* src/tgba/tgbaexplicit.cc, src/tgba/tgbaexplicit.hh
(tgba_explicit::create_transition(state*, const state*)): New function. * src/tgbaalgos/randomgraph.cc, src/tgbaalgos/randomgraph.hh: (random_graph): Revamp the algorithm to call rand() less often. * src/tgbatest/randtgba.cc: Add option -0 to easy profiling.
This commit is contained in:
parent
7917841fbe
commit
688587d700
6 changed files with 129 additions and 56 deletions
|
|
@ -65,8 +65,15 @@ namespace spot
|
|||
/// note = {Reprint of Master's thesis}
|
||||
/// }
|
||||
/// \endverbatim
|
||||
/// The only difference is that labels are on transitions, and that
|
||||
/// acceptance conditions are handled. Otherwise the logic is the same.
|
||||
///
|
||||
/// Although the intent is similar, there are some differences with
|
||||
/// between the above published algorithm and this implementation .
|
||||
/// First labels are on transitions, and acceptance conditions are
|
||||
/// generated too. Second, the number of successors of a node is
|
||||
/// chosen in \f$[1,n]\f$ following a normal distribution with mean
|
||||
/// \f$1+(n-1)d\f$ and variance \f$(n-1)d(1-d)\f$. (This is less
|
||||
/// accurate, but faster than considering all possible \a n
|
||||
/// successors one by one.)
|
||||
tgba*
|
||||
random_graph(int n, float d,
|
||||
const ltl::atomic_prop_set* ap, bdd_dict* dict,
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue