James-P-D/GeniusSquare

Computation time to show all solutions for given blockers pattern

adibacco opened this issue · 1 comments

How many combinations should the software test to list all possible solutions for a given blockers pattern?

The red piece can be rotated or flipped in 4 different ways and it should be tested on 36 different cells, obviously some positions could have a blocker or another already placed piece but in terms of iterations just the red piece accounts for 4*36 combinations, this value has to be multiplied by the combinations of other pieces. This is a huge number and then listing all possible solutions for a given blockers pattern is not possible in a reasonable time.

What's your opinion?

Hi @adibacco , sorry for the delay in responding, I don't check GitHub too often nowadays. Couple of points regarding your questions:

I don't think 4*36 is a huge number (144), and you have to remember that (a) there won't be the full 36 possible places to check for because of blocker pieces, and (b) once you have a single piece in place, the number of possible locations for the next piece is also smaller because you have blockers and the first piece, and so on, and so on.