dehnert
86c233f3df
fixed bug in sylvan
Former-commit-id: 3f3a3df83d
9 years ago
hbruintjes
2aaa519dfa
Temproary workaround for clang/gcc problem: Wrap std::function in a lambda for std::sort
Former-commit-id: 4148e73d69
9 years ago
hbruintjes
80430906f0
Use GMP is CLN is not found for RationalNumber
Former-commit-id: ba800d8342
9 years ago
dehnert
fb216e18d1
fixed another issue in expression evaluation
Former-commit-id: e28e1fdf28
9 years ago
dehnert
bb3e008611
added missing include
Former-commit-id: 27295accde
9 years ago
dehnert
fa1ac86ff2
fixed evaluation of floor/ceil in expressions
Former-commit-id: c3e64c552e
9 years ago
dehnert
3476df75e8
finally removed log4cplus and affected code parts
Former-commit-id: 2beea8195e
9 years ago
dehnert
5d9f225f9f
fixed serious bug in model generation from PRISM code; can't believe that didn't cause wrong models until now
Former-commit-id: 8fef881161
9 years ago
dehnert
60bbce0ba1
added two tests for exploration engine
Former-commit-id: 960393b229
9 years ago
dehnert
313edf44e1
added new uniform heuristic and changed probdiff to be the sum instead of product
Former-commit-id: 6fcb9ad80b
9 years ago
dehnert
db3d1df863
added a sh*t ton of debug output, didn't help
Former-commit-id: a1bf88c024
9 years ago
dehnert
f9193325d4
further refactoring of exploration (better name than 'learning') based engine
Former-commit-id: 72622d821d
9 years ago
dehnert
1424d536ca
renamed learning to exploration engine and started on a minor refactoring
Former-commit-id: 0fa973dfe5
9 years ago
dehnert
40b8892f7f
removed debug output
Former-commit-id: bd55256a50
9 years ago
dehnert
07e97e1977
added some statistics and options
Former-commit-id: 1b6a9c20f6
9 years ago
dehnert
6d421a6fbe
learning seems to work find on first larger example
Former-commit-id: 706981a362
9 years ago
dehnert
f1105aac2a
EC-detection appears to work now
Former-commit-id: 0bb1369b3e
9 years ago
dehnert
ce91fa7d5b
started to work on local EC-detection
Former-commit-id: 0f36a1bf78
9 years ago
dehnert
599a3e99c7
minimal probabilities now working (for some test cases)
Former-commit-id: 9f38386531
9 years ago
dehnert
8ea869ef14
changed detection of terminal states a bit
Former-commit-id: a9229fa174
9 years ago
dehnert
c2b287a1e1
more work on learning approach
Former-commit-id: 48aa9ddd2c
9 years ago
dehnert
1405cdfc46
debugged the refactoring a bit
Former-commit-id: 9df3d5d533
9 years ago
dehnert
5092435329
started refactoring learning model checker
Former-commit-id: b9e6015ae4
9 years ago
dehnert
62db38813b
started to refactor learning engine a bit
Former-commit-id: e908301152
9 years ago
dehnert
2f5f439f26
re-added (naive) splitter selection heuristic
Former-commit-id: 5c5166510d
9 years ago
dehnert
38ea181e3d
added tons of debug output. all small test models now show sane results
Former-commit-id: ecfa5ce433
9 years ago
dehnert
e4a5c1d0d6
more work on EC detection (again0
more work on EC detection (again)
Former-commit-id: 1b618f45ec
9 years ago
dehnert
1136ff0d37
fixed a failing test (uninitialized data issue)
Former-commit-id: ca0f456ba2
9 years ago
dehnert
b06419afe0
working towards EC detection
Former-commit-id: 78bbe54f81
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
9f52d9fa97
first working version (for DTMCs only)
Former-commit-id: d3c789596e
9 years ago
dehnert
034cf626a0
more work on learning-based engin
Former-commit-id: bbcf67abd1
9 years ago
dehnert
d802f0d9c6
worked a bit on the learning-based verification of MDPs
Former-commit-id: bc3c0885b2
9 years ago
dehnert
e6ec8d5b60
fixed formula building in some performance tests
Former-commit-id: 1f6c5f67db
9 years ago
dehnert
fd615289e0
outline of learning algorithm
Former-commit-id: d770d1b7dc
9 years ago
dehnert
8ed46ce1b8
started on learning-based verification
Former-commit-id: 24e9d81b15
9 years ago
dehnert
1fb943b658
moved some internal structs from model builder to their own files to make them reusable
Former-commit-id: a354059fe8
9 years ago
dehnert
ca354cffe4
moved preprocessing of PRISM program to utility to make it accessible from learning-based model checker
Former-commit-id: 704dde9ec5
9 years ago
dehnert
dae55eeb29
fixed some bugs and enabled markov automaton model checking from cli
Former-commit-id: 91b689d817
9 years ago
dehnert
7dee6d3da2
started on learning-based MDP model checking
Former-commit-id: 9a901e619b
9 years ago
dehnert
bb7d8ca3c5
added learning as new engine selection in options
Former-commit-id: e00c7ad75d
9 years ago
dehnert
4a19d81133
fixed a few bugs
Former-commit-id: 70d408e653
9 years ago
dehnert
6a99ab9ef9
expectation/variance now handled in formula parser
Former-commit-id: 9dbe09411c
9 years ago
dehnert
51402ec853
removed measure type and only added measure type to reward/time operators
Former-commit-id: 16e19fe349
9 years ago
dehnert
39acf24448
fix for weak bisimulation on CTMCs
Former-commit-id: 4eee2e0997
9 years ago
dehnert
016ab53f42
making the logic formulas better
Former-commit-id: bd5dd26c51
9 years ago
dehnert
5e1e5b55a1
renamed expected time formulas to time formulas
Former-commit-id: 50a11fe446
9 years ago
TimQu
6e8602413e
ModelInstantiator + test
Former-commit-id: f3c9980067
9 years ago
TimQu
69c5ba604e
Helper functions for parametric stuff
Former-commit-id: 288e4de3da
9 years ago
TimQu
a3aededd3a
public access to model ingredients: RewardModel and exitRates
Former-commit-id: b8dbe8576e
9 years ago