The parametric engine works under the assumption that models
are well-defined, i.e., that parameter valuations will lead
to graph-preserving probabilities, etc.
If we encounter a point in a region where these assumptions
don't hold, we now treat this by splitting the region. Regions
with such an improper midpoint that are below the precision
threshold are considered to be undefined and further processing
is aborted.
In particular, this treatment ensures that we never perform
policy iteration on a model with parameter instantiation
that is not well defined / supported. E.g., if there
were negative rewards for such a point, policy iteration
would previously not necessarily terminate.
As well, remove getNumTotalChoices() in favour of the getNumChoices() method
as used in the explicit models.
This will allow using the explicit model checkers prob0/prob1 methods.
This commit refactors the exportToDotFile infrastructure of
the explicit engine to allow the flexible decoration of the
nodes and edges in the DOT file, e.g., for annotating rewards,
marking certain states, etc.
We provide default implementations in explicit.Model for most methods,
only exportTransitionsToDotFile needs to be provided in derived classes
to allow DOT export.
Note that the abstract method in ModelExplicit
abstract void exportTransitionsToDotFile(int i, PrismLog out);
has been removed, which will lead to errors in derived classes
where implementations of this method have been marked with the
@Override annotation.
To fix this, simply replace the signature of your implementation of
void exportTransitionsToDotFile(int i, PrismLog out);
by
void exportTransitionsToDotFile(int i, PrismLog out, Iterable<explicit.graphviz.Decorator> decorators)
(as defined in explicit.Model). You can simply ignore the decorators
parameter at first. Later on, if you want to support decoration,
have a look at the implementations of this method in DTMCExplicit
and MDPExplicit for the proper handling.
git-svn-id: https://www.prismmodelchecker.org/svn/prism/prism/trunk@12115 bbc10eb1-c90d-0410-af57-cb519fbb1720
The getSuccessorsIterator provides an iterator over the set of successors,
i.e., deduplication is sometimes required (e.g. for MDPs).
We introduce here a SuccessorsIterator class that allows to make deduplication
optional, which may have benefits in performance. Additionally, SuccessorsIterator
implements a primitive OfInt iterator, which can avoid auto-boxing.
Subclasses of explicit.Model now have to provide an implementation for getSuccessors(int s),
getSuccessorsIterator(int s) is provided via a default method that automatically
requests deduplication.
Subclasses of explicit.NondetModel now have to provide an implementation for getSuccessors(int s, int).,
getSuccessorsIterator(int s, int i) is provided via a default method that automatically
requests deduplication.
Adapt the existing subclasses of explicit.Model.
Provides additional default methods and removes unneeded specializations in sub-classes:
boolean isSuccessor(int s1, int s2)
boolean allSuccessorsInSet(int s, BitSet set)
boolean someSuccessorsInSet(int s, BitSet set)
boolean allSuccessorsMatch(int s, IntPredicate p)
boolean someSuccessorsMatch(int s, IntPredicate p)
and for NondetModel:
boolean allSuccessorsInSet(int s, int i, BitSet set)
boolean someSuccessorsInSet(int s, int i, BitSet set)
boolean allSuccessorsMatch(int s, int i, IntPredicate p)
boolean someSuccessorsMatch(int s, int i, IntPredicate p)
git-svn-id: https://www.prismmodelchecker.org/svn/prism/prism/trunk@12085 bbc10eb1-c90d-0410-af57-cb519fbb1720
Dave Parker