 8 79 | 29 5 4 | 27 1 6 | '-------------------'-------------------'-----------------'

HoDoKu can find here two complex fish patterns:
- Finned Franken Swordfish: 9 c16b2 r268 fr1c5 fr5c6 => r6c5<>9
- Finned Franken Jellyfish: 8 r367b8 c1468 fr6c5 efr7c4 => r5c4<>8,
but that does not help very much because now the progress can be made only with the methods of Last Resort, that is, forcing chains and forcing nets.

Website · by **denis_berthier** » Tue Apr 06, 2021 1:50 pm

DEFISE wrote:0 fish but an "optimized" T&E(singles only):
8 singles
T&E(8r4c9) => error => -8r4c9
T&E(8r8c9) => error => -8r8c9 => 12 singles
T&E(8r6c1) => error => -8r6c1 => STTE.

Did you try to find the lengths of the 3 whips or braids allowing these eliminations?

by **Cenoman** » Tue Apr 06, 2021 3:53 pm

Code: Select all: +-------------------------+------------------------+-------------------------+ | 1 23579 C379 | 4 239 6 | 8 2357 257 | | 2589 23459 3489 | 3589 2389 7 | 6 235 1 | | 2578 2357 6 | 358 1 a'25-8 | 9 2357 4 | +-------------------------+------------------------+-------------------------+ | B58 6 C38 | 7 2348 1 | zb24 9 wAa258 | | 4 13579 C13789 | 35689 2389 b'2589 | c'127 25678 2578 | | 5789 D1579 2 | 5689 489 589 | E17-4 5678 3 | +-------------------------+------------------------+-------------------------+ | 6 14 14 | 28 7 3 | 5 28 9 | | 279 279 5 | 1 6 e'89 | 3 4 x78 | | 3 8 C79 |e'29 5 4 | d'y27 1 6 | +-------------------------+------------------------+-------------------------+

1. Kraken cell (258)r4c9 => -4 r6c7
(2)r4c9 - (2=4)r4c7
(5)r4c9 - (5=8)r4c1 - (8=3791)r1459c3 - r6c2 = (1)r6c7
(8)r4c9 - (8=7)r8c9 - (7=2)r9c7 - (2=4)r4c7
----------
=> -4 r6c7; 2 placements (+4 r6c5, r4c7)

2. (2)r3c6 = r5c6 - r5c7 = r9c7 - (2=98)b8p67 =>-8r3c6

Hidden Text: Show

3. Kraken cell (258)r4c9
(2)r4c9 - r5c7 = (2)r9c7
(5)r4c9 - (5=8)r4c1 - r3c1 = r3c4 - (8=2)r7c4
(8)r4c9 - (8=7)r8c9 - (7=2)r9c7
=>-2r9c4; 7 placements & lcls

Hidden Text: Show

4. XW (8)r36\c14 => -8 r24c1, r25c4; +5r4c1, NP(29)r28c1

Hidden Text: Show

5. (9=38)r14c3-(8=179)r6c127 =>-9r5c3
6. (2=5)r3c6-(5=3)r2c4-r2c8=(37)r13c8 =>-2r3c8; 10 placements & lcls

Hidden Text: Show

7. UR(28)r45c59 using mixed external-internal (8)r6c4 == (5)r5c9 - (5=6)r6c8 => -6 r6c4; ste

by **DEFISE** » Tue Apr 06, 2021 9:11 pm

denis_berthier wrote:
DEFISE wrote:0 fish but an "optimized" T&E(singles only):
8 singles
T&E(8r4c9) => error => -8r4c9
T&E(8r8c9) => error => -8r8c9 => 12 singles
T&E(8r6c1) => error => -8r6c1 => STTE.

Did you try to find the lengths of the 3 whips or braids allowing these eliminations?

Yes but I gave up because of the combinatorial explosion.
But if we don't limit ourselves to singles, apart from whips, it’s becomes possible with two whips[12] and
one whip[8].
However, knowing that this puzzle is in W7, here is a more reasonable solution:

Hidden Text: Show

Website · by **denis_berthier** » Wed Apr 07, 2021 4:47 am

DEFISE wrote:
denis_berthier wrote:
DEFISE wrote:0 fish but an "optimized" T&E(singles only):
8 singles
T&E(8r4c9) => error => -8r4c9
T&E(8r8c9) => error => -8r8c9 => 12 singles
T&E(8r6c1) => error => -8r6c1 => STTE.

Did you try to find the lengths of the 3 whips or braids allowing these eliminations?

Yes but I gave up because of the combinatorial explosion.

I also tried with braids but it started to take too long, so I stopped after length 15.
The solution in W8 is great. How many tries did you have to do to find it?

by **DEFISE** » Wed Apr 07, 2021 9:19 am

denis_berthier wrote:The solution in W8 is great. How many tries did you have to do to find it?

I did 15 executions of my "few steps" program with parameter max_length_whip = 8.
The summarized results were as follows:

Target(length whip), Target(length whip), Target(length whip), Target(length whip),
2L4C7(8),2L7C8(8),4L7C2(8),5L1C2(6),
2L4C7(8),8L8C9(7),1L7C3(7),8L5C5(8),
2L4C7(8),2L9C4(8),1L7C3(7),2L4C5(7),
2L4C7(8),2L9C4(8),4L2C3(7),2L1C9(5),
2L4C7(8),9L8C6(8),1L7C3(7),2L4C5(7),
2L4C7(8),2L5C7(7),4L7C2(8),8L5C5(8),
2L4C7(8),8L7C4(7),4L7C2(8),8L4C9(8),
2L4C7(8),8L7C4(7),1L7C3(7),2L1C9(5),
2L4C7(8),8L7C4(7),1L7C3(7),5L5C9(5),
2L4C7(8),2L9C4(8),4L2C3(7),2L4C5(7),
2L4C7(8),2L7C8(8),4L2C3(7),8L4C9(8),
2L4C7(8),8L7C4(7),4L7C2(8),2L1C9(5),
2L4C7(8),2L9C4(8),1L7C3(7),5L5C9(5),
2L4C7(8),8L7C4(7),1L7C3(7),8L5C5(8),
2L4C7(8),2L7C8(8),4L2C3(7),8L5C5(8),
I gave the 8th resolution.
Run time <= 0,4s per execution.
N.B : with parameter max_length_whip = 7 at least 8 whips are required per resolution !

The New Sudoku Players' Forum

An extreme puzzle with a lot of fish

An extreme puzzle with a lot of fish

Re: An extreme puzzle with a lot of fish

Re: An extreme puzzle with a lot of fish

Re: An extreme puzzle with a lot of fish

Re: An extreme puzzle with a lot of fish

Re: An extreme puzzle with a lot of fish

Re: An extreme puzzle with a lot of fish

Re: An extreme puzzle with a lot of fish

Re: An extreme puzzle with a lot of fish

Re: An extreme puzzle with a lot of fish