pandiagonal magic squares are interesting because they have a lot of internal error correction: values can be altered or erased but the correct square can usually be recovered. (is this true?)
what is an algorithm to randomly generate a pandiagonal magic square? ideally uniformly sample from all possible pandiagonal magic squares, but that seems difficult, so sampling from a substantial subset is good enough.
consider robustly encoding data as a collection of pandiagonal magic squares. each pandiagonal magic square is a glyph. what is a mapping (bijection) between numbers 1 to N and a collection of pandiagonal magic squares?
No comments :
Post a Comment