def make_driver():
ha, vstate, _ = _setup_system(L=2)
return nkx.TDVP(
ha,
vstate,
integrator,
propagation_type="imag",
)
def test_dt_bounds():
ha, vstate, _ = _setup_system(L=2, dtype=np.complex128)
te = nkx.TDVP(
ha,
vstate,
nkx.dynamics.RK23(dt=0.1, adaptive=True, dt_limits=(1e-2, None)),
propagation_type="real",
)
with pytest.warns(UserWarning, match="RK solver: dt reached lower bound"):
te.run(T=0.1, callback=_stop_after_one_step)
def test_one_fixed_step(integrator, propagation_type):
ha, vstate, _ = _setup_system(L=2)
te = nkx.TDVP(
ha,
vstate,
integrator,
qgt=nk.optimizer.qgt.QGTJacobianDense(holomorphic=True),
propagation_type=propagation_type,
)
te.run(T=0.01, callback=_stop_after_one_step)
assert te.t == 0.01
def test_run_twice():
# 1100
ha, vstate, _ = _setup_system(L=2)
driver = nkx.TDVP(
ha,
vstate,
nkx.dynamics.RK23(dt=0.01),
)
driver.run(0.03)
driver.run(0.03)
np.testing.assert_allclose(driver.t, 0.06)
def test_repr_and_info():
ha, vstate, _ = _setup_system(L=2)
driver = nkx.TDVP(
ha,
vstate,
nkx.dynamics.RK23(dt=0.01),
)
print(str(driver))
assert "TDVP" in str(driver)
info = driver.info()
print(info)
assert "TDVP" in info
assert "generator" in info
assert "integrator" in info
assert "RK23" in info
def test_change_integrator():
ha, vstate, _ = _setup_system(L=2)
driver = nkx.TDVP(
ha,
vstate,
nkx.dynamics.RK23(dt=0.01, adaptive=False),
)
driver.run(0.03)
np.testing.assert_allclose(driver.t, 0.03)
integrator = nkx.dynamics.Euler(dt=0.05)
driver.integrator = integrator
np.testing.assert_allclose(driver.t, 0.03)
np.testing.assert_allclose(driver.dt, 0.05)
driver.run(0.1)
np.testing.assert_allclose(driver.t, 0.13)
def test_change_norm():
ha, vstate, _ = _setup_system(L=2)
driver = nkx.TDVP(
ha,
vstate,
nkx.dynamics.RK23(dt=0.01, adaptive=False),
)
driver.run(0.03)
def norm(x):
return 0.0
driver.error_norm = norm
driver.run(0.02)
driver.error_norm = "qgt"
driver.run(0.02)
driver.error_norm = "maximum"
driver.run(0.02)
with pytest.raises(ValueError):
driver.error_norm = "assd"
def test_one_step_lindbladian(integrator):
def _setup_lindbladian_system():
L = 3
hi = nk.hilbert.Spin(s=0.5)**L
ha = nk.operator.LocalOperator(hi)
j_ops = []
for i in range(L):
ha += (0.3 / 2.0) * nk.operator.spin.sigmax(hi, i)
ha += ((2.0 / 4.0) * nk.operator.spin.sigmaz(hi, i) *
nk.operator.spin.sigmaz(hi, (i + 1) % L))
j_ops.append(nk.operator.spin.sigmam(hi, i))
# Create the liouvillian
lind = nk.operator.LocalLiouvillian(ha, j_ops)
# Create NDM and vstate
ma = nk.models.NDM()
sa = nk.sampler.MetropolisLocal(hilbert=nk.hilbert.DoubledHilbert(hi))
sa_obs = nk.sampler.MetropolisLocal(hilbert=hi)
vstate = nk.vqs.MCMixedState(sa,
ma,
sampler_diag=sa_obs,
n_samples=1000,
seed=SEED)
return lind, vstate
lind, vstate = _setup_lindbladian_system()
te = nkx.TDVP(
lind,
vstate,
integrator,
propagation_type="real",
linear_solver=partial(nk.optimizer.solver.svd, rcond=1e-3),
)
te.run(T=0.01, callback=_stop_after_one_step)
assert te.t > 0.0
Sx = sum([nk.operator.spin.sigmax(hi, i) for i in range(L)])
# Run the optimization for 300 iterations to determine the ground state, used as
# initial state of the time-evolution
gs.run(n_iter=300, out="example_ising1d_GS", obs={"Sx": Sx})
# Create integrator for time propagation
integrator = nkx.dynamics.RK23(dt=0.01, adaptive=True, rtol=1e-3, atol=1e-3)
print(integrator)
# Quenched hamiltonian: this has a different transverse field than `ha`
ha1 = nk.operator.Ising(hilbert=hi, graph=g, h=0.5)
te = nkx.TDVP(
ha1,
variational_state=vs,
integrator=integrator,
t0=0.0,
qgt=nk.optimizer.qgt.QGTJacobianDense(holomorphic=True),
error_norm="qgt",
)
log = nk.logging.JsonLog("example_ising1d_TE")
# perform the time-evolution saving the observable Sx at every `tstop` time
te.run(
T=1.0,
out=log,
show_progress=True,
obs={"Sx": Sx},
tstops=np.linspace(0.0, 1.0, 101, endpoint=True),
)
