moi people,

i've ported an optimizer to processing. optimizer as in meta heuristic combinatorial optimization of parameters from a certain function: in the same vein of genetic algorithms, simulated annealing, tabu search, reinforced learning, that kind of stuff.

this was originally developed in c# as part of my masters thesis to operate with multiparameter functions of different type of parameters (bool, int, float/double). then ported to processing for random graphic folery purposes only. then adapted to also handle single triangles, strips and polygons - trying to replicate roger alsing's monalisa with my own code.
http://rogeralsing.com/2008/12/07/genetic-programming-evolution-of-mona-lisa/

well, the code is getting kinda messy, and the results on graphic evolution didnt come out as i hoped, so i sort of abandoned the thing to dry in a corner.

i have plenty of ideas for evolve the optimizer itself internally into better results and more automated self-mutated configuration based on evolution strategies and particle swarm optimization but i think i should probably structure this whole code mess into a proper p5 library first.

there are a few AI/optimization libs for processing already, but they seem abit stale development wise. so i reckon this could also be useful for others in the p5 community and not just my freetime experimentacrappity smackarism.

i know how to code but i don't have that much experience handling library and inheritance stuff that could make this thing easy to use and update in the future.

so i'm looking for people who know what they're doing in that department and have some free time to look into prep'ing my spaghetti code into a proper lib or atleast give me some fundamental pointers on a few key issues where i have doubts and concearns regarding what is the best path to take.

let me know if you're interested.

love,
filipe cruz
ps / tpolm ^ enoughrecords