The New Sudoku Players' Forum

by **champagne** » Mon Jul 01, 2013 3:45 pm

Leren wrote:
Champagne, could you provide your solution to puzzles 187, 236 and 238 - they are either outside the scope of my code or I might have a bug.

98.7..6..7..5...8...6.....48..3...7..2...........1.5..5....9..7.38....9....8....5;#187

Hidden Text: Show

98.7.....6...5.9....5....7.4...3...2..75..8.......4.5..1..2......98..5.........31;#236

Hidden Text: Show

98.7.....6.....7....7.65...4...8......86..9.......3.42.1...6.3...98..6......2...1;#238

Hidden Text: Show

by **daj95376** » Mon Jul 01, 2013 4:34 pm

champagne wrote:
Leren wrote:
Champagne, could you provide your solution to puzzles 187, 236 and 238 - they are either outside the scope of my code or I might have a bug.

98.7..6..7..5...8...6.....48..3...7..2...........1.5..5....9..7.38....9....8....5;#187
Code: Select all
A B C |D E F |G H I 9 8 12345 |7 234 1234 |6 1235 123 7 14 1234 |5 2346 12346 |1239 8 1239 123 15 6 |129 2389 1238 |7 1235 4 ------------------------------------------------------ 8 14569 1459 |3 24569 2456 |1249 7 1269 1346 2 134579 |469 456789 45678 |1349 1346 13689 346 4679 3479 |2469 1 24678 |5 2346 23689 ------------------------------------------------------ 5 146 124 |1246 2346 9 |8 12346 7 1246 3 8 |1246 57 57 |124 9 126 1246 79 79 |8 2346 12346 |1234 12346 5 16 Truths = {1N5 2N57 3N1 4N7 5N17 6N1 7N5 8N1567 9N157 } 16 Links = {1c17 2c157 3c157 4c157 5r8 6c15 7r8 9c7 } 7 elims 2r3c5 2r4c5 3r3c5 4r4c5 4r5c5 6r4c5 6r5c5

Unbelievable!!! You went to all this trouble to duplicate the eliminations of a Naked Quad in [c5].

by **champagne** » Mon Jul 01, 2013 5:27 pm

daj95376 wrote:Unbelievable!!! You went to all this trouble to duplicate the eliminations of a Naked Quad in [c5].

don't worry danny, it's clear that many of these "limit down" r0 logic have a relatively simple solution using classical moves. Several players as David, Ronk, JC Van Hay already stated that.

In my solver, working as solver, I would never launch the r0 search before all basic moves have been done.

In that investigation step, it's difficult to put the filters at the right level, and, moreover, testing a process can be done without any restriction on a sample eligible to simpler solutions

BTW, if these processes were open to a one column set of cells, the naked quad would come. Looking for more complex patterns, that was hidden

EDIT some additional comments

1) possibly leren does more than me in preliminary eliminations and that could explain that he did not find that r0 logic
2) it makes sense for the r0 search do to more preliminary eliminations and I'll change the code to go in that direction, easy if it is to add quads and jellyfish, not that much problems to add xcyles and may be xycycles (to check) , but I don't have the code for "Sue de coq" pattern seen by JC Van Hay in one of these patterns.

EDIT 3 I made a first test having added as preliminary eliminations (subject to deeper checking)
. quads
. any kind of fish
. XY Wing and XYZ wing

7 puzzles disappeared from the list of 50 201 puzzles, but that lot has an average high rating.
The ratio for lower rating should be much higher.

by **Leren** » Tue Jul 02, 2013 10:15 am

Hi Champagne,

I can now produce solutions for puzzles 236 and 238. As I understand it there is no non-trivial Rank 0 pattern for puzzle 187 (the published patterns only reproduce basic moves).

That completes your exception list for me. Have you made any further progress ?

Leren

by **champagne** » Tue Jul 02, 2013 8:02 pm

Hi leren,

I had still on my side to finish the code for addition of cells.
It's done, but I have now to redo the tests.

My next point will be to digest all that work to see if I can improve the process to the point where it will be possible to use it as main process for the selection of puzzles having a rank 0 logic.

The current runtime can not compete with my previous process.
But I have anyway to test whether that new process can select puzzles not seen by the current one.

The rating of the big file should be over in 2 or 3 days (one batch still running out of 7). At that time, I have to decide what to do.

by **Leren** » Fri Jul 05, 2013 7:21 am

Hi Champagne,

I've got my batch process running and I've processed all puzzles in your 04c multi_fish rank 0.txt and 03 V V loop seen.txt files.

In the first file there are 6333 puzzles and I have found MSLS solutions for all but 1 of them. All the solved puzzles used 4 x 4, 4 x 5 or 5 x 4
cell grids with at most 1 given/solved cell, and no extra cells, so they solved in about the same time as conventional R,C,X Rank 0 Multifish. I also confirmed that there are 20 puzzles inn that
file that have 3 digit base solutions but no 4 digit equivalents.

I had similar results with the V loop file which had 2013 puzzles (all solved). Interestingly 839 of the V loop puzzles had a 3 digit base solution as well as a 4 digit base solution. I even found a MSLS solution for
the "French puzzle" that featured some pages back in this thread, although it was unique in that it had 4 given solved cells in its grid.

I was going to post some statistics of these runs but it's proving too difficult to post the relevant tables in a readable format. I guess the next step will be to code for 2 and 3 digit R,C,X Multifish and apply this before
the MSLS process is applied. I suspect puzzles with appropriate properties will solve quickly with that process and the MSLS process will not have have to used in an inefficient manner as I think we both found with your 241 puzzles.

Leren

by **champagne** » Fri Jul 05, 2013 9:32 am

Hi leren

Leren wrote:I've got my batch process running and I've processed all puzzles in your 04c multi_fish rank 0.txt and 03 V V loop seen.txt files.

sounds very good. I am interesting in getting your average runtime for these files

Leren wrote:I had similar results with the V loop file which had 2013 puzzles (all solved).

Regarding puzzles having a SK loop, the typical signature is a 16 cells base in the 4 boxes. I have difficulties to see that as "in the scope" of your process. If I am right, this means that for each of these puzzles you found an alternative cells base

Leren wrote:I guess the next step will be to code for 2 and 3 digit R,C,X Multifish and apply this before
the MSLS process is applied. I suspect puzzles with appropriate properties will solve quickly with that process and the MSLS process will not have have to used in an inefficient manner as I think we both found with your 241 puzzles.

I think we have 2 main issues :

a) extracting puzzles having a solving potential linked to the r0 logic
b) for such puzzles finding the "best" r0 logic;

From our earlier exchanges, I got the feeling that puzzles rating below 8.0 will generally offer r0 logic redundant with some relatively easy solving rules (locked sets, fishes, X cycles ...)
Also, for puzzles with a higher rating, the preliminary steps should cover all these simple rules;

It will be interesting to see what comes in the area of ratings 7.5_9.5

One key point for extraction of puzzles is to see whether the "cells based" process find puzzles of interest not seen with the RCX search (rows, columns or "X" base). depending on the runtime, this can be done on the entire file of on sample files.

As the sample file created to explore the ratings below the potential hardest should be stable. this could be the right file to explore. I should be in a position to post the pieces of the sample in a split per rating next week.

Working on the b) issue is easier and open to anybody, including manual players. The nExt issue then will be to extract various patterns from that file. JC already expressed some interest for a pure 3x3 logic that you seem to have found. May be you could post the corresponding puzzles

by **David P Bird** » Fri Jul 05, 2013 10:37 am

Hi Champagne,

There are always several ways SK Loops can be represented as locked sets with and without box covers.

My search method looks for diagonals of givens in the usual rectangle of 4 boxes such that 4 digits are covered by the two rows and columns in each of the bands of boxes – that involves 12 cells not 16.

Using the AALS chain it's sometimes possible to prove that every pair of linking digits must contain one truth. I don't think that this can be shown just using strong and weak cover sets, but I could be wrong. How useful this may be is doubtful, but it shows promise when a JExocet pattern also exists (which is reasonably common).

When an Almost SK loop exists IMO it's easier to extend the row and column cover sets that are tried first to find if a larger locked set exists.

It's therefore a question of judgment whether it's worth looking for an SK loop or just letting your new procedure handle these cases.

On another subject

I found that my method of only defining house cover sets can follow AICs provided all the links inside cells are the same type (weak or strong). I'm now exploring a method of allowing the user to identify exception cells to allow a mixture of weak and strong cell cover sets to be used.

David

by **champagne** » Fri Jul 05, 2013 11:57 am

David P Bird wrote:It's therefore a question of judgment whether it's worth looking for an SK loop or just letting your new procedure handle these cases.

Hi David,

A SK loop is so easy to see that I would first look for it.

Regarding your approach, I went several times through your text, but I did not so far identify a promising alternative to leren's design.

by **David P Bird** » Fri Jul 05, 2013 12:32 pm

champagne wrote:Regarding your approach, I went several times through your text, but I did not so far identify a promising alternative to leren's design.

No problem – you're using a brute force iteration of all possibilities which is your choice.

We all reject the dancing links solver because its solutions can't be explained.
You're producing a solution that can be explained, but not in a way that someone can be expected to find manually.
My aim is to find solutions that a determined player could be capable of finding.
It's our choice where we draw our boundaries.

BYW I still have doubts that your samples are representative because they come in families with strong similarities.

David

PS doesn't JC deserve some credt?

by **champagne** » Fri Jul 05, 2013 4:02 pm

David P Bird wrote:You're producing a solution that can be explained, but not in a way that someone can be expected to find manually.

The only merit of these computer built solution is to extract puzzles whose logic can be explored for a later manual research.

Speaking of the r0 logic, it's not so bad. The r0 logic seems to be there in less than 3% of the puzzles.

David P Bird wrote:BTW I still have doubts that your samples are representative because they come in families with strong similarities.

It can be that specific conditions must be there to have a r0 logic. Then, it makes sense to have similarities.

David P Bird wrote:[
PS doesn't JC deserve some credit?

IMO all contributors to that topic (in that thread or in others) deserve some credit.

by **Leren** » Sat Jul 06, 2013 11:39 am

Champagne Wrote: JC already expressed some interest for a pure 3x3 logic that you seem to have found. May be you could post the corresponding puzzles

Hidden Text: Show

98.7.....76....5....4.6....3..9..7....8.2..4.........1.5.6..3......4..7......1..2 498 GP H74 124 R C 206 MSLS 1 124 r35689 c1247 20 12r3 1r5 24r6 12r8 4r9 568c1 379c2 358c4 689c7
98.7.....76....5....4.6....3..9..7....8.2..1.........4.5.6..3......1..7......4..2 1588 GP H292 124 R C 727 MSLS 1 124 r35689 c1247 20 12r3 4r5 12r6 24r8 1r9 568c1 379c2 358c4 689c7
98.7.....76....5....4....9.3..8..9......6...4.....2.1..5.3..8....7.4...2.....1... 3619 GP H763 124 R C 1678 MSLS 1 124 r35689 c1247 20 12r3 12r5 4r6 1r8 24r9 568c1 379c2 569c4 367c7
98.7.....76....5....4......5..8..3......6..2......1..4.3.6..8....7.2..9......4..1 4085 GP H872 124 R C X 1930 MSLS 1 124 r35689 c1247 20 12r3 14r5 2r6 14r8 2r9 368c1 579c2 359c4 679c7
98.7.....76....5....4....9.3..6..9......8...4.....2.1..5.3..6....7.4...2.....1... 5819 GP H1079 124 R C X 2462 MSLS 1 124 r35689 c1247 20 12r3 12r5 4r6 1r8 24r9 568c1 379c2 589c4 378c7
98.7.....65....4....3.6....4..9..5......2..1......6..3.7.6..9......3...5.....1.2. 7334 GP H1302 123 R C X 2704 MSLS 1 123 r35689 c1247 20 12r3 3r5 12r6 12r8 3r9 578c1 469c2 458c4 678c7
98.7.....76....5....4....9.3..8..9......6...2.....4.1..5.3..8....7.2...4.....1... 7623 GP H1481 124 R C X 2888 MSLS 1 124 r35689 c1247 20 12r3 14r5 2r6 1r8 24r9 568c1 379c2 569c4 367c7
98.7.....76....5....4.8....5..3..6......2..9......4..1.3.6..9....9..1..2.......4. 7643 GP H1509 124 R C X 2903 MSLS 1 124 r35689 c1247 20 12r3 14r5 2r6 4r8 12r9 368c1 579c2 589c4 378c7
98.7.....76....5....4......5..3..7....8.4..2......1..9.3.8..6......5..4......2..1 7681 GP H1548 124 R C X 2934 MSLS 1 124 r35689 c1247 20 12r3 1r5 24r6 12r8 4r9 368c1 579c2 569c4 389c7
98.7.....76....5....4.6....5..8..3......2..1......4..9.3.6..8....7....4......1..2 7839 GP H1664 124 R C X 3049 MSLS 1 124 r35689 c1247 20 12r3 4r5 12r6 12r8 4r9 368c1 579c2 359c4 679c7
.2.......456.8....78......4.....53.....9...1.8...4...5.....19.....3...5..4..7...2 8141 col H475 139 X 3283 MSLS 1 139 r14578 c1259 + r1c3 20 139r1 19r4 3r5 3r7 19r8 256c1 67c2 256c5 678c9
98.7.....76....5....4.8....5..3..6......2..9......4..1.7.6..9....9..1..2.......4. 8379 GP H2030 124 R C X 3464 MSLS 1 124 r35689 c1247 20 12r3 14r5 2r6 4r8 12r9 368c1 359c2 589c4 378c7
98.7.....65....4....3.9....4..8..9....5.2..6......1..3.7.5..8......3..2.........1 8638 GP H2246 123 R C X 3655 MSLS 1 123 r35689 c1247 20 12r3 13r5 2r6 1r8 23r9 578c1 469c2 469c4 567c7
98.7.....65....4....3.9....4..8..5......3..6......2..1.7.6..8....6.....3....1..2. 8717 GP H2239 123 R C X 3720 MSLS 1 123 r35689 c1247 20 12r3 12r5 3r6 12r8 3r9 578c1 469c2 459c4 679c7
98.7.....65....4....3.6....7..5..9....5....2......1..3.4.9..7......3..1......2..8 8766 GP H2252 123 R C X 3752 MSLS 1 123 r35689 c1247 20 12r3 13r5 2r6 2r8 13r9 458c1 679c2 468c4 568c7
98.7.....76....5....4.6....5..9..3......2..1......4..8.3.6..9....7....2......1..4 8781 GP H2292 124 R C X 3763 MSLS 1 124 r35689 c1247 20 12r3 4r5 12r6 14r8 2r9 368c1 579c2 358c4 678c7
98.7.....76....5....4....8.5..6..8......3..2......1..4.3.5..6....9..4.......2..1. 8906 GP H2415 124 R C X 3874 MSLS 1 124 r35689 c1247 20 12r3 14r5 2r6 12r8 4r9 368c1 579c2 389c4 379c7
98.7.....76....8....5.6....4..3..9....8.2..1.........5.3.6..7......5..6......1..2 8919 GP H2428 125 R C X 3883 MSLS 1 125 r35689 c1247 20 12r3 5r5 12r6 12r8 5r9 368c1 479c2 489c4 346c7
98.7.....76....5....4.6..7.3..9..8....8.2...1.......4..5.6..9......1...7.....4.2. 12339 GP kz0 124 R C X 4816 MSLS 1 124 r35689 c1247 20 12r3 4r5 12r6 24r8 1r9 568c1 379c2 358c4 367c7

Hi Champagne, I think you were referring to the puzzles I found in your Rank 0 file that had 3 digit Rank 0 bases but no 4 digit bases. You had already flagged these as such but I didn't see them until I had completed the batch run. I confirmed that in each case I can't find a 4 digit base either via an call cells or an RCX Rank 0.

I am interesting in getting your average runtime for these files

I've run a quick test to check the relative times for my all cells and RCX Rank 0 routines. I don't think the absolute times would mean much as I am sure that my platform is much slower than yours. Nevertheless, if I limit the given/solved cells in the all cells process to 1 and don't allow for extra cells then the all cells process runs in about 2/3 of the time of my RCX process. If I allow for more given/solved cells and extra cells the all cells process slows up considerably.

Leren

by **champagne** » Wed Jul 10, 2013 5:31 am

Leren wrote:
I've got my batch process running and I've processed all puzzles in your 04c multi_fish rank 0.txt and 03 V V loop seen.txt files.

In the first file there are 6333 puzzles and I have found MSLS solutions for all but 1 of them. All the solved puzzles used 4 x 4, 4 x 5 or 5 x 4
cell grids with at most 1 given/solved cell, and no extra cells, so they solved in about the same time as conventional R,C,X Rank 0 Multifish. I also confirmed that there are 20 puzzles inn that
file that have 3 digit base solutions but no 4 digit equivalents.

Hi leren,

back to that post, one remark,

The most interesting IMO is not to prove that existing r0 logic have a corresponding cell-base. This has been established by blue in very general terms.
It was good to check the validity of the process through that property, but the next step is to collect puzzles having a r0 logic not yet seen.

The run time is a barrier, but exploring a collection of puzzle can be done slowly.

The next step, to go in the direction of David, would be to find what are the similarities and to define pattern based families as it has been done with the JEs.

I'll start ASAP a full exploration of the data base of potential hardest. I am running a test on a part of the grey zone and I have the feeling that we can expect a better response using the cells search.

by **champagne** » Thu Jul 11, 2013 6:56 am

here a sample file of puzzles rating in the range 95_99
the entire file except the last one has a response using the r0 cells_base search and nothing using the RCX search.
The last one is reverse : no response with the cells-base search, but a RCX solution

may-be leren will have other results.

I am also interested in getting the rows/columns/boxes equivalence to the cells solution

Hidden Text: Show

9..8..7...8..9..6...5..4...8..3..2...2.....7...4..5..81..2...3..3..8.9....7..3..1;cells truths/covers ;16;95;10;10;
9..8..7...8..9..6...5..4...8..6..3...3..8..7...4..5..62..1...9..9..2.1....1.....7;cells truths/covers ;18;98;98;83;
9..8..7...8..9..6...5..4..38..4..6...7..2..8...3..8..51...4.2...6.9...1...7......;cells truths/covers ;17;97;10;10;
9..8..7...8..9..6...5..4..38..9..2...9..2..7...4..5...2..7...1..1...2..7....3.8..;cells truths/covers ;16;95;10;10;
9..8..7...8..9..6...5..4..98..7..3...2..8..1...4..5...1..3...7..7..1.2....3.....1;cells truths/covers ;18;96;96;89;
98.7.......6.5.9.........436...9.2...2...8.....51...3.1..2....4.6..1.8....2..5.1.;cells truths/covers ;18;97;97;17;
98.7.......6.5.9.......6.4.6...3.2...3......1..2..7.5.5..1...2..6..2.5....3..5..4;cells truths/covers ;17;95;95;71;
98.7.......6.5.9.......8.4.6...3.2...3......1..2..7.5.5..1...2..6..2.5....3..5..4;cells truths/covers ;17;95;95;71;
98.7.......6.9.7.......5.4.6...8.4....7..3..5...5...2.1..2....4.6..7.1....9..1.3.;cells truths/covers ;13;97;97;94;
98.7.......6.9.7.......5.8.6...4...3.7...24....46...9.1......25.6.1.......9.2.6..;cells truths/covers ;17;97;10;10;
98.7.......7.6.5.......4.738..........2.4...5.....124.7..1....2.2...96....3.2..5.;cells truths/covers ;19;96;88;17;
98.7.....6.....5....7.6..8.4...9..7...83..6.......2..1.7..8..4...94..3.......1..2;cells truths/covers ;21;95;10;10;
98.7.....6.....85...5.6...94...3...5..74..6.......2.1..5..7...3..43..5.......1.2.;cells truths/covers ;17;98;10;10;
98.7.....6...5.8....4..6.7.5....4.9...93..4......2...1.9...8.4...65..7......1...2;cells truths/covers ;23;95;66;66;
98.7.....6...5.8....4..8.7.5....4.9...39..4......2...1.9...6.4...65..7......1...2;cells truths/covers ;15;95;10;10;
98.7.....6...9.5....7..4...3....2.5...89..6......4...1.2...7.3...64..9......1...6;cells truths/covers ;16;95;10;10;
98.7.....6...9.7....7..5.9.7....4....6..3...2..86..5...2..5..1...69..8.......1..3;cells truths/covers ;13;99;99;66;
98.7.....6..5.......4.3.8..2....31...9.1...2...1.2...8.6......4..3.4..8......53.1;cells truths/covers ;21;95;10;10;
98.7.....6..5..4....3.6..8.3...9..6..9....5....24....1.3..8..2......9..8.....73.6;cells truths/covers ;15;96;10;10;

98.7.....6...5.8....5..4.9.3....5.8...79..5......2...1.1...6.3...95..6..........2;2;3;97;10;10;floors; 589;ntruths; 6;

by **Leren** » Thu Jul 11, 2013 8:33 am

Hi Champagne,

Here are my results of a quick run through your list of puzzles:

Hidden Text: Show

9..8..7...8..9..6...5..4...8..3..2...2.....7...4..5..81..2...3..3..8.9....7..3..1;cells truths/covers ;16;95;10;10; MSLS 1 : Base 28; c1247 r359 + r3c5 r5c3 r5c5 r5c9 r9c5 : 16 Links; 2c1 8c7 ; 1367r3 134569r5 4569r9 ;
9..8..7...8..9..6...5..4...8..6..3...3..8..7...4..5..62..1...9..9..2.1....1.....7;cells truths/covers ;18;98;98;83; MSLS 1 : Base 89; r4578 c369 + r1c3 r2c3 r1c6 r2c6 r1c9 r2c9 : 18 Links; 9r4 9r5 8r7 8r8 ; 2367c3 12367c6 12345c9 ;
9..8..7...8..9..6...5..4..38..4..6...7..2..8...3..8..51...4.2...6.9...1...7......;cells truths/covers ;17;97;10;10; MSLS 1 : Base 89; r1458 c369 + r2c3 r2c6 r7c6 r9c6 r2c9 : 17 Links; 9r4 9r5 8r8 ; 1246c3 123567c6 1247c9 ;
9..8..7...8..9..6...5..4..38..9..2...9..2..7...4..5...2..7...1..1...2..7....3.8..;cells truths/covers ;16;95;10;10; MSLS 1 : Base 89; r2579 c369 + r1c3 r4c3 r1c6 r4c6 r1c9 r4c9 : 18 Links; 8r5 89r7 9r9 ; 1367c3 1367c6 12456c9 ; 2b1 ;
9..8..7...8..9..6...5..4..98..7..3...2..8..1...4..5...1..3...7..7..1.2....3.....1;cells truths/covers ;18;96;96;89; MSLS 1 : Base 78; r12578 c369 + r4c3 r4c6 r4c9 : 18 Links; 7r2 7r5 8r7 8r8 ; 1269c3 12369c6 23456c9 ;
98.7.......6.5.9.........436...9.2...2...8.....51...3.1..2....4.6..1.8....2..5.1.;cells truths/covers ;18;97;97;17;
98.7.......6.5.9.......6.4.6...3.2...3......1..2..7.5.5..1...2..6..2.5....3..5..4;cells truths/covers ;17;95;95;71; MSLS 1 : Base 25; c3567 r157 + r1c8 r5c1 r5c8 r7c2 r7c9 : 17 Links; 5c3 2c6 ; 1346r1 46789r5 346789r7 ;
98.7.......6.5.9.......8.4.6...3.2...3......1..2..7.5.5..1...2..6..2.5....3..5..4;cells truths/covers ;17;95;95;71; MSLS 1 : Base 25; c1368 r158 + r1c5 r1c7 r5c5 r5c7 r8c4 r8c9 : 17 Links; 5c3 2c6 ; 1346r1 46789r5 134789r8 ;
98.7.......6.9.7.......5.4.6...8.4....7..3..5...5...2.1..2....4.6..7.1....9..1.3.;cells truths/covers ;13;97;97;94; MSLS 1 : Base 2345; c689 r1248 + r4c4 : 13 Links; 24c6 5c8 23c9 ; 6r1 8r2 179r4 89r8 ; 1b3 ;
98.7.......6.9.7.......5.8.6...4...3.7...24....46...9.1......25.6.1.......9.2.6..;cells truths/covers ;17;97;10;10; MSLS 1 : Base 69; r12589 c689 + r4c6 r6c6 r4c8 r6c9 : 18 Links; 6r1 6r5 9r8 ; 13478c6 13457c8 12478c9 ;
98.7.......7.6.5.......4.738..........2.4...5.....124.7..1....2.2...96....3.2..5.;cells truths/covers ;19;96;88;17;
98.7.....6.....5....7.6..8.4...9..7...83..6.......2..1.7..8..4...94..3.......1..2;cells truths/covers ;21;95;10;10;
98.7.....6.....85...5.6...94...3...5..74..6.......2.1..5..7...3..43..5.......1.2.;cells truths/covers ;17;98;10;10;
98.7.....6...5.8....4..6.7.5....4.9...93..4......2...1.9...8.4...65..7......1...2;cells truths/covers ;23;95;66;66;
98.7.....6...5.8....4..8.7.5....4.9...39..4......2...1.9...6.4...65..7......1...2;cells truths/covers ;15;95;10;10;
98.7.....6...9.5....7..4...3....2.5...89..6......4...1.2...7.3...64..9......1...6;cells truths/covers ;16;95;10;10; MSLS 1 : Base 69; r1268 c268 + r3c2 r5c2 r9c2 r5c6 r5c8 r9c8 : 17 Links; 6r1 69r6 ; 13457c2 1358c6 12478c8 ;
98.7.....6...9.7....7..5.9.7....4....6..3...2..86..5...2..5..1...69..8.......1..3;cells truths/covers ;13;99;99;66;
98.7.....6..5.......4.3.8..2....31...9.1...2...1.2...8.6......4..3.4..8......53.1;cells truths/covers ;21;95;10;10; MSLS 1 : Base 34; r3469 c1248 + r7c1 r8c1 r8c2 r8c4 r7c8 : 20 Links; 4r4 34r6 4r9 ; 1578c1 1257c2 2689c4 5679c8 ;
98.7.....6..5..4....3.6..8.3...9..6..9....5....24....1.3..8..2......9..8.....73.6;cells truths/covers ;15;96;10;10; MSLS 1 : Base 89; r4578 c347 + r1c3 r2c3 r9c4 r3c7 : 15 Links; 8r4 8r5 9r7 ; 14567c3 1236c4 127c7 ;

98.7.....6...5.8....5..4.9.3....5.8...79..5......2...1.1...6.3...95..6..........2;2;3;97;10;10;floors; 589;ntruths; 6;

As you can see I'm not getting an all cells response in all of them. Perhaps you can provide more details of your all cells solution for, say, this puzzle for which I did not get a response.

98.7.......6.5.9.........436...9.2...2...8.....51...3.1..2....4.6..1.8....2..5.1.;cells truths/covers ;18;97;97;17;

That way I can either widen the scope of my search or look for a deficiency or bug in my code.

Leren

The New Sudoku Players' Forum

Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume

Re: Exotic patterns a resume