def test_profile_with_history():
objective = rosen_for_sensi(max_sensi_order=2, integrated=False)['obj']
with warnings.catch_warnings():
warnings.simplefilter("ignore")
(problem, result, optimizer) = create_optimization_results(objective,
dim_full=5)
profile_options = profile.ProfileOptions(
min_step_size=0.0005,
delta_ratio_max=0.05,
default_step_size=0.005,
ratio_min=0.03,
)
problem.fix_parameters(
[0, 3],
[
result.optimize_result.list[0].x[0],
result.optimize_result.list[0].x[3],
],
)
problem.objective.history = pypesto.MemoryHistory({'trace_record': True})
profile.parameter_profile(
problem=problem,
result=result,
optimizer=optimizer,
profile_index=np.array([0, 2, 4]),
result_index=0,
profile_options=profile_options,
filename=None,
)
def history(request) -> pypesto.History:
if request.param == "memory":
history = pypesto.MemoryHistory(options={'trace_record': True})
elif request.param == "csv":
file = tempfile.mkstemp(suffix='.csv')[1]
history = pypesto.CsvHistory(file, options={'trace_record': True})
else:
history = pypesto.History()
for _ in range(10):
result = {FVAL: np.random.randn(), GRAD: np.random.randn(7)}
history.update(np.random.randn(7), (0, 1), 'mode_fun', result)
history.finalize()
return history
def test_trace_all(self):
history = pypesto.MemoryHistory(options=pypesto.HistoryOptions(
trace_record=True,
trace_record_grad=True,
trace_record_hess=True,
trace_record_res=True,
trace_record_sres=True,
trace_record_chi2=True,
trace_record_schi2=True,
))
self.obj.history = history
self.check_history()
def test_profile_with_history():
objective = test_objective.rosen_for_sensi(max_sensi_order=2,
integrated=False)['obj']
with warnings.catch_warnings():
warnings.simplefilter("ignore")
(problem, result, optimizer) = \
create_optimization_results(objective)
profile_options = profile.ProfileOptions(min_step_size=0.0005,
delta_ratio_max=0.05,
default_step_size=0.005,
ratio_min=0.03)
problem.objective.history = pypesto.MemoryHistory({'trace_record': True})
profile.parameter_profile(problem=problem,
result=result,
optimizer=optimizer,
profile_index=np.array(
[1, 1, 1, 0, 0, 1, 0, 1, 0, 0, 0]),
result_index=0,
profile_options=profile_options)
