testbed_1.py

should_print = True

from qto.model import LinearConstrainedBinaryOptimization as LcboModel
from qto.solvers.optimizers import CobylaOptimizer, AdamOptimizer
from qto.solvers.qiskit import (
    HeaSolver, PenaltySolver, CyclicSolver, ChocoSolver,
    QtoSolver, QtoSimplifySolver, QtoSimplifyDiscardSolver, QtoSimplifyDiscardSegmentedSolver, QtoSimplifyDiscardSegmentedFilterSolver,
    AerProvider, AerGpuProvider, DdsimProvider, FakeBrisbaneProvider, FakeKyivProvider, FakeTorinoProvider, 
)

# model ----------------------------------------------
m = LcboModel()
x = m.addVars(5, name="x")
m.setObjective((x[0] + x[1])* x[3] + x[2], "max")
# m.addConstr(x[0] + x[1] + x[2] == 2)
# m.addConstr(x[0] + x[1] == 1)
# exit()
m.addConstr(x[0] + x[1] - x[2] == 0)
m.addConstr(x[2] + x[3] - x[4] == 1)



print(m.lin_constr_mtx)
# exit()
# m.set_penalty_lambda(0)
print(m)
optimize = m.optimize()
print(f"optimize_cost: {optimize}\n\n")
# sovler ----------------------------------------------
opt = CobylaOptimizer(max_iter=2)
aer = DdsimProvider()
gpu = AerGpuProvider()
fake = FakeBrisbaneProvider()
# opt = AdamOptimizer(max_iter=200)
solver = QtoSimplifyDiscardSolver(
    prb_model=m,  # 问题模型
    optimizer=opt,  # 优化器
    provider=aer,  # 提供器（backend + 配对 pass_mannager ）
    num_layers=1,
    # mcx_mode="linear",
)
print(solver.circuit_analyze(['depth', 'width', 'culled_depth', 'num_one_qubit_gates']))
# print(solver.search())
result = solver.solve()
eval = solver.evaluation()
print(result)
print(eval)
print(opt.cost_history)