choicedesign.algorithms

Optimisation algorithms for efficient experimental designs.

Implements three column-wise search heuristics:

• _swapalg — Random Swapping: picks a random attribute column and swaps two row values.

• _rscalg — RSC (Relabelling, Swapping, Cycling): applies one of three random column moves per iteration.

• _federovalg — Modified Federov: replaces one row at a time with the best candidate from the full factorial of attribute levels.

All three algorithms share the same stopping-criteria interface (iter_lim, noimprov_lim, time_lim) and accept a generic criterion callable, so they work with D-, A-, or C-error interchangeably.