SudokuSolver Forum

Good point, I'll whange this in the next release.

_________________
Jean-Christophe
"When you have eliminated the impossible, whatever remains, however improbable, must be the truth." Sherlock Holmes.

Using JC location of the UR Type1, after some hunting to make it appear, I could solve the 4D.
It was interesting to discover and learn about those Unique Rectangles. Thanks for the references

_________________
Nothing can both be and not be

Using Uniqueness Tests in any kind of overlapping puzzles is generally dangerous, unless you 100% know when it can be used and when not.
Refer to Regular Samurai #3.

_________________
Quis custodiet ipsos custodes?

Normal: [D  Y-m-d,  G:i]     PM->email: [D, d M Y H:i:s]

Uniqueness techniques are not dangerous as long as:
1. It as been proven the puzzle has a unique solution.
2. you know where they can be used.

For overlapping puzzles, they generally cannot be used in the overlapping cells.

_________________
Jean-Christophe
"When you have eliminated the impossible, whatever remains, however improbable, must be the truth." Sherlock Holmes.

And AFAIK elsewhere only when all overlapping cells in a (9x9) grid are solved.

_________________
Quis custodiet ipsos custodes?

Normal: [D  Y-m-d,  G:i]     PM->email: [D, d M Y H:i:s]

An example where UR cannot be used. This is a solved samurai with a unique solution.
These two rectangles do not form URs.
The rectangle at the left in g3:r35c23 do not form a UR because the two upper cells (r3c23) also belong to the two columns of the upper-left sub-grid. One cannot swap the 2s and 8s because there would be two 2s and two 8s in these two columns in the upper-left sub-grid.
Similarly for g3:r27c89 which overlap 2*2 columns in the outer sub-grids.

_________________
Jean-Christophe
"When you have eliminated the impossible, whatever remains, however improbable, must be the truth." Sherlock Holmes.

And without using Jean-Christophe location of the UR Type1, I was unable to find the XY-chain .
I used JSudoku to point it out Moving on until a two-string kite appeared! And finally, 4D was done
Nice to be able to track and solve specifics. Thanks Børge and Jean-Christophe for pointing them.

_________________
Nothing can both be and not be

A UR-1 is not necessary for solving 4D. Personally I use whatever techniques makes me solve a Sudoku, and do not feel obliged to use only what a logical solver needed, especially since the solver applies the techniques in a specified sequence. If you use Chains you can avoid most other techniques except Almost Locked Sets, which only and always are required to solve the Wizard editions. Finding a Chain in a Vanilla Sudoku is often hard, so finding one in a 10 grid CS is of course "10 times harder".

This weeks Clueless Special Master edition should be an interesting one. The hardest technique required is Naked/Hidden subsets, but there are lots of them and very many Intersections. The required techniques for the Guru edition reads like a small horror story. Here the required techniques according to JSudoku 1.3b1:

C) Master (81 Intersections, 9 Naked Pairs, 9 Naked Triplets, 5 Naked Quads, 1 Hidden Pair).
D) Guru (72 Intersections, 9 Naked Pairs, 4 Naked Triplets, 1 Naked Quad, 2 Hidden Pairs, 2 Turbot Fishes, 1 Finned X-Wing, 1 Finned Swordfish, 1 Grouped TF, 1 Y-Wing, 4 XY-X-Chains up to 3 links).

_________________
Quis custodiet ipsos custodes?

Normal: [D  Y-m-d,  G:i]     PM->email: [D, d M Y H:i:s]