m_b_metcalf wrote:Sorry about that, but 'enxio' gives the appearance of being masculine (as opposed to, say, 'enxia', cf. 'claudio' and 'claudia').
Haha!

enxio27 wrote:I'm still at a loss to figure out how y'all are coming up with these high EPs and EDs. Even vicinity searching doesn't seem to be helping me much. Maybe I'm using the wrong seeds, but I don't know how to select better ones.
m_b_metcalf wrote:I use filters in my vicinity searches, for instance, that no result have an ED corresponding to singles, or that ER be at least >~8.5. This hugely reduces the number of puzzles that have to be further processed. Do you have any filters?
I'm not sure how to make a filter within the search. And perhaps I'm not going about the vicinity search the correct (or best) way. Should I have more than one seed puzzle in my seed file? I have been using only one seed puzzle per search (usually one of the highest-rated puzzles played to that point). To perform the vicinity search, I have been using:
- Code: Select all
sudoku-64.exe -go{-4+4}p -euniq()*(clues==23)*(minimal==1) -f%%v <seed.txt >puzzles.txt
and then using serate to rate them:
- Code: Select all
java -cp SudokuExplainer.jar diuf.sudoku.test.serate --format=%%r/%%p/%%d --input=puzzles.txt --output=rating.txt
Then I import the puzzles and their ratings into a spreadsheet and sort them by ER, EP, and ED. At that point, I can use filters to narrow down the field for visual selection. Very clumsy and time consuming, of course, but I don't yet know how to do it more efficiently.