Browse Source

fixed unit tests

changed shield specification in tests
refactoring
Thomas Knoll 1 year ago
parent
commit
6b8ceedccb
  1. 10
      examples/shields/01_pre_shield_extraction.py
  2. 9
      examples/shields/02_post_shield_extraction.py
  3. 6
      examples/shields/03_optimal_shield_extraction.py
  4. 7
      examples/shields/04_pre_shield_export.py
  5. 5
      examples/shields/05_post_shield_export.py
  6. 5
      examples/shields/06_optimal_shield_export.py
  7. 7
      examples/shields/07_pre_shield_simulator.py
  8. 5
      examples/shields/08_post_shield_simulator.py
  9. 5
      examples/shields/09_pre_shield_decision_tree.py
  10. 5
      examples/shields/10_post_shield_decision_tree.py
  11. 3
      lib/stormpy/__init__.py
  12. 9
      src/core/core.cpp
  13. 8
      src/dft/analysis.cpp
  14. 1
      src/logic/formulae.cpp

10
examples/shields/01_pre_shield_extraction.py

@ -32,18 +32,18 @@ def pre_shield_extraction():
initial_state = model.initial_states[0]
assert initial_state == 0
test = formulas[0].shielding_expression
print(test)
shield_specification = stormpy.logic.ShieldExpression(stormpy.logic.ShieldingType.PRE_SAFETY, stormpy.logic.ShieldComparison.RELATIVE, 0.9) #TODO Parameter for shield expression would be nice to have
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True, shield_expression=shield_specification) #, shielding_expression=shield_specification)
shield_specification = stormpy.logic.ShieldExpression(stormpy.logic.ShieldingType.PRE_SAFETY, stormpy.logic.ShieldComparison.RELATIVE, 0.9)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True, shield_expression=shield_specification)
assert result.has_scheduler
assert result.has_shield
shield = result.shield
pre_scheduler = shield.construct()
for state_id in model.states:
choices = shield.construct().get_choice(state_id)
choices = pre_scheduler.get_choice(state_id)
print(F"Allowed choices in state {state_id}, are {choices.choice_map} ")

9
examples/shields/02_post_shield_extraction.py

@ -20,7 +20,7 @@ for allowed actions.
def post_shield_extraction():
path = stormpy.examples.files.prism_mdp_lava_simple
formula_str = "<PostSafety, gamma=0.9> Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
formula_str = "Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
program = stormpy.parse_prism_program(path)
formulas = stormpy.parse_properties_for_prism_program(formula_str, program)
@ -33,14 +33,17 @@ def post_shield_extraction():
initial_state = model.initial_states[0]
assert initial_state == 0
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True)
shield_specification = stormpy.logic.ShieldExpression(stormpy.logic.ShieldingType.POST_SAFETY, stormpy.logic.ShieldComparison.RELATIVE, 0.9)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True, shield_expression=shield_specification)
assert result.has_scheduler
assert result.has_shield
shield = result.shield
post_scheduler = shield.construct()
for state_id in model.states:
choices = shield.construct().get_choice(state_id)
choices = post_scheduler.get_choice(state_id)
print(F"Applied corrections in state {state_id}, are {choices.choice_map} ")

6
examples/shields/03_optimal_shield_extraction.py

@ -22,7 +22,8 @@ def optimal_shield_extraction():
options.set_build_all_labels()
model = stormpy.build_sparse_model_with_options(program, options)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True)
shield_specification = stormpy.logic.ShieldExpression(stormpy.logic.ShieldingType.OPTIMAL)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True, shield_expression=shield_specification)
assert result.has_scheduler
assert result.has_shield
@ -30,9 +31,10 @@ def optimal_shield_extraction():
shield = result.shield
state_ids = [x for x in model.states]
scheduler = shield.construct()
for state_id in state_ids[0:50]:
choices = shield.construct().get_choice(state_id)
choices = scheduler.get_choice(state_id)
print(F"Corrections in state {state_id}, are {choices.choice_map} ")

7
examples/shields/04_pre_shield_export.py

@ -17,7 +17,7 @@ to a file
def pre_schield():
path = stormpy.examples.files.prism_mdp_lava_simple
formula_str = "<PreSafety, lambda=0.9> Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
formula_str = "Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
program = stormpy.parse_prism_program(path)
formulas = stormpy.parse_properties_for_prism_program(formula_str, program)
@ -30,7 +30,10 @@ def pre_schield():
initial_state = model.initial_states[0]
assert initial_state == 0
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True)
shield_specification = stormpy.logic.ShieldExpression(stormpy.logic.ShieldingType.PRE_SAFETY, stormpy.logic.ShieldComparison.Absolute, 0.9)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True, shield_expression=shield_specification)
assert result.has_scheduler
assert result.has_shield

5
examples/shields/05_post_shield_export.py

@ -17,7 +17,7 @@ to a file
def pre_schield():
path = stormpy.examples.files.prism_mdp_lava_simple
formula_str = "<PostSafety, gamma=1> Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
formula_str = "Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
program = stormpy.parse_prism_program(path)
formulas = stormpy.parse_properties_for_prism_program(formula_str, program)
@ -30,7 +30,8 @@ def pre_schield():
initial_state = model.initial_states[0]
assert initial_state == 0
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True)
shield_specification = stormpy.logic.ShieldExpression(stormpy.logic.ShieldingType.POST_SAFETY, stormpy.logic.ShieldComparison.RELATIVE, 0.9)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True, shield_expression=shield_specification)
assert result.has_scheduler
assert result.has_shield

5
examples/shields/06_optimal_shield_export.py

@ -16,7 +16,7 @@ to a file
def optimal_shield_export():
path = stormpy.examples.files.prism_smg_lights
formula_str = "<Optimal> <<shield>> R{\"differenceWithInterferenceCost\"}min=? [ LRA ]"
formula_str = "<<shield>> R{\"differenceWithInterferenceCost\"}min=? [ LRA ]"
program = stormpy.parse_prism_program(path)
formulas = stormpy.parse_properties_for_prism_program(formula_str, program)
@ -27,7 +27,8 @@ def optimal_shield_export():
options.set_build_all_labels()
model = stormpy.build_sparse_model_with_options(program, options)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True)
shield_specification = stormpy.logic.ShieldExpression(stormpy.logic.ShieldingType.OPTIMAL)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True, shield_expression=shield_specification)
assert result.has_scheduler
assert result.has_shield

7
examples/shields/07_pre_shield_simulator.py

@ -15,7 +15,7 @@ Simulating a model with the usage of a pre shield
def example_pre_shield_simulator():
path = stormpy.examples.files.prism_mdp_cliff_walking
formula_str = "<PreSafety, lambda=0.9> Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
formula_str = "Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
program = stormpy.parse_prism_program(path)
formulas = stormpy.parse_properties_for_prism_program(formula_str, program)
@ -28,7 +28,10 @@ def example_pre_shield_simulator():
initial_state = model.initial_states[0]
assert initial_state == 0
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True)
shield_specification = stormpy.logic.ShieldExpression(stormpy.logic.ShieldingType.PRE_SAFETY, stormpy.logic.ShieldComparison.RELATIVE, 0.9)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True, shield_expression=shield_specification)
assert result.has_scheduler
assert result.has_shield

5
examples/shields/08_post_shield_simulator.py

@ -14,7 +14,7 @@ Simulating a model with the usage of a post shield
def example_post_shield_simulator():
path = stormpy.examples.files.prism_mdp_lava_simple
formula_str = "<PostSafety, gamma=0.9> Pmax=? [G !\"AgentIsInLavaAndNotDone\"]; Pmax=? [ F \"AgentIsInLavaAndNotDone\" ];"
formula_str = "Pmax=? [G !\"AgentIsInLavaAndNotDone\"]; Pmax=? [ F \"AgentIsInLavaAndNotDone\" ];"
program = stormpy.parse_prism_program(path)
formulas = stormpy.parse_properties_for_prism_program(formula_str, program)
@ -27,7 +27,8 @@ def example_post_shield_simulator():
initial_state = model.initial_states[0]
assert initial_state == 0
result = stormpy.model_checking(model, formulas[0])
shield_specification = stormpy.logic.ShieldExpression(stormpy.logic.ShieldingType.POST_SAFETY, stormpy.logic.ShieldComparison.RELATIVE, 0.9)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True, shield_expression=shield_specification)
result2 = stormpy.model_checking(model, formulas[1], extract_scheduler=True)
assert result.has_shield

5
examples/shields/09_pre_shield_decision_tree.py

@ -22,7 +22,7 @@ from stormpy.decision_tree import create_decision_tree
def export_shield_as_dot():
path = stormpy.examples.files.prism_mdp_lava_simple
formula_str = "<PreSafety, lambda=0.9> Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
formula_str = "Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
program = stormpy.parse_prism_program(path)
formulas = stormpy.parse_properties_for_prism_program(formula_str, program)
@ -34,7 +34,8 @@ def export_shield_as_dot():
options.set_build_with_choice_origins(True)
model = stormpy.build_sparse_model_with_options(program, options)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True)
shield_specification = stormpy.logic.ShieldExpression(stormpy.logic.ShieldingType.PRE_SAFETY, stormpy.logic.ShieldComparison.RELATIVE, 0.9)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True, shield_expression=shield_specification)
assert result.has_shield

5
examples/shields/10_post_shield_decision_tree.py

@ -12,7 +12,7 @@ from stormpy.decision_tree import create_decision_tree
def export_shield_as_dot():
path = stormpy.examples.files.prism_mdp_lava_simple
formula_str = "<PostSafety, gamma=1> Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
formula_str = "Pmax=? [G !\"AgentIsInLavaAndNotDone\"]"
program = stormpy.parse_prism_program(path)
formulas = stormpy.parse_properties_for_prism_program(formula_str, program)
@ -24,7 +24,8 @@ def export_shield_as_dot():
options.set_build_with_choice_origins(True)
model = stormpy.build_sparse_model_with_options(program, options)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True)
shield_specification = stormpy.logic.ShieldExpression(stormpy.logic.ShieldingType.POST_SAFETY, stormpy.logic.ShieldComparison.RELATIVE, 0.9)
result = stormpy.model_checking(model, formulas[0], extract_scheduler=True, shield_expression=shield_specification)
assert result.has_shield

3
lib/stormpy/__init__.py

@ -325,9 +325,6 @@ def check_model_sparse(model, property, only_initial_states=False, extract_sched
if shield_expression is not None:
task.set_shielding_expression(shield_expression)
elif property.is_shielding_property():
task.set_shielding_expression(property.shielding_expression)
return core._model_checking_sparse_engine(model, task, environment=environment)

9
src/core/core.cpp

@ -119,9 +119,9 @@ void define_build(py::module& m) {
m.def("_build_sparse_parametric_model_from_drn", &storm::api::buildExplicitDRNModel<storm::RationalFunction>, "Build the parametric model from DRN", py::arg("file"), py::arg("options") = storm::parser::DirectEncodingParserOptions());
m.def("build_sparse_model_from_explicit", &storm::api::buildExplicitModel<double>, "Build the model model from explicit input", py::arg("transition_file"), py::arg("labeling_file"), py::arg("state_reward_file") = "", py::arg("transition_reward_file") = "", py::arg("choice_labeling_file") = "");
m.def("make_sparse_model_builder", &storm::api::makeExplicitModelBuilder<double>, "Construct a builder instance", py::arg("model_description"), py::arg("options"), py::arg("action_mask"));
m.def("make_sparse_model_builder_exact", &storm::api::makeExplicitModelBuilder<storm::RationalNumber>, "Construct a builder instance", py::arg("model_description"), py::arg("options"), py::arg("action_mask"));
m.def("make_sparse_model_builder_parametric", &storm::api::makeExplicitModelBuilder<double>, "Construct a builder instance", py::arg("model_description"), py::arg("options"), py::arg("action_mask"));
m.def("make_sparse_model_builder", &storm::api::makeExplicitModelBuilder<double>, "Construct a builder instance", py::arg("model_description"), py::arg("options"), py::arg("action_mask") = nullptr);
m.def("make_sparse_model_builder_exact", &storm::api::makeExplicitModelBuilder<storm::RationalNumber>, "Construct a builder instance", py::arg("model_description"), py::arg("options"), py::arg("action_mask") = nullptr);
m.def("make_sparse_model_builder_parametric", &storm::api::makeExplicitModelBuilder<storm::RationalFunction>, "Construct a builder instance", py::arg("model_description"), py::arg("options"), py::arg("action_mask") = nullptr);
py::class_<storm::builder::ExplicitModelBuilder<double>>(m, "ExplicitModelBuilder", "Model builder for sparse models")
.def("build", &storm::builder::ExplicitModelBuilder<double>::build, "Build the model")
@ -150,6 +150,9 @@ void define_build(py::module& m) {
.def("set_build_choice_labels", &storm::builder::BuilderOptions::setBuildChoiceLabels, "Build with choice labels", py::arg("new_value")=true)
.def("set_build_all_labels" , &storm::builder::BuilderOptions::setBuildAllLabels, "Build with all state labels", py::arg("new_value")=true)
.def("set_build_all_reward_models", &storm::builder::BuilderOptions::setBuildAllRewardModels, "Build with all reward models", py::arg("new_value")=true);
py::class_<storm::generator::ActionMask<double>, std::shared_ptr<storm::generator::ActionMask<double>>> actionmask(m, "ActionMaskDouble");
}
void define_optimality_type(py::module& m) {

8
src/dft/analysis.cpp

@ -9,8 +9,8 @@
// Thin wrapper for DFT analysis
template<typename ValueType>
std::vector<ValueType> analyzeDFT(storm::storage::DFT<ValueType> const& dft, std::vector<std::shared_ptr<storm::logic::Formula const>> const& properties, bool symred, bool allowModularisation, storm::utility::RelevantEvents const& relevantEvents) {
typename storm::modelchecker::DFTModelChecker<ValueType>::dft_results dftResults = storm::api::analyzeDFT(dft, properties, symred, allowModularisation, relevantEvents, 0.0, storm::builder::ApproximationHeuristic::DEPTH, false);
std::vector<ValueType> analyzeDFT(storm::storage::DFT<ValueType> const& dft, std::vector<std::shared_ptr<storm::logic::Formula const>> const& properties, bool symred, bool allowModularisation, storm::utility::RelevantEvents const& relevantEvents, bool allowDCForRelevant) {
typename storm::modelchecker::DFTModelChecker<ValueType>::dft_results dftResults = storm::api::analyzeDFT(dft, properties, symred, allowModularisation, relevantEvents, allowDCForRelevant ,0.0, storm::builder::ApproximationHeuristic::DEPTH, false);
std::vector<ValueType> results;
for (auto result : dftResults) {
@ -27,7 +27,7 @@ void define_analysis(py::module& m) {
.def("is_relevant", &storm::utility::RelevantEvents::isRelevant, "Check whether the given name is a relevant event", py::arg("name"))
;
m.def("analyze_dft", &analyzeDFT<double>, "Analyze the DFT", py::arg("dft"), py::arg("properties"), py::arg("symred")=true, py::arg("allow_modularisation")=false, py::arg("relevant_events")=storm::utility::RelevantEvents());
m.def("analyze_dft", &analyzeDFT<double>, "Analyze the DFT", py::arg("dft"), py::arg("properties"), py::arg("symred")=true, py::arg("allow_modularisation")=false, py::arg("relevant_events")=storm::utility::RelevantEvents(), py::arg("allow_dc_for_relevant")=false);
m.def("transform_dft", &storm::api::applyTransformations<double>, "Apply transformations on DFT", py::arg("dft"), py::arg("unique_constant_be"), py::arg("binary_fdeps"));
@ -35,5 +35,5 @@ void define_analysis(py::module& m) {
m.def("is_well_formed", &storm::api::isWellFormed<double>, "Check whether DFT is well-formed.", py::arg("dft"), py::arg("check_valid_for_analysis") = true);
m.def("compute_relevant_events", &storm::api::computeRelevantEvents<double>, "Compute relevant event ids from properties and additional relevant names", py::arg("dft"), py::arg("properties"), py::arg("additional_relevant_names") = std::vector<std::string>(), py::arg("allow_dc_relevant") = false);
m.def("compute_relevant_events", &storm::api::computeRelevantEvents<double>, "Compute relevant event ids from properties and additional relevant names", py::arg("dft"), py::arg("properties"), py::arg("additional_relevant_names") = std::vector<std::string>());
}

1
src/logic/formulae.cpp

@ -19,6 +19,7 @@ void define_formulae(py::module& m) {
.value("PRE_SAFETY", storm::logic::ShieldingType::PreSafety)
.value("OPTIMAL_PRE", storm::logic::ShieldingType::OptimalPre)
.value("OPTIMAL_POST", storm::logic::ShieldingType::OptimalPost)
.value("OPTIMAL", storm::logic::ShieldingType::OptimalPost)
;
py::enum_<storm::logic::ShieldComparison>(m, "ShieldComparison")

Loading…
Cancel
Save