Browse Source

started on some optimizations for conditionals in MDPs

tempestpy_adaptions
dehnert 7 years ago
parent
commit
5fafe835cb
  1. 39
      src/storm/modelchecker/prctl/helper/SparseMdpPrctlHelper.cpp
  2. 2
      src/storm/utility/graph.h

39
src/storm/modelchecker/prctl/helper/SparseMdpPrctlHelper.cpp

@ -975,18 +975,29 @@ namespace storm {
}
}
storm::storage::BitVector allStates(fixedTargetStates.size(), true);
// Extend the target states by computing all states that have probability 1 to go to a target state
// under *all* schedulers.
fixedTargetStates = storm::utility::graph::performProb1A(transitionMatrix, transitionMatrix.getRowGroupIndices(), backwardTransitions, allStates, fixedTargetStates);
// We solve the max-case and later adjust the result if the optimization direction was to minimize.
storm::storage::BitVector initialStatesBitVector(transitionMatrix.getRowGroupCount());
initialStatesBitVector.set(initialState);
storm::storage::BitVector allStates(initialStatesBitVector.size(), true);
std::vector<ValueType> conditionProbabilities = std::move(computeUntilProbabilities(OptimizationDirection::Maximize, transitionMatrix, backwardTransitions, allStates, conditionStates, false, false, minMaxLinearEquationSolverFactory).values);
// Extend the condition states by computing all states that have probability 1 to go to a condition state
// under *all* schedulers.
storm::storage::BitVector extendedConditionStates = storm::utility::graph::performProb1A(transitionMatrix, transitionMatrix.getRowGroupIndices(), backwardTransitions, allStates, conditionStates);
STORM_LOG_DEBUG("Computing probabilities to satisfy condition.");
std::vector<ValueType> conditionProbabilities = std::move(computeUntilProbabilities(OptimizationDirection::Maximize, transitionMatrix, backwardTransitions, allStates, extendedConditionStates, false, false, minMaxLinearEquationSolverFactory).values);
// If the conditional probability is undefined for the initial state, we return directly.
if (storm::utility::isZero(conditionProbabilities[initialState])) {
return std::unique_ptr<CheckResult>(new ExplicitQuantitativeCheckResult<ValueType>(initialState, storm::utility::infinity<ValueType>()));
}
STORM_LOG_DEBUG("Computing probabilities to reach target.");
std::vector<ValueType> targetProbabilities = std::move(computeUntilProbabilities(OptimizationDirection::Maximize, transitionMatrix, backwardTransitions, allStates, fixedTargetStates, false, false, minMaxLinearEquationSolverFactory).values);
storm::storage::BitVector statesWithProbabilityGreater0E(transitionMatrix.getRowGroupCount(), true);
@ -999,10 +1010,15 @@ namespace storm {
}
// Determine those states that need to be equipped with a restart mechanism.
storm::storage::BitVector problematicStates = storm::utility::graph::performProb0E(transitionMatrix, transitionMatrix.getRowGroupIndices(), backwardTransitions, allStates, conditionStates | fixedTargetStates);
STORM_LOG_DEBUG("Computing problematic states.");
storm::storage::BitVector pureResetStates = storm::utility::graph::performProb0A(backwardTransitions, allStates, fixedTargetStates);
// FIXME: target | condition as target states here?
storm::storage::BitVector problematicStates = storm::utility::graph::performProb0E(transitionMatrix, transitionMatrix.getRowGroupIndices(), backwardTransitions, allStates, fixedTargetStates);
// Otherwise, we build the transformed MDP.
storm::storage::BitVector relevantStates = storm::utility::graph::getReachableStates(transitionMatrix, initialStatesBitVector, allStates, conditionStates | fixedTargetStates);
storm::storage::BitVector relevantStates = storm::utility::graph::getReachableStates(transitionMatrix, initialStatesBitVector, allStates, extendedConditionStates | fixedTargetStates | pureResetStates);
STORM_LOG_TRACE("Found " << relevantStates.getNumberOfSetBits() << " relevant states for conditional probability computation.");
std::vector<uint_fast64_t> numberOfStatesBeforeRelevantStates = relevantStates.getNumberOfSetBitsBeforeIndices();
storm::storage::sparse::state_type newGoalState = relevantStates.getNumberOfSetBits();
storm::storage::sparse::state_type newStopState = newGoalState + 1;
@ -1014,17 +1030,24 @@ namespace storm {
for (auto state : relevantStates) {
builder.newRowGroup(currentRow);
if (fixedTargetStates.get(state)) {
builder.addNextValue(currentRow, newGoalState, conditionProbabilities[state]);
if (!storm::utility::isZero(conditionProbabilities[state])) {
builder.addNextValue(currentRow, newGoalState, conditionProbabilities[state]);
}
if (!storm::utility::isOne(conditionProbabilities[state])) {
builder.addNextValue(currentRow, newFailState, storm::utility::one<ValueType>() - conditionProbabilities[state]);
}
++currentRow;
} else if (conditionStates.get(state)) {
builder.addNextValue(currentRow, newGoalState, targetProbabilities[state]);
} else if (extendedConditionStates.get(state)) {
if (!storm::utility::isZero(targetProbabilities[state])) {
builder.addNextValue(currentRow, newGoalState, targetProbabilities[state]);
}
if (!storm::utility::isOne(targetProbabilities[state])) {
builder.addNextValue(currentRow, newStopState, storm::utility::one<ValueType>() - targetProbabilities[state]);
}
++currentRow;
} else if (pureResetStates.get(state)) {
builder.addNextValue(currentRow, numberOfStatesBeforeRelevantStates[initialState], storm::utility::one<ValueType>());
++currentRow;
} else {
for (uint_fast64_t row = transitionMatrix.getRowGroupIndices()[state]; row < transitionMatrix.getRowGroupIndices()[state + 1]; ++row) {
for (auto const& successorEntry : transitionMatrix.getRow(row)) {
@ -1051,9 +1074,11 @@ namespace storm {
++currentRow;
// Finally, build the matrix and dispatch the query as a reachability query.
STORM_LOG_DEBUG("Computing conditional probabilties.");
storm::storage::BitVector newGoalStates(newFailState + 1);
newGoalStates.set(newGoalState);
storm::storage::SparseMatrix<ValueType> newTransitionMatrix = builder.build();
STORM_LOG_DEBUG("Transformed model has " << newTransitionMatrix.getRowGroupCount() << " states and " << newTransitionMatrix.getNonzeroEntryCount() << " transitions.");
storm::storage::SparseMatrix<ValueType> newBackwardTransitions = newTransitionMatrix.transpose(true);
std::vector<ValueType> goalProbabilities = std::move(computeUntilProbabilities(OptimizationDirection::Maximize, newTransitionMatrix, newBackwardTransitions, storm::storage::BitVector(newFailState + 1, true), newGoalStates, false, false, minMaxLinearEquationSolverFactory).values);

2
src/storm/utility/graph.h

@ -416,7 +416,7 @@ namespace storm {
storm::storage::BitVector performProb1A(storm::models::sparse::NondeterministicModel<T, RM> const& model, storm::storage::SparseMatrix<T> const& backwardTransitions, storm::storage::BitVector const& phiStates, storm::storage::BitVector const& psiStates);
template <typename T>
storm::storage::BitVector performProb1A( storm::storage::SparseMatrix<T> const& transitionMatrix, std::vector<uint_fast64_t> const& nondeterministicChoiceIndices, storm::storage::SparseMatrix<T> const& backwardTransitions, storm::storage::BitVector const& phiStates, storm::storage::BitVector const& psiStates);
storm::storage::BitVector performProb1A(storm::storage::SparseMatrix<T> const& transitionMatrix, std::vector<uint_fast64_t> const& nondeterministicChoiceIndices, storm::storage::SparseMatrix<T> const& backwardTransitions, storm::storage::BitVector const& phiStates, storm::storage::BitVector const& psiStates);
template <typename T>
std::pair<storm::storage::BitVector, storm::storage::BitVector> performProb01Min(storm::storage::SparseMatrix<T> const& transitionMatrix, std::vector<uint_fast64_t> const& nondeterministicChoiceIndices, storm::storage::SparseMatrix<T> const& backwardTransitions, storm::storage::BitVector const& phiStates, storm::storage::BitVector const& psiStates) ;

Loading…
Cancel
Save