Built-in Models

qiotoolkit has the following built-in models:

Model	Description
ising	The Ising model consists of binary variables \(`\in\{\pm1\}`\) with a cost function that is the sum of weighted variable-products.
pubo, qubo	The PUBO ("polynomial unconstrained binary optimization") and QUBO ("quadratic ...") models have binary variables \(`\in[0,1]`\) and a cost function that sums terms of 2 (qubo) or more variables (pubo).
blume-capel	The Blume-Capel model extends the Ising model to Spin-1 (i.e, 3 states, \(`[-1,0,1]`\)).
potts-model	Each variable in the Potts model can have one of \(`p`\) values \(`[0,1,\ldots,p-1]`\) and the cost function compares variables for equality. That is, there is no notion of neighboring values
clock-model	Each clock variable (or "rotor") can have on of \(`p`\) discrete values which can be interpreted as a discretized angle (periodic boundaries). The cost function compares how similarly aligned interacting variables are.
tsp	The travelling salesman problem asks to find the shortest tour visiting all nodes in a graph.
poly	Cost function constructed from nested polynomial terms with mutable parameters. Experimental

State Spaces

In addition to the built-in models, components for Partition and Permutation are provided as pre-defined state-spaces.