qaoalib 0.3.0

A package for QAOA Maxcut calculations

qaoalib

Implementations of VQA simulations for combinatorial optimization (mainly with graph and Max-cut).

v0.2 works towards adapting the circuit simulations with new Qiskit primitives: Sampler and Estimator (starting from Qiskit 1.0).

How to install

You can install from the PyPI:

pip install --upgrade qaoalib

Usage

Solving Max-cut with VQE:

import networkx as nx
from qaoalib.solver import VQE
from qiskit_optimization.applications import Maxcut

G = nx.random_regular_graph(3, 6) # 3-regular graph wtih 6 nodes
qp = Maxcut(G).to_quadratic_program()
ansatz = # some preferred ansatz
solver = VQE(qp, ansatz)
result = solver.solve()
print(result)

Usage (legacy)

Calculate Max-cut expectation with Qmc or QmcFastKron (faster version):

import networkx as nx
from qaoalib.qaoa import Qmc, QmcFastKron

G = nx.fast_gnp_random_graph(10, 0.5)  # Random graph with 10 nodes
params = [0.2, 0.4, 0.3, 0.5]          # 4 params, depth = 2

qmc = Qmc(G, params) # or QmcFastKron(G, params)
qmc.run()
print(qmc.expectation)

Plot landscapes of the QAOA Max-cut expectation function:

import networkx as nx
from qaoalib.qaoa.landscape import HybridFast

G = nx.fast_gnp_random_graph(10, 0.5)
prev_params = [0.1, 0.2] # previous parameters (gamma1, beta1)

ins = HybridFast(G, prev_params) # plots the landscape wrt gamma2 and beta2 with previous parameters given.
ins.create_grid()
ins.show_landscape()