dehnert
3ad85ba0e6
fixes and improvements for game-based abstraction
7 years ago
dehnert
c6a5d5a74d
started on refining menu games based on explicit results
7 years ago
dehnert
3f6a8fed92
fixed some issues in qualitative sparse solution of game-based abstraction
7 years ago
dehnert
9665f4fa30
sparse qualitative solving of menu games
7 years ago
dehnert
edbe3b1952
more work on explicit game solving
7 years ago
TimQu
a1c10cac37
filtering reward zero states for MDPs
8 years ago
dehnert
29903bef04
more work on general abstraction refinement framework
8 years ago
dehnert
e557a8e069
started on EC elimination for hybrid engine
8 years ago
dehnert
5fafe835cb
started on some optimizations for conditionals in MDPs
8 years ago
dehnert
9bda631795
symbolic MDP helper respecting solver requirements
8 years ago
dehnert
7c24607427
started on symbolic solver requirements
8 years ago
TimQu
2f49255db6
Improved storage::Scheduler. We can now consider arbitrary finite memory schedulers, potentially employing randomization.
8 years ago
TimQu
dc079b3196
moved a function to graph.h
8 years ago
TimQu
a887fd9224
added possibiliy to perfromProbGreater0A only for a selected set of choices
8 years ago
dehnert
b258f1e52d
some more warnings gone
9 years ago
dehnert
eac2735068
fixed more warnings
9 years ago
dehnert
5b09b91ae1
fixed more warnings
9 years ago
Tom Janson
f390aeadf3
rm broken Dijkstra from graph.cpp
9 years ago
Tom Janson
b71ef02692
comments and fixes (?) to graph.cpp's Dijkstra
This implementation seemed pretty wrong in multiple ways;
I attempted to fix it (a long time ago) (see diff, you'll see what I'm
talking about), then gave up.
Luckily (?) the code is unused, just sitting there, sad and broken.
9 years ago
Tom Janson
6d1608a147
Dijkstra fixed, maybe
TODO: check; improve
Things that aren't going well:
- On the example graph BRP-16-2, all nodes have distance 1. I that
possible??
- The initial states list themself as their own predecessor. That's
bad, because it's simply false (unless there is a self-loop).
Former-commit-id: 06f9a28306
[formerly e7e2385e0d
]
Former-commit-id: 3505011831
10 years ago
dehnert
19fd72cfb6
optimized reuse of qualitative values
9 years ago
dehnert
bf5018b858
post-merge fixes
9 years ago
Sebastian Junges
d246517757
removed src prefix in all includes
9 years ago
Sebastian Junges
e1d201c85e
c++ code compiles again after rename
9 years ago
Sebastian Junges
3a7ee7867b
rename files (does not compile)
9 years ago
dehnert
bc1eff959f
graph algorithms for games now also compute player 2 prob0/1 states and the generated strategies are adapted accordingly
Former-commit-id: da52581328
9 years ago
dehnert
cc550984b3
enabling qualitative answers of game-based model checker
Former-commit-id: b5eca1d671
9 years ago
dehnert
7d50a6b839
graph algorithms for games can now produce player strategies even if they can pick any choice (if requested)
Former-commit-id: 98119f274d
9 years ago
dehnert
1bb116dd1c
some more work on game-based model checker
Former-commit-id: 50399c3d7c
9 years ago
dehnert
9878c1bdc3
fixed some tests that were failing because of (now) proper bottom state computation
Former-commit-id: ecc8dfb065
9 years ago
dehnert
15a4d4757f
added feature to linear equation solver factories to take posession of the matrix to forward it to the solvers
Former-commit-id: ed183f1820
9 years ago
dehnert
82d4164c39
added obeying a state ordering to elimination linear equation solver
Former-commit-id: 5a62842963
9 years ago
Mavo
c3f9d4cfe0
Includes for successful compiling from python
Former-commit-id: 3af33cd206
9 years ago
TimQu
67cc067f35
fixed computeSchedulerProbGreater0E.
Previously, it did not enforce that psiStates are actually reached. For instance, it was ok to chose a probability 1 selfloop.
Former-commit-id: 518a3b33a9
9 years ago
dehnert
0f6e6e4da1
added feature to compute step-bounded until probabilities in parametric models
Former-commit-id: 172e87cb55
10 years ago
sjunges
1e1400d68d
merge
Former-commit-id: eb9efc4bb2
10 years ago
dehnert
dd5af80d5a
work towards easier deployment of other ordering heuristics
Former-commit-id: f48dff3630
10 years ago
dehnert
19029cd905
functional tests compile and run again, yay!
Former-commit-id: 60d3ce16b9
10 years ago
dehnert
8bf0f3c87e
apparently, changing the DD interface implies some other changes as well...
Former-commit-id: c5cedc720f
10 years ago
dehnert
7fa7381047
trying to get the new infrastructure of the DD abstraction layer integrated into the other parts of storm
Former-commit-id: 80a6634565
10 years ago
dehnert
4b4c11048f
fixed bug in prob1E. added functions to retrieve schedulers for prob0E, probGreater0E and prob1E states of MDPs
Former-commit-id: aea5fe773b
10 years ago
sjunges
765748c7cf
graph stuff for other reward models..
Former-commit-id: e7d4d9b95d
10 years ago
sjunges
53dc99deb0
prob01A for non.detmodels added (convenience function)
Former-commit-id: 213c8b1b49
10 years ago
dehnert
781610b05d
extended tests for validity of returned strategies
Former-commit-id: fb6a1c23f0
10 years ago
dehnert
1c42ed792b
fixed some bugs, added some test, added some prob1 algorithm, and did some stuff, you know?
Former-commit-id: 00fa21d1fe
10 years ago
dehnert
972795912a
added some convenience accessor methods in symbolic model/games. added return type for prob01 for games that can also store strategies. added tests for prob0 for games
Former-commit-id: f0a8b156ca
10 years ago
dehnert
6c804732e1
introduced (probably buggy) versions of existsAbstractRepresentative on BDDs and prob0 for games
Former-commit-id: 5e7225fe29
10 years ago
sjunges
35a154f67f
Splitted graph in h and cpp`
Former-commit-id: e22ab7f8eb
10 years ago
dehnert
fbd05cd780
more and more bugfixes
Former-commit-id: 7f1ba98797
10 years ago
sjunges
7a050434d9
bugfixes for NondeterministicModel, improvements for StateActionPair, graph and initialize
Former-commit-id: 4531d784a1
10 years ago