Talk:Random Latin squares: Difference between revisions
→Uniformity over all possible
Line 110:
Ideally I would find examples of what is missing and try and work out an additional algorithm that generates from the whole set. .... If time allowed.<br>
--[[User:Paddy3118|Paddy3118]] ([[User talk:Paddy3118|talk]]) 14:55, 9 June 2019 (UTC)
: It might be worth mentioning that non-uniform solutions are acceptable in the task description. From my cursory research, it appears that generating a uniformly-random latin square is a difficult problem which either requires the exponential-time brute force approach of generating random permutations for rows and restarting the row if a clash occurs, or else requires implementing the involved Jacobson and Matthews algorithm. This also absolves us of some mathematical pedantry. :) --[[User:Chunes|Chunes]] ([[User talk:Chunes|talk]]) 15:14, 9 June 2019 (UTC)
| |||