Optimizer problem

crosspollinator · Post by **crosspollinator** » Tue Jun 27, 2006 5:51 pm

I entered the following puzzle in SudoCue, and clicked on Optimize - Keep Symmetry. It didn't find a more minimal puzzle. However, if you remove the 5 from the center square, the puzzle is still valid and symmetrical.

Code: Select all

. . .|1 3 9|. . .
. . 6|. . .|. . .
. . 4|. . .|8 3 .
-----+-----+-----
2 . .|4 . 6|. . 7
7 . .|. 5 .|. . 6
1 . .|9 . 3|. . 2
-----+-----+-----
. 5 3|. . .|7 . .
. . .|. . .|9 . .
. . .|8 2 5|. . .

crosspollinator · Post by **crosspollinator** » Wed Jun 28, 2006 7:13 pm

I think I might understand what's happening. In this case, when the Optimizer removes the center clue, the Analyzer score indicates that the resulting puzzle is easier than the original. I don't think that it should work that way. When you remove a clue from a puzzle, the puzzle always gets harder. The Optimizer should not compare the Analyzer scores of two puzzles where one puzzle is the same as the other, except for having fewer clues.

Post by **Ruud** » Wed Jun 28, 2006 7:27 pm

I will look into this issue when I have a little more time. The optimizer should only look at the number of clues, and try to minimalize the puzzle.

cheers,

Ruud

crosspollinator · Post by **crosspollinator** » Thu Jun 29, 2006 6:51 pm

It turns out this isn't a bug. Removing the center clue results in a puzzle with multiple solutions. I didn't realize it because the Analyzer results looked the same as for valid puzzles, due to an apparent bug in the Analyzer which causes it to terminate prematurely on some valid puzzles. I think that this is the same bug that I reported in the "Problems with Analyzer" thread on this forum.