Farfalle

Post puzzles for others to solve here.

Farfalle

Postby shye » Mon Oct 04, 2021 1:36 am

Code: Select all
+-------+-------+-------+
| . 4 8 | 1 . 6 | . . . |
| 5 1 . | . . 3 | 9 4 . |
| 2 . . | . 4 . | 8 1 . |
+-------+-------+-------+
| 6 . . | . . 1 | . 9 8 |
| . 5 . | 8 . . | 6 . . |
| 8 . . | 9 6 . | . . 4 |
+-------+-------+-------+
| 4 . . | . . . | . . 2 |
| 1 . . | . 2 . | . 6 9 |
| . 9 2 | 6 . 5 | 4 8 . |
+-------+-------+-------+
.481.6...51...394.2...4.81.6....1.98.5.8..6..8..96...44.......21...2..69.926.548.

estimated rating: 7.7
two-step solution for me :D
Last edited by shye on Sat Nov 06, 2021 11:45 pm, edited 1 time in total.
User avatar
shye
 
Posts: 275
Joined: 12 June 2021

Re: Farfalle

Postby denis_berthier » Mon Oct 04, 2021 6:54 am

.
Code: Select all
Resolution state after Singles and whips[1]:
   +-------------------+-------------------+-------------------+
   ! 379   4     8     ! 1     579   6     ! 237   237   357   !
   ! 5     1     67    ! 2     8     3     ! 9     4     67    !
   ! 2     367   3679  ! 57    4     79    ! 8     1     3567  !
   +-------------------+-------------------+-------------------+
   ! 6     237   347   ! 3457  357   1     ! 237   9     8     !
   ! 379   5     13479 ! 8     37    247   ! 6     237   137   !
   ! 8     237   137   ! 9     6     27    ! 12357 2357  4     !
   +-------------------+-------------------+-------------------+
   ! 4     3678  3567  ! 37    1379  789   ! 1357  357   2     !
   ! 1     378   357   ! 347   2     478   ! 357   6     9     !
   ! 37    9     2     ! 6     137   5     ! 4     8     137   !
   +-------------------+-------------------+-------------------+
133 candidates.


Reducing the number of steps is meaningful only wrt their maximal allowed size*.
The simplest-first strategy allows an easy solution in S+W4:
Code: Select all
biv-chain[3]: r3c6{n7 n9} - c3n9{r3 r5} - r5n4{c3 c6} ==> r5c6≠7
biv-chain[3]: r5c5{n3 n7} - r6c6{n7 n2} - r5n2{c6 c8} ==> r5c8≠3
biv-chain[4]: r3c6{n7 n9} - c3n9{r3 r5} - r5n4{c3 c6} - b5n2{r5c6 r6c6} ==> r6c6≠7
singles ==> r6c6=2, r5c6=4, r8c4=4, r4c3=4, r5c8=2, r1c7=2, r4c2=2
biv-chain[3]: r7c4{n3 n7} - r8c6{n7 n8} - b7n8{r8c2 r7c2} ==> r7c2≠3
z-chain[3]: r9c1{n7 n3} - r8n3{c2 c7} - r8n5{c7 .} ==> r8c3≠7
biv-chain[4]: r7n6{c3 c2} - b7n8{r7c2 r8c2} - r8c6{n8 n7} - r7c4{n7 n3} ==> r7c3≠3
z-chain[4]: r4c7{n7 n3} - r8c7{n3 n5} - r7c8{n5 n3} - r7c4{n3 .} ==> r7c7≠7
z-chain[4]: r4c7{n3 n7} - r8c7{n7 n5} - r7c8{n5 n7} - r7c4{n7 .} ==> r7c7≠3
z-chain[4]: r2c3{n7 n6} - r7c3{n6 n5} - r7c7{n5 n1} - r6n1{c7 .} ==> r6c3≠7
t-whip[4]: c7n7{r6 r8} - r8c6{n7 n8} - r8c2{n8 n3} - r6c2{n3 .} ==> r6c8≠7
biv-chain[2]: c8n7{r7 r1} - r2n7{c9 c3} ==> r7c3≠7
z-chain[3]: c3n7{r3 r5} - c3n9{r5 r3} - r3c6{n9 .} ==> r3c2≠7
biv-chain[3]: r6c2{n7 n3} - r3c2{n3 n6} - r2c3{n6 n7} ==> r5c3≠7
whip[1]: c3n7{r3 .} ==> r1c1≠7
finned-swordfish-in-rows: n7{r5 r9 r1}{c5 c1 c9} ==> r3c9≠7
finned-swordfish-in-rows: n7{r5 r9 r1}{c5 c1 c9} ==> r2c9≠7
singles ==> r2c9=6, r2c3=7
whip[1]: r3n7{c6 .} ==> r1c5≠7
biv-chain[4]: c5n1{r9 r7} - c5n9{r7 r1} - r1c1{n9 n3} - r9c1{n3 n7} ==> r9c5≠7
biv-chain[3]: c9n1{r5 r9} - r9c5{n1 n3} - r5c5{n3 n7} ==> r5c9≠7
whip[1]: b6n7{r6c7 .} ==> r8c7≠7
naked-pairs-in-a-row: r8{c3 c7}{n3 n5} ==> r8c2≠3
biv-chain[3]: r9c1{n3 n7} - r8n7{c2 c6} - r7c4{n7 n3} ==> r9c5≠3
singles ==> r9c5=1, r7c7=1, r5c9=1, r6c3=1
whip[1]: b8n3{r7c5 .} ==> r7c8≠3
finned-x-wing-in-columns: n3{c8 c2}{r6 r1} ==> r1c1≠3
singles ==> r1c1=9, r1c5=5, r3c4=7, r3c6=9, r7c4=3, r4c4=5, r7c5=9, r3c9=5, r5c3=9
finned-swordfish-in-columns: n3{c3 c2 c7}{r8 r3 r6} ==> r6c8≠3
stte


Even after stretching the maximum length of chains to 8, there's no 1-step or 2-step solution in W8.

Here's a solution in W8 with only 3 non-W1 steps, based on my (slow) implementation (https://github.com/denis-berthier/CSP-Rules-V2.1) of Defise's fewer steps technique described here: http://forum.enjoysudoku.com/reducing-the-number-of-steps-t39234.html

biv-chain-cn[4]: c3n9{r5 r3} - c6n9{r3 r7} - c6n8{r7 r8} - c6n4{r8 r5} ==> r5c3≠4
singles ==> r4c3=4, r5c6=4, r6c6=2, r4c2=2, r5c8=2, r1c7=2, r8c4=4
whip[8]: c3n9{r5 r3} - c6n9{r3 r7} - r7n8{c6 c2} - c2n6{r7 r3} - r3n3{c2 c9} - r1n3{c9 c1} - r9n3{c1 c5} - r5n3{c5 .} ==> r5c3≠1
singles ==> r6c3=1, r5c9=1, r7c7=1, r9c5=1
whip[1]: b8n3{r7c5 .} ==> r7c2≠3, r7c3≠3, r7c8≠3
z-chain[4]: c8n3{r1 r6} - r6c2{n3 n7} - c1n7{r5 r9} - c9n7{r9 .} ==> r1c8≠7
stte

This solution was obtained after only 3 tries.


[Edit] (*) If no limit is set on the size of patterns involved, the puzzle is blown in a single step:
FORCING[3]-T&E(W1) applied to trivalue candidates n1r9c9, n3r9c9 and n7r9c9 :
===> 7 values decided in the three cases: n4r5c6 n7r2c3 n4r8c4 n2r5c8 n7r4c7 n4r4c3 n2r1c7
===> 47 candidates eliminated in the three cases: n7r1c1 n7r1c5 n3r1c7 n7r1c7 n2r1c8 n6r2c3 n7r2c9 n7r3c2 n7r3c3 n3r3c9 n7r3c9 n7r4c2 n3r4c3 n7r4c3 n4r4c4 n7r4c4 n7r4c5 n2r4c7 n3r4c7 n3r5c3 n4r5c3 n7r5c3 n2r5c6 n7r5c6 n3r5c8 n7r5c8 n7r5c9 n3r6c2 n7r6c3 n2r6c7 n7r6c7 n2r6c8 n7r6c8 n3r7c2 n7r7c2 n3r7c3 n7r7c3 n3r7c5 n7r7c6 n3r7c7 n7r7c7 n7r8c3 n3r8c4 n7r8c4 n4r8c6 n7r8c7 n7r9c5
Code: Select all
   +-------------+-------------+-------------+
   ! 39  4   8   ! 1   59  6   ! 2   37  357 !
   ! 5   1   7   ! 2   8   3   ! 9   4   6   !
   ! 2   36  369 ! 57  4   79  ! 8   1   56  !
   +-------------+-------------+-------------+
   ! 6   23  4   ! 35  35  1   ! 7   9   8   !
   ! 379 5   19  ! 8   37  4   ! 6   2   13  !
   ! 8   27  13  ! 9   6   27  ! 135 35  4   !
   +-------------+-------------+-------------+
   ! 4   68  56  ! 37  179 89  ! 15  357 2   !
   ! 1   378 35  ! 4   2   78  ! 35  6   9   !
   ! 37  9   2   ! 6   13  5   ! 4   8   137 !
   +-------------+-------------+-------------+

stte
denis_berthier
2010 Supporter
 
Posts: 3982
Joined: 19 June 2007
Location: Paris

Re: Farfalle

Postby DEFISE » Mon Oct 04, 2021 10:10 am

3 steps in W5:

Singles: 2r2c4, 8r2c5
Block/Line: 5r6b6 => -5r4c7
Block/Line : 5c9b3 => -5r1c7 -5r1c8
whip[4]: r5n4{c6 c3}- c3n9{r5 r3}- c6n9{r3 r7}- c6n8{r7 .} => -4r8c6
Singles: 4r8c4, 4r4c3, 4r5c6, 2r5c8, 2r1c7, 2r4c2, 2r6c6
whip[5]: c4n3{r4 r7}- c8n3{r7 r1}- c9n3{r1 r9}- c1n3{r9 r5}- c5n3{r5 .} => -3r4c7
Singles: 7r4c7, 7r5c5
Block/Line : 7b2r3 => -7r3c2 -7r3c3 -7r3c9
whip[2]: c1n7{r1 r9}- c9n7{r9 .} => -7r1c8
STTE
DEFISE
 
Posts: 270
Joined: 16 April 2020
Location: France

Re: Farfalle

Postby marek stefanik » Mon Oct 04, 2021 10:42 am

Code: Select all
.------------------.-------------------.-------------------.
|*379  4     8     | 1     G579    6   |*2357  *2357 G37–5 |
| 5    1     67    | 27     78     3   | 9      4    *67   |
| 2    367   3679  | 57     4      79  | 8      1    *3567 |
:------------------+-------------------+-------------------:
| 6    237   347   | 23457  357    1   | 2357   9     8    |
|G379  5     13479 | 8     #37     247 | 6      237  G137  |
| 8    237   137   | 9      6      27  | 12357  2357  4    |
:------------------+-------------------+-------------------:
| 4    3678  3567  | 37     13789  789 | 1357   357   2    |
| 1    378   357   | 347    2      478 | 357    6     9    |
|*37   9     2     | 6     G137    5   | 4      8    *137  |
'------------------'-------------------'-------------------'
dual broken wing: whichever digit is true in r5c5 is left with r1c9 as its only guardian => –5r1c9

Multi-links: Show
Code: Select all
.---------.---------.---------.
| *  .  . | .  .  . | *  *  * |
| .  .  . | .  .  . | .  .  * |
| .  .  . | .  .  . | .  .  * |
:---------+---------+---------:
| .  .  . | .  .  . | .  .  . |
| .  .  . | .  .  . | .  .  . |
| .  .  . | .  .  . | .  .  . |
:---------+---------+---------:
| .  .  . | .  .  . | .  .  . |
| .  .  . | .  .  . | .  .  . |
| *  .  . | .  .  . | .  .  * |
'---------'---------'---------'
(2L) X = r19c19b3 / 2


Code: Select all
.---------.---------.---------.
| *  .  . | .  *  . | *  *  * |
| .  .  . | .  .  . | .  .  * |
| .  .  . | .  .  . | .  .  * |
:---------+---------+---------:
| .  .  . | .  .  . | .  .  . |
| *  .  . | .  *  . | .  .  * |
| .  .  . | .  .  . | .  .  . |
:---------+---------+---------:
| .  .  . | .  .  . | .  .  . |
| .  .  . | .  .  . | .  .  . |
| *  .  . | .  *  . | .  .  * |
'---------'---------'---------'
(3L) S = r159c159b3 / 2


11 truths: 37r1, 37r9, 37c1, 37c9, 37b3, r5c5
11 links: (2L)3X, (2L)7X, (3L)3S, (3L)7S, r1c9

Then we arrive at this point:
Code: Select all
.-------------.-----------.--------------.
| 9   4    8  | 1  5   6  | 2    37   37 |
| 5   1   *7  | 2  8  *3  | 9    4    6  |
| 2   36 b*3+6|*7  4   9  | 8    1    5  |
:-------------+-----------+--------------:
| 6   2    4  | 5  37  1  | 37   9    8  |
| 37  5    9  | 8  37  4  | 6    2    1  |
| 8   37   1  | 9  6   2  | 357  357  4  |
:-------------+-----------+--------------:
| 4  d67–8c56 |*3  9 a*7+8| 1    57   2  |
| 1   378  35 | 4  2   78 | 357  6    9  |
| 37  9    2  | 6  1   5  | 4    8    37 |
'-------------'-----------'--------------'
Reverse BUG+2: 8r7c6 = 6r3c3 – 6r7c3 = 6r7c2 => –8r7c2, stte

Marek
marek stefanik
 
Posts: 358
Joined: 05 May 2021

Re: Farfalle

Postby denis_berthier » Mon Oct 04, 2021 1:06 pm

DEFISE wrote:3 steps in W5:
whip[4]: r5n4{c6 c3}- c3n9{r5 r3}- c6n9{r3 r7}- c6n8{r7 .} => -4r8c6
Singles: 4r8c4, 4r4c3, 4r5c6, 2r5c8, 2r1c7, 2r4c2, 2r6c6
whip[5]: c4n3{r4 r7}- c8n3{r7 r1}- c9n3{r1 r9}- c1n3{r9 r5}- c5n3{r5 .} => -3r4c7
Singles: 7r4c7, 7r5c5
Block/Line : 7b2r3 => -7r3c2 -7r3c3 -7r3c9
whip[2]: c1n7{r1 r9}- c9n7{r9 .} => -7r1c8
STTE


Good.
I had tried 1-step and 2-step in W8 and I inadvertently kept 8 as the max length allowed for fewer steps.

After seeing your solution, I checked what I'd get in W5. First try gave me 3 steps also:
biv-chain-cn[4]: c3n9{r5 r3} - c6n9{r3 r7} - c6n8{r7 r8} - c6n4{r8 r5} ==> r5c3≠4
singles ==> r4c3=4, r5c6=4, r6c6=2, r4c2=2, r5c8=2, r1c7=2, r8c4=4
whip-cn[5]: c4n3{r4 r7} - c8n3{r7 r1} - c9n3{r3 r9} - c5n3{r9 r5} - c1n3{r5 .} ==> r4c7≠3
singles ==> r4c7=7, r5c5=7
whip[1]: b2n7{r3c6 .} ==> r3c2≠7, r3c3≠7, r3c9≠7
whip-cn[4]: c4n7{r3 r7} - c8n7{r7 r1} - c9n7{r2 r9} - c1n7{r9 .} ==> r3c4≠5
stte

My second step is the same as yours.

My first step is the same as what I found in W8.
I checked why the second step in W8 was a whip[8]: this long whip[8] has score 20, while the whip-cn[5] here has score 19. As a result, the method could in no case find the whip-cn[5] if length 8 was allowed.
This illustrates three things :
- it is essential to choose the maximum allowed length very carefully (my bad in my previous solution);
- steepest descent can't guarantee the best result (which is already well-known by anyone familiar with steepest descent);
- as a consequence, deviations from steepest descent (as appears to exist in my implementation of the method) may not be fundamentally bad.

Note: I also tried in W4, but the best I got after 3 tries was 5 steps.
denis_berthier
2010 Supporter
 
Posts: 3982
Joined: 19 June 2007
Location: Paris

Re: Farfalle

Postby P.O. » Mon Oct 04, 2021 5:30 pm

Code: Select all
after singles and intersections:

379    4      8      1      579    6      237    237    357             
5      1      67     2      8      3      9      4      67             
2      367   d367-9  57     4     c7+9    8      1      3567           
6      237    347    3457   357    1      237    9      8               
379    5     e1347+9 8      37    f2+47   6      237    137             
8      237    137    9      6      27     12357  2357   4               
4      3678   3567   37     1379  b7+89   1357   357    2               
1      378    357    347    2     a×47-8  357    6      9               
37     9      2      6      137    5      4      8      137 

c6n8{r8 r7} - c6n9{r7 r3} - c3n9{r3 r5} - r5n4{c3 c6} => r8c6 <> 4
singles:
( r1c7b3 n2 r4c2b4 n2 r4c3b4 n4 r5c8b6 n2 r6c6b5 n2 r5c6b5 n4 r8c4b8 n4 )

Code: Select all
 379      4     8           1     579     6     2     37    357           
 5        1     67          2     8       3     9     4     67             
 2        367   3679        57    4       79    8     1     3567           
 6        2     4          b-357 b-357    1    c+37   9     8             
j-(3×7)+9 5    j+7-(139)    8    a+3-7    4     6     2    i+13×7           
 8        37    137         9     6       2     1357  357   4             
 4        3678  3567        37    1379    789   1357  357   2             
 1       e*378 e*357        4     2       78   d-357  6     9             
f3+7      9     2           6    g+1-(37) 5     4     8    h-137   

r5c5{n7 n3} - r4n3{c4c5 c7} - r8n3{c7 c2c3} - r9c1{n3 n7} - r9c5{n7 n1} - c9n1{r9 r5} - r5c3{n1 n7} => r5c1 r5c9 <> 7

Code: Select all
379   4          8         1    h5-79   6     2     37    357           
5     1          ×67       2     8      3     9     4     67             
2    ja+3-(67) je+6-(379) i5+7   4     d7+9   8     1     35×67           
6     2          4         357   357    1     37    9     8             
39    5         f137+9     8    g3+7    4     6     2     13             
8     37         137       9     6      2     1357  357   4             
4    b3+678      3567      37    1379  c7+89  1357  357   2             
1     378        357       4     2      78    357   6     9             
37    9          2         6     137    5     4     8     137   

c2n6{r3 r7} - r7n8{c2 c6} - c6n9{r7 r3} - c3n9{r3 r5} - r5n7{c3 c5} - b2n7{r1c5 r3c4} - r3c3{n7 n6} => r2c3 r3c9 <> 6
ste.
P.O.
 
Posts: 1405
Joined: 07 June 2021

Re: Farfalle

Postby Cenoman » Mon Oct 04, 2021 8:13 pm

My first step is just marek's awsome finding, written into an AIC:
Code: Select all
 +-----------------------+----------------------+------------------------+
 |  379*  4      8       |  1     d579#   6     |  237*    237* ea37-5#  |
 |  5     1      67      |  2      8      3     |  9       4      67*    |
 |  2     367    3679    |  57     4      79    |  8       1      3567*  |
 +-----------------------+----------------------+------------------------+
 |  6     237    347     |  3457   357    1     |  237     9      8      |
 |db379#  5      13479   |  8     c37     247   |  6       237  db137#   |
 |  8     237    137     |  9      6      27    |  12357   2357   4      |
 +-----------------------+----------------------+------------------------+
 |  4     3678   3567    |  37     1379   789   |  1357    357    2      |
 |  1     378    357     |  347    2      478   |  357     6      9      |
 |  37*   9      2       |  6    db137#   5     |  4       8      137*   |
 +-----------------------+----------------------+------------------------+

1. In cells tagged "*", 5-link oddagons (3) & (7), r19, c19, b3 having four and five guardians resp. (3)r5c19, r9c5, r1c9 and (7)r5c19, r19c5, r1c9. Using derived strong links between guardians, the following chain can be inferred:
(3)r1c9 == (3)r5c19, r9c5 - (3=7)r5c5 - (7)r19c5, r5c19 == (7)r1c9 => -5 r1c9; 22 placements & basics

Code: Select all
 +------------------+-----------------+-------------------+
 |  9    4     8    |  1    5    6    |  2     37*   37*  |
 |  5    1     7    |  2    8    3    |  9     4     6    |
 |  2    36    36   |  7    4    9    |  8     1     5    |
 +------------------+-----------------+-------------------+
 |  6    2     4    |  5    37*  1    |  37*   9     8    |
 |  37*  5     9    |  8    37*  4    |  6     2     1    |
 |  8    37    1    |  9    6    2    |  357   5-37  4    |
 +------------------+-----------------+-------------------+
 |  4    678   56   |  3    9    78   |  1     57    2    |
 |  1    378   35   |  4    2    78   |  357   6     9    |
 |  37*  9     2    |  6    1    5    |  4     8     37*  |
 +------------------+-----------------+-------------------+

2. Remote pair (37)r1c8, r4c7 (r1c89, r9c19, r5c15, r4c57) =>-37r6c8; ste
Cenoman
Cenoman
 
Posts: 2769
Joined: 21 November 2016
Location: France

Re: Farfalle

Postby shye » Tue Oct 05, 2021 1:46 am

marek stefanik wrote:dual broken wing: whichever digit is true in r5c5 is left with r1c9 as its only guardian => –5r1c9

Cenoman wrote:My first step is just marek's awsome finding, written into an AIC

awesome! happy to see you both got it ✦ ヮ ✦
i spent some time trying to find a way to represent it in xsudo, but decided wording it would be easier. the multi-links representation is very clever!!

for the second step mine was pretty much the same as cenomans (though it was a 6-cell RP instead, still using mostly the same cells), so i thought i'd share a different one instead
Code: Select all
.-------------.-----------.--------------.
| 9   4    8  | 1  5   6  | 2    37   37 |
| 5   1    7  | 2  8   3  | 9    4    6  |
| 2  #36  #36 | 7  4   9  | 8    1    5  |
:-------------+-----------+--------------:
| 6   2    4  | 5  37  1  | 7-3  9    8  |
| 37  5    9  | 8  37  4  | 6    2    1  |
| 8  #37   1  | 9  6   2  |#357 ~357~ 4  |
:-------------+-----------+--------------:
| 4   678  56 | 3  9   78 | 1    57   2  |
| 1  #378 #35 | 4  2   78 |#357  6    9  |
| 37  9    2  | 6  1   5  | 4    8    37 |
'-------------'-----------'--------------'

finned swordfish
3 in r368 covered by c237, fin in r6c8
=> -3r4c7
User avatar
shye
 
Posts: 275
Joined: 12 June 2021


Return to Puzzles