PottsPlayground 0.2.0

For creating and solving Potts-model representations of combinatorial optimization problems

PottsPlayground is for constructing and simulating combinatorial optimization problems represented as Potts Models. It includes a system for constructing weight matrices in Python and a C++/CUDA extension for minimizing the Potts model energy on a CPU or GPU using various flavors of simulated annealing. It is intended as a demonstration of what the Potts model is and is not capable of, and as a tool for further research with the Potts model of computation.

Read the full docs here.

Built-in combinatorial problems can be generated and solved easily:

import PottsPlayground
PottsTask = PottsPlayground.Tasks.TravelingSalesman(ncities=10, ConstraintFactor=1.5)
temp = PottsPlayground.Schedules.SawtoothTempLog2Space(MaxTemp=1., MinTemp=0.05, nTeeth=3, nIters=1e4)
results = PottsPlayground.Anneal(PottsTask, temp)
final_best_soln = results['MinStates'][-1,:]
PottsTask.DisplayState(final_best_soln)

There is also a BaseTask class that can be used to try out any Potts Model problem representations you can think of, and anneal it using the high-performance backend.

Features:

• Structured construction of Potts Models

• Automatic conversion from Potts to Ising models

• Minor Embedding of Ising models

• GPU accelerated annealing and sampling

• Parallel replicates

• Parallel updating

• Comprehensive statistical reporting