Browse Source

introduced memory structure

tempestpy_adaptions
TimQu 8 years ago
parent
commit
fc97c1fc9d
  1. 6
      src/storm/storage/SparseMatrix.cpp
  2. 96
      src/storm/storage/memorystructure/MemoryStructure.cpp
  3. 65
      src/storm/storage/memorystructure/MemoryStructure.h
  4. 36
      src/storm/storage/memorystructure/MemoryStructureBuilder.cpp
  5. 49
      src/storm/storage/memorystructure/MemoryStructureBuilder.h
  6. 327
      src/storm/storage/memorystructure/SparseModelMemoryProduct.cpp
  7. 67
      src/storm/storage/memorystructure/SparseModelMemoryProduct.h

6
src/storm/storage/SparseMatrix.cpp

@ -14,6 +14,7 @@
#include "storm/storage/BitVector.h"
#include "storm/utility/constants.h"
#include "storm/utility/ConstantsComparator.h"
#include "storm/utility/vector.h"
#include "storm/exceptions/InvalidStateException.h"
#include "storm/exceptions/NotImplementedException.h"
@ -546,10 +547,7 @@ namespace storm {
// If there is no current row grouping, we need to create it.
if (!this->rowGroupIndices) {
STORM_LOG_ASSERT(trivialRowGrouping, "Only trivial row-groupings can be constructed on-the-fly.");
this->rowGroupIndices = std::vector<index_type>(this->getRowCount() + 1);
for (uint_fast64_t group = 0; group <= this->getRowCount(); ++group) {
this->rowGroupIndices.get()[group] = group;
}
this->rowGroupIndices = storm::utility::vector::buildVectorForRange(0, this->getRowGroupCount() + 1);
}
return rowGroupIndices.get();
}

96
src/storm/storage/memorystructure/MemoryStructure.cpp

@ -0,0 +1,96 @@
#include "storm/storage/memorystructure/MemoryStructure.h"
#include "storm/logic/Formulas.h"
#include "storm/utility/macros.h"
#include "storm/storage/memorystructure/SparseModelMemoryProduct.h"
#include "storm/exceptions/InvalidOperationException.h"
namespace storm {
namespace storage {
MemoryStructure::MemoryStructure(TransitionMatrix const& transitionMatrix, storm::models::sparse::StateLabeling const& memoryStateLabeling) : transitions(transitionMatrix), stateLabeling(memoryStateLabeling) {
// intentionally left empty
}
MemoryStructure::MemoryStructure(TransitionMatrix&& transitionMatrix, storm::models::sparse::StateLabeling&& memoryStateLabeling) : transitions(std::move(transitionMatrix)), stateLabeling(std::move(memoryStateLabeling)) {
// intentionally left empty
}
MemoryStructure::TransitionMatrix const& MemoryStructure::getTransitionMatrix() const {
return this->transitions;
}
storm::models::sparse::StateLabeling const& MemoryStructure::getStateLabeling() const {
return stateLabeling;
}
MemoryStructure MemoryStructure::product(MemoryStructure const& rhs) const {
uint_fast64_t lhsNumStates = this->getTransitionMatrix().size();
uint_fast64_t rhsNumStates = rhs.getTransitionMatrix().size();
uint_fast64_t resNumStates = lhsNumStates * rhsNumStates;
// Transition matrix
TransitionMatrix resultTransitions(resNumStates, std::vector<std::shared_ptr<storm::logic::Formula const>>(resNumStates));
uint_fast64_t resState = 0;
for (uint_fast64_t lhsState = 0; lhsState < lhsNumStates; ++lhsState) {
for (uint_fast64_t rhsState = 0; rhsState < rhsNumStates; ++rhsState) {
assert (resState == (lhsState * rhsNumStates) + rhsState);
auto resStateTransitions = resultTransitions[resState];
for (uint_fast64_t lhsTransitionTarget = 0; lhsTransitionTarget < lhsNumStates; ++lhsTransitionTarget) {
auto& lhsTransition = this->getTransitionMatrix()[lhsState][lhsTransitionTarget];
if (lhsTransition) {
for (uint_fast64_t rhsTransitionTarget = 0; rhsTransitionTarget < rhsNumStates; ++rhsTransitionTarget) {
auto& rhsTransition = this->getTransitionMatrix()[rhsState][rhsTransitionTarget];
if (rhsTransition) {
uint_fast64_t resTransitionTarget = (lhsTransitionTarget * rhsNumStates) + rhsTransitionTarget;
resStateTransitions[resTransitionTarget] = std::make_shared<storm::logic::BinaryBooleanStateFormula const>(storm::logic::BinaryBooleanStateFormula::OperatorType::And, lhsTransition,rhsTransition);
}
}
}
}
++resState;
}
}
// State Labels
storm::models::sparse::StateLabeling resultLabeling(resNumStates);
for (std::string lhsLabel : this->getStateLabeling().getLabels()) {
storm::storage::BitVector const& lhsLabeledStates = this->getStateLabeling().getStates(lhsLabel);
storm::storage::BitVector resLabeledStates(resNumStates, false);
for (auto const& lhsState : lhsLabeledStates) {
for (uint_fast64_t rhsState = 0; rhsState < rhsNumStates; ++rhsState) {
resState = (lhsState * rhsNumStates) + rhsState;
resLabeledStates.set(resState, true);
}
}
resultLabeling.addLabel(lhsLabel, std::move(resLabeledStates));
}
for (std::string rhsLabel : rhs.getStateLabeling().getLabels()) {
STORM_LOG_THROW(!resultLabeling.containsLabel(rhsLabel), storm::exceptions::InvalidOperationException, "Failed to build the product of two memory structures: State labelings are not disjoint as both structures contain the label " << rhsLabel << ".");
storm::storage::BitVector const& rhsLabeledStates = rhs.getStateLabeling().getStates(rhsLabel);
storm::storage::BitVector resLabeledStates(resNumStates, false);
for (auto const& rhsState : rhsLabeledStates) {
for (uint_fast64_t lhsState = 0; lhsState < lhsNumStates; ++lhsState) {
resState = (lhsState * rhsNumStates) + rhsState;
resLabeledStates.set(resState, true);
}
}
resultLabeling.addLabel(rhsLabel, std::move(resLabeledStates));
}
return MemoryStructure(std::move(resultTransitions), std::move(resultLabeling));
}
template <typename ValueType>
SparseModelMemoryProduct<ValueType> MemoryStructure::product(storm::models::sparse::Model<ValueType> const& sparseModel) const {
return SparseModelMemoryProduct<ValueType>(sparseModel, *this);
}
template SparseModelMemoryProduct<double> MemoryStructure::product(storm::models::sparse::Model<double> const& sparseModel) const;
template SparseModelMemoryProduct<storm::RationalNumber> MemoryStructure::product(storm::models::sparse::Model<storm::RationalNumber> const& sparseModel) const;
}
}

65
src/storm/storage/memorystructure/MemoryStructure.h

@ -0,0 +1,65 @@
#pragma once
#include <vector>
#include <memory>
#include "storm/logic/Formula.h"
#include "storm/models/sparse/StateLabeling.h"
#include "storm/models/sparse/Model.h"
namespace storm {
namespace storage {
template <typename ValueType>
class SparseModelMemoryProduct;
/*!
* This class represents a (deterministic) memory structure that can be used to encode certain events
* (such as reaching a set of goal states) into the state space of the model.
*/
class MemoryStructure {
public:
typedef std::vector<std::vector<std::shared_ptr<storm::logic::Formula const>>> TransitionMatrix;
/*!
* Creates a memory structure with the given transition matrix and the given memory state labeling.
* The initial state is always the state with index 0.
* The transition matrix is assumed to contain propositional state formulas. The entry
* transitionMatrix[m][n] specifies the set of model states which trigger a transition from memory
* state m to memory state n.
* Transitions are assumed to be deterministic and complete, i.e., the formulas in
* transitionMatrix[m] form a partition of the state space of the considered model.
*
* @param transitionMatrix The transition matrix
* @param memoryStateLabeling A labeling of the memory states to specify, e.g., accepting states
*/
MemoryStructure(TransitionMatrix const& transitionMatrix, storm::models::sparse::StateLabeling const& memoryStateLabeling);
MemoryStructure(TransitionMatrix&& transitionMatrix, storm::models::sparse::StateLabeling&& memoryStateLabeling);
TransitionMatrix const& getTransitionMatrix() const;
storm::models::sparse::StateLabeling const& getStateLabeling() const;
/*!
* Builds the product of this memory structure and the given memory structure.
* The resulting memory structure will have the state labels of both given structures.
* Throws an exception if the state labelings are not disjoint.
*/
MemoryStructure product(MemoryStructure const& rhs) const;
/*!
* Builds the product of this memory structure and the given sparse model.
* An exception is thrown if the state labelings of this memory structure and the given model are not disjoint.
*/
template <typename ValueType>
SparseModelMemoryProduct<ValueType> product(storm::models::sparse::Model<ValueType> const& sparseModel) const;
private:
TransitionMatrix transitions;
storm::models::sparse::StateLabeling stateLabeling;
};
}
}

36
src/storm/storage/memorystructure/MemoryStructureBuilder.cpp

@ -0,0 +1,36 @@
#include "storm/storage/memorystructure/MemoryStructureBuilder.h"
#include "storm/logic/FragmentSpecification.h"
#include "storm/utility/macros.h"
#include "storm/exceptions/InvalidOperationException.h"
namespace storm {
namespace storage {
MemoryStructureBuilder::MemoryStructureBuilder(uint_fast64_t const& numberOfStates) : transitions(numberOfStates, std::vector<std::shared_ptr<storm::logic::Formula const>>(numberOfStates)), stateLabeling(numberOfStates) {
// Intentionally left empty
}
void MemoryStructureBuilder::setTransition(uint_fast64_t const& startState, uint_fast64_t const& goalState, std::shared_ptr<storm::logic::Formula const> formula) {
STORM_LOG_THROW(startState < transitions.size(), storm::exceptions::InvalidOperationException, "Invalid index of start state: " << startState << ". There are only " << transitions.size() << " states in this memory structure.");
STORM_LOG_THROW(goalState < transitions.size(), storm::exceptions::InvalidOperationException, "Invalid index of goal state: " << startState << ". There are only " << transitions.size() << " states in this memory structure.");
STORM_LOG_THROW(formula->isInFragment(storm::logic::propositional()), storm::exceptions::InvalidOperationException, "The formula '" << *formula << "' is not propositional");
transitions[startState][goalState] = formula;
}
void MemoryStructureBuilder::setLabel(uint_fast64_t const& state, std::string const& label) {
STORM_LOG_THROW(state < transitions.size(), storm::exceptions::InvalidOperationException, "Can not add label to state with index " << state << ". There are only " << transitions.size() << " states in this memory structure.");
if (!stateLabeling.containsLabel(label)) {
stateLabeling.addLabel(label);
}
stateLabeling.addLabelToState(label, state);
}
MemoryStructure MemoryStructureBuilder::build() {
return MemoryStructure(std::move(transitions), std::move(stateLabeling));
}
}
}

49
src/storm/storage/memorystructure/MemoryStructureBuilder.h

@ -0,0 +1,49 @@
#pragma once
#include <vector>
#include <memory>
#include "storm/logic/Formula.h"
#include "storm/models/sparse/StateLabeling.h"
#include "storm/storage/memorystructure/MemoryStructure.h"
namespace storm {
namespace storage {
class MemoryStructureBuilder {
public:
/*!
* Initializes a new builder for a memory structure
* @param numberOfStates The number of states the resulting memory structure should have
*/
MemoryStructureBuilder(uint_fast64_t const& numberOfStates);
/*!
* Specifies a transition. The formula should be a propositional formula
*/
void setTransition(uint_fast64_t const& startState, uint_fast64_t const& goalState, std::shared_ptr<storm::logic::Formula const> formula);
/*!
* Sets a label to the given state.
*/
void setLabel(uint_fast64_t const& state, std::string const& label);
/*!
* Builds the memory structure.
* @note Calling this invalidates this builder.
* @note When calling this method, the specified transitions should be deterministic and complete. This is not checked.
*/
MemoryStructure build();
private:
MemoryStructure::TransitionMatrix transitions;
storm::models::sparse::StateLabeling stateLabeling;
};
}
}

327
src/storm/storage/memorystructure/SparseModelMemoryProduct.cpp

@ -0,0 +1,327 @@
#include "storm/storage/memorystructure/SparseModelMemoryProduct.h"
#include <boost/optional.hpp>
#include "storm/models/sparse/Dtmc.h"
#include "storm/models/sparse/Mdp.h"
#include "storm/models/sparse/Ctmc.h"
#include "storm/models/sparse/MarkovAutomaton.h"
#include "storm/modelchecker/propositional/SparsePropositionalModelChecker.h"
#include "storm/modelchecker/results/ExplicitQualitativeCheckResult.h"
#include "storm/utility/constants.h"
#include "storm/utility/macros.h"
#include "storm/exceptions/InvalidOperationException.h"
namespace storm {
namespace storage {
template <typename ValueType>
SparseModelMemoryProduct<ValueType>::SparseModelMemoryProduct(storm::models::sparse::Model<ValueType> const& sparseModel, storm::storage::MemoryStructure const& memoryStructure) : model(sparseModel), memory(memoryStructure) {
// intentionally left empty
}
template <typename ValueType>
std::shared_ptr<storm::models::sparse::Model<ValueType>> SparseModelMemoryProduct<ValueType>::build() {
std::vector<uint_fast64_t> memorySuccessors = computeMemorySuccessors();
storm::storage::BitVector reachableStates = computeReachableStates(memorySuccessors);
// Compute the mapping to the states of the result
uint_fast64_t reachableStateCount = 0;
toResultStateMapping = std::vector<uint_fast64_t> (model.getNumberOfStates() * memory.getTransitionMatrix().size(), std::numeric_limits<uint_fast64_t>::max());
for (auto const& reachableState : reachableStates) {
toResultStateMapping[reachableState] = reachableStateCount;
++reachableStateCount;
}
// Build the model components
storm::storage::SparseMatrix<ValueType> transitionMatrix = model.getTransitionMatrix().hasTrivialRowGrouping() ?
buildDeterministicTransitionMatrix(reachableStates, memorySuccessors) :
buildNondeterministicTransitionMatrix(reachableStates, memorySuccessors);
storm::models::sparse::StateLabeling labeling = buildStateLabeling(transitionMatrix);
std::unordered_map<std::string, storm::models::sparse::StandardRewardModel<ValueType>> rewardModels = buildRewardModels(transitionMatrix, memorySuccessors);
return buildResult(std::move(transitionMatrix), std::move(labeling), std::move(rewardModels));
}
template <typename ValueType>
uint_fast64_t const& SparseModelMemoryProduct<ValueType>::getResultState(uint_fast64_t const& modelState, uint_fast64_t const& memoryState) {
return toResultStateMapping[modelState * memory.getTransitionMatrix().size() + memoryState];
}
template <typename ValueType>
std::vector<uint_fast64_t> SparseModelMemoryProduct<ValueType>::computeMemorySuccessors() const {
uint_fast64_t modelStateCount = model.getNumberOfStates();
uint_fast64_t memoryStateCount = memory.getTransitionMatrix().size();
std::vector<uint_fast64_t> result(modelStateCount * memoryStateCount, std::numeric_limits<uint_fast64_t>::max());
storm::modelchecker::SparsePropositionalModelChecker<storm::models::sparse::Model<ValueType>> mc(model);
for (uint_fast64_t memoryState = 0; memoryState < memoryStateCount; ++memoryState) {
for (uint_fast64_t transitionGoal = 0; transitionGoal < memoryStateCount; ++transitionGoal) {
auto const& transition = memory.getTransitionMatrix()[memoryState][transitionGoal];
if (transition) {
storm::storage::BitVector const& modelStates = mc.check(*transition)->asExplicitQualitativeCheckResult().getTruthValuesVector();
for (auto const& modelState : modelStates) {
result[modelState * memoryStateCount + memoryState] = transitionGoal;
}
}
}
}
return result;
}
template <typename ValueType>
storm::storage::BitVector SparseModelMemoryProduct<ValueType>::computeReachableStates(std::vector<uint_fast64_t> const& memorySuccessors) const {
uint_fast64_t modelStateCount = model.getNumberOfStates();
uint_fast64_t memoryStateCount = memory.getTransitionMatrix().size();
// Explore the reachable states via with DFS.
// A state s on the stack corresponds to the model state (s / memoryStateCount) and memory state (s % memoryStateCount)
storm::storage::BitVector reachableStates(modelStateCount * memoryStateCount, false);
std::vector<uint_fast64_t> stack;
for (auto const& modelInit : model.getInitialStates()) {
// Note: 0 is always the initial state of a memory structure.
uint_fast64_t stateIndex = modelInit * memoryStateCount + memorySuccessors[modelInit * memoryStateCount];
reachableStates.set(stateIndex, true);
stack.push_back(stateIndex);
}
while (!stack.empty()) {
uint_fast64_t stateIndex = stack.back();
stack.pop_back();
uint_fast64_t modelState = stateIndex / memoryStateCount;
uint_fast64_t memoryState = stateIndex % memoryStateCount;
for (auto const& modelTransition : model.getTransitionMatrix().getRowGroup(modelState)) {
if (!storm::utility::isZero(modelTransition.getValue())) {
uint_fast64_t successorModelState = modelTransition.getColumn();
uint_fast64_t successorMemoryState = memorySuccessors[successorModelState * memoryStateCount + memoryState];
uint_fast64_t successorStateIndex = successorModelState * memoryStateCount + successorMemoryState;
if (!reachableStates.get(successorStateIndex)) {
reachableStates.set(successorStateIndex, true);
stack.push_back(successorStateIndex);
}
}
}
}
return reachableStates;
}
template <typename ValueType>
storm::storage::SparseMatrix<ValueType> SparseModelMemoryProduct<ValueType>::buildDeterministicTransitionMatrix(storm::storage::BitVector const& reachableStates, std::vector<uint_fast64_t> const& memorySuccessors) const {
uint_fast64_t memoryStateCount = memory.getTransitionMatrix().size();
uint_fast64_t numResStates = reachableStates.getNumberOfSetBits();
uint_fast64_t numResTransitions = 0;
for (auto const& stateIndex : reachableStates) {
numResTransitions += model.getTransitionMatrix().getRow(stateIndex / memoryStateCount).getNumberOfEntries();
}
storm::storage::SparseMatrixBuilder<ValueType> builder(numResStates, numResStates, numResTransitions, true);
uint_fast64_t currentRow = 0;
for (auto const& stateIndex : reachableStates) {
uint_fast64_t modelState = stateIndex / memoryStateCount;
uint_fast64_t memoryState = stateIndex % memoryStateCount;
for (auto const& entry : model.getTransitionMatrix().getRow(modelState)) {
uint_fast64_t const& successorMemoryState = memorySuccessors[entry.getColumn() * memoryStateCount + memoryState];
uint_fast64_t transitionTargetStateIndex = entry.getColumn() * memoryStateCount + successorMemoryState;
builder.addNextValue(currentRow, toResultStateMapping[transitionTargetStateIndex], entry.getValue());
}
++currentRow;
}
return builder.build();
}
template <typename ValueType>
storm::storage::SparseMatrix<ValueType> SparseModelMemoryProduct<ValueType>::buildNondeterministicTransitionMatrix(storm::storage::BitVector const& reachableStates, std::vector<uint_fast64_t> const& memorySuccessors) const {
uint_fast64_t memoryStateCount = memory.getTransitionMatrix().size();
uint_fast64_t numResStates = reachableStates.getNumberOfSetBits();
uint_fast64_t numResChoices = 0;
uint_fast64_t numResTransitions = 0;
for (auto const& stateIndex : reachableStates) {
uint_fast64_t modelState = stateIndex / memoryStateCount;
for (uint_fast64_t modelRow = model.getTransitionMatrix().getRowGroupIndices()[modelState]; modelRow < model.getTransitionMatrix().getRowGroupIndices()[modelState + 1]; ++modelRow) {
++numResChoices;
numResTransitions += model.getTransitionMatrix().getRow(modelRow).getNumberOfEntries();
}
}
storm::storage::SparseMatrixBuilder<ValueType> builder(numResChoices, numResStates, numResTransitions, true, true, numResStates);
uint_fast64_t currentRow = 0;
for (auto const& stateIndex : reachableStates) {
uint_fast64_t modelState = stateIndex / memoryStateCount;
uint_fast64_t memoryState = stateIndex % memoryStateCount;
builder.newRowGroup(currentRow);
for (uint_fast64_t modelRow = model.getTransitionMatrix().getRowGroupIndices()[modelState]; modelRow < model.getTransitionMatrix().getRowGroupIndices()[modelState + 1]; ++modelRow) {
for (auto const& entry : model.getTransitionMatrix().getRow(modelRow)) {
uint_fast64_t const& successorMemoryState = memorySuccessors[entry.getColumn() * memoryStateCount + memoryState];
uint_fast64_t transitionTargetStateIndex = entry.getColumn() * memoryStateCount + successorMemoryState;
builder.addNextValue(currentRow, toResultStateMapping[transitionTargetStateIndex], entry.getValue());
}
++currentRow;
}
}
return builder.build();
}
template <typename ValueType>
storm::models::sparse::StateLabeling SparseModelMemoryProduct<ValueType>::buildStateLabeling(storm::storage::SparseMatrix<ValueType> const& resultTransitionMatrix) const {
uint_fast64_t modelStateCount = model.getNumberOfStates();
uint_fast64_t memoryStateCount = memory.getTransitionMatrix().size();
uint_fast64_t numResStates = resultTransitionMatrix.getRowGroupCount();
storm::models::sparse::StateLabeling resultLabeling(numResStates);
for (std::string modelLabel : model.getStateLabeling().getLabels()) {
storm::storage::BitVector const& modelLabeledStates = model.getStateLabeling().getStates(modelLabel);
storm::storage::BitVector resLabeledStates(numResStates, false);
for (auto const& modelState : modelLabeledStates) {
for (uint_fast64_t memoryState = 0; memoryState < memoryStateCount; ++memoryState) {
uint_fast64_t const& resState = toResultStateMapping[modelState * memoryStateCount + memoryState];
// Check if the state exists in the result (i.e. if it is reachable)
if (resState < numResStates) {
resLabeledStates.set(resState, true);
}
}
}
resultLabeling.addLabel(modelLabel, std::move(resLabeledStates));
}
for (std::string memoryLabel : memory.getStateLabeling().getLabels()) {
STORM_LOG_THROW(!resultLabeling.containsLabel(memoryLabel), storm::exceptions::InvalidOperationException, "Failed to build the product of model and memory structure: State labelings are not disjoint as both structures contain the label " << memoryLabel << ".");
storm::storage::BitVector const& memoryLabeledStates = memory.getStateLabeling().getStates(memoryLabel);
storm::storage::BitVector resLabeledStates(numResStates, false);
for (auto const& memoryState : memoryLabeledStates) {
for (uint_fast64_t modelState = 0; modelState < modelStateCount; ++modelState) {
uint_fast64_t const& resState = toResultStateMapping[modelState * memoryStateCount + memoryState];
// Check if the state exists in the result (i.e. if it is reachable)
if (resState < numResStates) {
resLabeledStates.set(resState, true);
}
}
}
resultLabeling.addLabel(memoryLabel, std::move(resLabeledStates));
}
return resultLabeling;
}
template <typename ValueType>
std::unordered_map<std::string, storm::models::sparse::StandardRewardModel<ValueType>> SparseModelMemoryProduct<ValueType>::buildRewardModels(storm::storage::SparseMatrix<ValueType> const& resultTransitionMatrix, std::vector<uint_fast64_t> const& memorySuccessors) const {
std::unordered_map<std::string, storm::models::sparse::StandardRewardModel<ValueType>> result;
uint_fast64_t memoryStateCount = memory.getTransitionMatrix().size();
uint_fast64_t numResStates = resultTransitionMatrix.getRowGroupCount();
for (auto const& rewardModel : model.getRewardModels()) {
boost::optional<std::vector<ValueType>> stateRewards;
if (rewardModel.second.hasStateRewards()) {
stateRewards = std::vector<ValueType>(numResStates, storm::utility::zero<ValueType>());
uint_fast64_t modelState = 0;
for (auto const& modelStateReward : rewardModel.second.getStateRewardVector()) {
if (!storm::utility::isZero(modelStateReward)) {
for (uint_fast64_t memoryState = 0; memoryState < memoryStateCount; ++memoryState) {
uint_fast64_t const& resState = toResultStateMapping[modelState * memoryStateCount + memoryState];
// Check if the state exists in the result (i.e. if it is reachable)
if (resState < numResStates) {
stateRewards.get()[resState] = modelStateReward;
}
}
}
++modelState;
}
}
boost::optional<std::vector<ValueType>> stateActionRewards;
if (rewardModel.second.hasStateActionRewards()) {
stateActionRewards = std::vector<ValueType>(resultTransitionMatrix.getRowCount(), storm::utility::zero<ValueType>());
uint_fast64_t modelState = 0;
uint_fast64_t modelRow = 0;
for (auto const& modelStateActionReward : rewardModel.second.getStateActionRewardVector()) {
if (!storm::utility::isZero(modelStateActionReward)) {
while (modelRow >= model.getTransitionMatrix().getRowGroupIndices()[modelState + 1]) {
++modelState;
}
uint_fast64_t rowOffset = modelRow - model.getTransitionMatrix().getRowGroupIndices()[modelState];
for (uint_fast64_t memoryState = 0; memoryState < memoryStateCount; ++memoryState) {
uint_fast64_t const& resState = toResultStateMapping[modelState * memoryStateCount + memoryState];
// Check if the state exists in the result (i.e. if it is reachable)
if (resState < numResStates) {
stateActionRewards.get()[resultTransitionMatrix.getRowGroupIndices()[resState] + rowOffset] = modelStateActionReward;
}
}
}
++modelRow;
}
}
boost::optional<storm::storage::SparseMatrix<ValueType>> transitionRewards;
if (rewardModel.second.hasTransitionRewards()) {
storm::storage::SparseMatrixBuilder<ValueType> builder(resultTransitionMatrix.getRowCount(), resultTransitionMatrix.getColumnCount());
uint_fast64_t stateIndex = 0;
for (auto const& resState : toResultStateMapping) {
if (resState < numResStates) {
uint_fast64_t modelState = stateIndex / memoryStateCount;
uint_fast64_t memoryState = stateIndex % memoryStateCount;
uint_fast64_t rowGroupSize = resultTransitionMatrix.getRowGroupSize(resState);
for (uint_fast64_t rowOffset = 0; rowOffset < rowGroupSize; ++rowOffset) {
uint_fast64_t resRowIndex = resultTransitionMatrix.getRowGroupIndices()[resState] + rowOffset;
uint_fast64_t modelRowIndex = model.getTransitionMatrix().getRowGroupIndices()[modelState] + rowOffset;
for (auto const& entry : rewardModel.second.getTransitionRewardMatrix().getRow(modelRowIndex)) {
uint_fast64_t const& successorMemoryState = memorySuccessors[entry.getColumn() * memoryStateCount + memoryState];
uint_fast64_t transitionTargetStateIndex = entry.getColumn() * memoryStateCount + successorMemoryState;
builder.addNextValue(resRowIndex, toResultStateMapping[transitionTargetStateIndex], entry.getValue());
}
}
}
++stateIndex;
}
transitionRewards = builder.build();
}
result.insert(std::make_pair(rewardModel.first, storm::models::sparse::StandardRewardModel<ValueType>(std::move(stateRewards), std::move(stateActionRewards), std::move(transitionRewards))));
}
return result;
}
template <typename ValueType>
std::shared_ptr<storm::models::sparse::Model<ValueType>> SparseModelMemoryProduct<ValueType>::buildResult(storm::storage::SparseMatrix<ValueType>&& matrix, storm::models::sparse::StateLabeling&& labeling, std::unordered_map<std::string, storm::models::sparse::StandardRewardModel<ValueType>>&& rewardModels) const {
switch (model.getType()) {
case storm::models::ModelType::Dtmc:
return std::make_shared<storm::models::sparse::Dtmc<ValueType>> (std::move(matrix), std::move(labeling), std::move(rewardModels));
case storm::models::ModelType::Mdp:
return std::make_shared<storm::models::sparse::Mdp<ValueType>> (std::move(matrix), std::move(labeling), std::move(rewardModels));
case storm::models::ModelType::Ctmc:
return std::make_shared<storm::models::sparse::Ctmc<ValueType>> (std::move(matrix), std::move(labeling), std::move(rewardModels));
case storm::models::ModelType::MarkovAutomaton:
{
// We also need to translate the exit rates and the Markovian states
uint_fast64_t numResStates = matrix.getRowGroupCount();
uint_fast64_t memoryStateCount = memory.getTransitionMatrix().size();
std::vector<ValueType> resultExitRates;
resultExitRates.reserve(matrix.getRowGroupCount());
storm::storage::BitVector resultMarkovianStates(numResStates, false);
auto const& modelExitRates = dynamic_cast<storm::models::sparse::MarkovAutomaton<ValueType> const&>(model).getExitRates();
auto const& modelMarkovianStates = dynamic_cast<storm::models::sparse::MarkovAutomaton<ValueType> const&>(model).getMarkovianStates();
uint_fast64_t stateIndex = 0;
for (auto const& resState : toResultStateMapping) {
if (resState < numResStates) {
assert(resState == resultExitRates.size());
uint_fast64_t modelState = stateIndex / memoryStateCount;
resultExitRates.push_back(modelExitRates[modelState]);
if (modelMarkovianStates.get(modelState)) {
resultMarkovianStates.set(resState, true);
}
}
}
++stateIndex;
return std::make_shared<storm::models::sparse::MarkovAutomaton<ValueType>> (std::move(matrix), std::move(labeling), std::move(resultMarkovianStates), std::move(resultExitRates), true, std::move(rewardModels));
}
}
}
template class SparseModelMemoryProduct<double>;
template class SparseModelMemoryProduct<storm::RationalNumber>;
template class SparseModelMemoryProduct<storm::RationalFunction>;
}
}

67
src/storm/storage/memorystructure/SparseModelMemoryProduct.h

@ -0,0 +1,67 @@
#pragma once
#include <memory>
#include <string>
#include <unordered_map>
#include "storm/storage/BitVector.h"
#include "storm/storage/memorystructure/MemoryStructure.h"
#include "storm/models/sparse/Model.h"
#include "storm/models/sparse/StandardRewardModel.h"
namespace storm {
namespace storage {
/*!
* This class builds the product of the given sparse model and the given memory structure.
* This is similar to the well-known product of a model with a deterministic rabin automaton.
* Note: we already do one memory-structure-step for the initial state, i.e., if s is an initial state of
* the given model and s satisfies memoryStructure.getTransitionMatrix[0][n], then (s,n) will be the corresponding
* initial state of the resulting model.
*
* The states of the resulting sparse model will have the original state labels plus the labels of this
* memory structure.
* An exception is thrown if the state labelings are not disjoint.
*/
template <typename ValueType>
class SparseModelMemoryProduct {
public:
SparseModelMemoryProduct(storm::models::sparse::Model<ValueType> const& sparseModel, storm::storage::MemoryStructure const& memoryStructure);
// Invokes the building of the product
std::shared_ptr<storm::models::sparse::Model<ValueType>> build();
// Retrieves the state of the resulting model that represents the given memory and model state
uint_fast64_t const& getResultState(uint_fast64_t const& modelState, uint_fast64_t const& memoryState);
private:
// Computes for each pair of memory and model state the successor memory state
// The resulting vector maps (modelState * memoryStateCount) + memoryState to the corresponding successor state of the memory structure
std::vector<uint_fast64_t> computeMemorySuccessors() const;
// Computes the reachable states of the resulting model
storm::storage::BitVector computeReachableStates(std::vector<uint_fast64_t> const& memorySuccessors) const;
// Methods that build the model components
storm::storage::SparseMatrix<ValueType> buildDeterministicTransitionMatrix(storm::storage::BitVector const& reachableStates, std::vector<uint_fast64_t> const& memorySuccessors) const;
storm::storage::SparseMatrix<ValueType> buildNondeterministicTransitionMatrix(storm::storage::BitVector const& reachableStates, std::vector<uint_fast64_t> const& memorySuccessors) const;
storm::models::sparse::StateLabeling buildStateLabeling(storm::storage::SparseMatrix<ValueType> const& resultTransitionMatrix) const;
std::unordered_map<std::string, storm::models::sparse::StandardRewardModel<ValueType>> buildRewardModels(storm::storage::SparseMatrix<ValueType> const& resultTransitionMatrix, std::vector<uint_fast64_t> const& memorySuccessors) const;
// Builds the resulting model
std::shared_ptr<storm::models::sparse::Model<ValueType>> buildResult(storm::storage::SparseMatrix<ValueType>&& matrix, storm::models::sparse::StateLabeling&& labeling, std::unordered_map<std::string, storm::models::sparse::StandardRewardModel<ValueType>>&& rewardModels) const;
// Maps (modelState * memoryStateCount) + memoryState to the state in the result that represents (memoryState,modelState)
std::vector<uint_fast64_t> toResultStateMapping;
storm::models::sparse::Model<ValueType> const& model;
storm::storage::MemoryStructure const& memory;
};
}
}
Loading…
Cancel
Save