def __search(sync_net, ini, fin, cost_function, skip, ret_tuple_as_trans_desc=False,
max_align_time_trace=sys.maxsize):
start_time = time.time()
decorate_transitions_prepostset(sync_net)
decorate_places_preset_trans(sync_net)
closed = set()
ini_state = utils.DijkstraSearchTuple(0, ini, None, None, 0)
open_set = [ini_state]
heapq.heapify(open_set)
visited = 0
queued = 0
traversed = 0
trans_empty_preset = set(t for t in sync_net.transitions if len(t.in_arcs) == 0)
while not len(open_set) == 0:
if (time.time() - start_time) > max_align_time_trace:
return None
curr = heapq.heappop(open_set)
current_marking = curr.m
already_closed = current_marking in closed
if already_closed:
continue
if current_marking == fin:
# from pympler.asizeof import asizeof
# from pm4py.util import measurements
# measurements.Measurements.ALIGN_TIME.append(asizeof(open_set))
return utils.__reconstruct_alignment(curr, visited, queued, traversed,
ret_tuple_as_trans_desc=ret_tuple_as_trans_desc)
closed.add(current_marking)
visited += 1
enabled_trans = copy(trans_empty_preset)
for p in current_marking:
for t in p.ass_trans:
if t.sub_marking <= current_marking:
enabled_trans.add(t)
trans_to_visit_with_cost = [(t, cost_function[t]) for t in enabled_trans if not (
t is not None and utils.__is_log_move(t, skip) and utils.__is_model_move(t, skip))]
for t, cost in trans_to_visit_with_cost:
traversed += 1
new_marking = utils.add_markings(current_marking, t.add_marking)
if new_marking in closed:
continue
queued += 1
tp = utils.DijkstraSearchTuple(curr.g + cost, new_marking, curr, t, curr.l + 1)
heapq.heappush(open_set, tp)
def __build_entities(self):
"""
Builds entities useful to define the marking equation
"""
self.Aeq = None
self.__build_encodings()
if self.full_bootstrap_required:
petri_utils.decorate_transitions_prepostset(self.sync_net)
petri_utils.decorate_places_preset_trans(self.sync_net)
def __build_entities(self):
"""
Builds entities useful to define the marking equation
"""
transitions = self.incidence_matrix.transitions
self.inv_indices = {y: x for x, y in transitions.items()}
self.inv_indices = [
self.inv_indices[i] for i in range(len(self.inv_indices))
]
self.ini_vec = np.matrix(self.incidence_matrix.encode_marking(
self.ini)).transpose()
self.fin_vec = np.matrix(self.incidence_matrix.encode_marking(
self.fin)).transpose()
if self.full_bootstrap_required:
petri_utils.decorate_transitions_prepostset(self.net)
petri_utils.decorate_places_preset_trans(self.net)
def __search(net, ini, fin, variants, exponent=2, marking_limit=1000):
'''
This function is the A* algorithm for multi-alignments.
'''
decorate_transitions_prepostset(net)
decorate_places_preset_trans(net)
mymemory = {}
closed = {}
ini_state = utils.DijkstraSearchTupleForAntiAndMulti(0, ini, [])
open_set = [ini_state]
heapq.heapify(open_set)
visited = 0
queued = 0
traversed = 0
best, sizeAA = None, None
distanceTime = 0
trans_empty_preset = set(t for t in net.transitions if len(t.in_arcs) == 0)
def costfunction(t, curr_ma, withFrac=True):
'''
Compute the maximal distance of the current multi-alignment + t and the variants.
:param t: transition that we want to add in the current multi-alignment
:param curr_ma: current prefix of multi-alignment
:param withFrac: the fraction that prevents the best suffice, this parameter is false at the end of the algorithm.
'''
if t is None:
str_aa = [a.label for a in curr_ma]
elif len(curr_ma) == 0:
str_aa = [t.label]
else:
str_aa = [a.label for a in curr_ma] + [t.label]
all = []
for v in variants:
if str(v + str_aa) not in mymemory.keys() or not withFrac:
size_of_alignment, distance = discountedEditDistance(
str_aa, v, exponent)
if withFrac:
mymemory[str(v + str_aa)] = distance - (
exponent**(-len(str_aa) + 1) - exponent**
(-(len(str_aa) + len(v)))) / (exponent - 1)
else:
mymemory[str(v + str_aa)] = distance
all.append(mymemory[str(v + str_aa)])
return max(all)
# ----------------------------
# a* algorithm starts here
while not len(open_set) == 0:
curr = heapq.heappop(open_set)
if best and best.g < curr.g:
continue
if curr.m == fin:
curr.g = costfunction(None, curr.r, withFrac=False)
# in case there only one run
if not best and (len(heapq.nsmallest(1, open_set)) == 0
or heapq.nsmallest(1, open_set)[0].g > curr.g):
return {
'multi-alignment': curr.r,
'cost': curr.g,
'visited_states': visited,
'queued_states': queued,
'traversed_arcs': traversed,
"max_distance_to_log": getMaxDist(curr.r, variants)
}
# other runs might be interesting
elif not best or (best and best.g > curr.g):
best = curr
continue
closed[curr.m] = 1 if (
curr.m) not in closed.keys() else closed[curr.m] + 1
visited += 1
possible_enabling_transitions = copy(trans_empty_preset)
for p in curr.m:
for t in p.ass_trans:
possible_enabling_transitions.add(t)
enabled_trans = [
t for t in possible_enabling_transitions if t.sub_marking <= curr.m
]
trans_to_visit_with_cost = [(t, costfunction(t, curr.r, withFrac=True))
for t in enabled_trans if t is not None]
for t, cost in trans_to_visit_with_cost:
traversed += 1
new_marking = utils.add_markings(curr.m, t.add_marking)
if ((new_marking) in closed.keys()
and closed[new_marking] > marking_limit) or cost == curr.g:
continue
queued += 1
tp = utils.DijkstraSearchTupleForAntiAndMulti(
cost, new_marking, curr.r + [t])
heapq.heappush(open_set, tp)
return {
'multi-alignment': best.r,
'cost': best.g,
'visited_states': visited,
'queued_states': queued,
'traversed_arcs': traversed,
"max_distance_to_log": getMaxDist(best.r, variants)
}
def __search(sync_net, ini, fin, stop, cost_function, skip):
decorate_transitions_prepostset(sync_net)
decorate_places_preset_trans(sync_net)
incidence_matrix = construct(sync_net)
ini_vec, fin_vec, cost_vec = utils.__vectorize_initial_final_cost(
incidence_matrix, ini, fin, cost_function)
closed = set()
ini_state = utils.SearchTuple(0, 0, 0, ini, None, None, None, True)
open_set = [ini_state]
heapq.heapify(open_set)
visited = 0
queued = 0
traversed = 0
# return all the prefix markings of the optimal alignments as set
ret_markings = None
# keep track of the optimal cost of an alignment (to trim search when needed)
optimal_cost = None
while not len(open_set) == 0:
curr = heapq.heappop(open_set)
current_marking = curr.m
# trim alignments when we already reached an optimal alignment and the
# current cost is greater than the optimal cost
if optimal_cost is not None and curr.f > optimal_cost:
break
already_closed = current_marking in closed
if already_closed:
continue
if stop <= current_marking:
# add the current marking to the set
# of returned markings
if ret_markings is None:
ret_markings = set()
ret_markings.add(current_marking)
# close the marking
closed.add(current_marking)
# set the optimal cost
optimal_cost = curr.f
continue
closed.add(current_marking)
visited += 1
enabled_trans = set()
for p in current_marking:
for t in p.ass_trans:
if t.sub_marking <= current_marking:
enabled_trans.add(t)
trans_to_visit_with_cost = [
(t, cost_function[t]) for t in enabled_trans
if not (t is None or utils.__is_log_move(t, skip) or (
utils.__is_model_move(t, skip) and not t.label[1] is None))
]
for t, cost in trans_to_visit_with_cost:
traversed += 1
new_marking = utils.add_markings(current_marking, t.add_marking)
if new_marking in closed:
continue
g = curr.g + cost
queued += 1
new_f = g
tp = utils.SearchTuple(new_f, g, 0, new_marking, curr, t, None,
True)
heapq.heappush(open_set, tp)
return ret_markings
def __search(sync_net, ini, fin, cost_function, skip, ret_tuple_as_trans_desc=False,
max_align_time_trace=sys.maxsize):
start_time = time.time()
decorate_transitions_prepostset(sync_net)
decorate_places_preset_trans(sync_net)
incidence_matrix = inc_mat_construct(sync_net)
ini_vec, fin_vec, cost_vec = utils.__vectorize_initial_final_cost(incidence_matrix, ini, fin, cost_function)
closed = set()
a_matrix = np.asmatrix(incidence_matrix.a_matrix).astype(np.float64)
g_matrix = -np.eye(len(sync_net.transitions))
h_cvx = np.matrix(np.zeros(len(sync_net.transitions))).transpose()
cost_vec = [x * 1.0 for x in cost_vec]
use_cvxopt = False
if lp_solver.DEFAULT_LP_SOLVER_VARIANT == lp_solver.CVXOPT_SOLVER_CUSTOM_ALIGN or lp_solver.DEFAULT_LP_SOLVER_VARIANT == lp_solver.CVXOPT_SOLVER_CUSTOM_ALIGN_ILP:
use_cvxopt = True
if use_cvxopt:
# not available in the latest version of PM4Py
from cvxopt import matrix
a_matrix = matrix(a_matrix)
g_matrix = matrix(g_matrix)
h_cvx = matrix(h_cvx)
cost_vec = matrix(cost_vec)
h, x = utils.__compute_exact_heuristic_new_version(sync_net, a_matrix, h_cvx, g_matrix, cost_vec, incidence_matrix,
ini,
fin_vec, lp_solver.DEFAULT_LP_SOLVER_VARIANT,
use_cvxopt=use_cvxopt)
ini_state = utils.SearchTuple(0 + h, 0, h, ini, None, None, x, True)
open_set = [ini_state]
heapq.heapify(open_set)
visited = 0
queued = 0
traversed = 0
lp_solved = 1
trans_empty_preset = set(t for t in sync_net.transitions if len(t.in_arcs) == 0)
while not len(open_set) == 0:
if (time.time() - start_time) > max_align_time_trace:
return None
curr = heapq.heappop(open_set)
current_marking = curr.m
while not curr.trust:
if (time.time() - start_time) > max_align_time_trace:
return None
already_closed = current_marking in closed
if already_closed:
curr = heapq.heappop(open_set)
current_marking = curr.m
continue
h, x = utils.__compute_exact_heuristic_new_version(sync_net, a_matrix, h_cvx, g_matrix, cost_vec,
incidence_matrix, curr.m,
fin_vec, lp_solver.DEFAULT_LP_SOLVER_VARIANT,
use_cvxopt=use_cvxopt)
lp_solved += 1
# 11/10/19: shall not a state for which we compute the exact heuristics be
# by nature a trusted solution?
tp = utils.SearchTuple(curr.g + h, curr.g, h, curr.m, curr.p, curr.t, x, True)
# 11/10/2019 (optimization ZA) heappushpop is slightly more efficient than pushing
# and popping separately
curr = heapq.heappushpop(open_set, tp)
current_marking = curr.m
# max allowed heuristics value (27/10/2019, due to the numerical instability of some of our solvers)
if curr.h > lp_solver.MAX_ALLOWED_HEURISTICS:
continue
# 12/10/2019: do it again, since the marking could be changed
already_closed = current_marking in closed
if already_closed:
continue
# 12/10/2019: the current marking can be equal to the final marking only if the heuristics
# (underestimation of the remaining cost) is 0. Low-hanging fruits
if curr.h < 0.01:
if current_marking == fin:
return utils.__reconstruct_alignment(curr, visited, queued, traversed,
ret_tuple_as_trans_desc=ret_tuple_as_trans_desc,
lp_solved=lp_solved)
closed.add(current_marking)
visited += 1
enabled_trans = copy(trans_empty_preset)
for p in current_marking:
for t in p.ass_trans:
if t.sub_marking <= current_marking:
enabled_trans.add(t)
trans_to_visit_with_cost = [(t, cost_function[t]) for t in enabled_trans if not (
t is not None and utils.__is_log_move(t, skip) and utils.__is_model_move(t, skip))]
for t, cost in trans_to_visit_with_cost:
traversed += 1
new_marking = utils.add_markings(current_marking, t.add_marking)
if new_marking in closed:
continue
g = curr.g + cost
queued += 1
h, x = utils.__derive_heuristic(incidence_matrix, cost_vec, curr.x, t, curr.h)
trustable = utils.__trust_solution(x)
new_f = g + h
tp = utils.SearchTuple(new_f, g, h, new_marking, curr, t, x, trustable)
heapq.heappush(open_set, tp)
def __search(sync_net, ini, fin, cost_function, skip, ret_tuple_as_trans_desc=False,
max_align_time_trace=sys.maxsize):
start_time = time.time()
decorate_transitions_prepostset(sync_net)
decorate_places_preset_trans(sync_net)
incidence_matrix = inc_mat_construct(sync_net)
ini_vec, fin_vec, cost_vec = utils.__vectorize_initial_final_cost(incidence_matrix, ini, fin, cost_function)
closed = set()
heu_dict = {}
heu_max_ind_dict = {}
mtcgt_dict = {}
parameters = {}
parameters[marking_equation.Parameters.FULL_BOOTSTRAP_REQUIRED] = False
parameters[marking_equation.Parameters.INCIDENCE_MATRIX] = incidence_matrix
parameters[marking_equation.Parameters.COSTS] = cost_function
visited = 0
queued = 0
traversed = 0
me = marking_equation.build(sync_net, ini, fin, parameters=parameters)
h, x = me.solve()
lp_solved = 1
# try to see if the firing sequence is already fine
firing_sequence, reach_fm, explained_events = me.get_firing_sequence(x)
if reach_fm:
return __reconstruct_alignment(firing_sequence, h, visited, queued, traversed,
ret_tuple_as_trans_desc=ret_tuple_as_trans_desc, lp_solved=lp_solved)
mm, index = __get_model_marking_and_index(ini)
__update_heu_dict(heu_dict, heu_max_ind_dict, mm, index, h, x, firing_sequence, incidence_matrix, cost_vec)
ini_state = utils.TweakedSearchTuple(0 + h, 0, h, ini, None, None, x, True, False)
open_set = [ini_state]
heapq.heapify(open_set)
trans_empty_preset = set(t for t in sync_net.transitions if len(t.in_arcs) == 0)
while not len(open_set) == 0:
if (time.time() - start_time) > max_align_time_trace:
return None
curr = heapq.heappop(open_set)
current_marking = curr.m
while not curr.trust:
if (time.time() - start_time) > max_align_time_trace:
return None
already_closed = current_marking in closed
if already_closed:
curr = heapq.heappop(open_set)
current_marking = curr.m
continue
if curr.t not in mtcgt_dict:
lp_solved += 1
mtcgt = __min_total_cost_given_trans(me, ini, incidence_matrix, curr.t)
mtcgt_dict[curr.t] = mtcgt
else:
mtcgt = mtcgt_dict[curr.t]
h1 = max(mtcgt - curr.g, 0)
if h1 > curr.h:
tp = utils.TweakedSearchTuple(curr.g + h1, curr.g, h1, curr.m, curr.p, curr.t, curr.x, False, False)
curr = heapq.heappushpop(open_set, tp)
current_marking = curr.m
continue
mm, index = __get_model_marking_and_index(curr.m)
h2, x2, trust2 = __get_heu_from_dict(heu_dict, heu_max_ind_dict, mm, index)
if h2 is not None and h2 > curr.h:
tp = utils.TweakedSearchTuple(curr.g + h2, curr.g, h2, curr.m, curr.p, curr.t, x2, trust2, False)
curr = heapq.heappushpop(open_set, tp)
current_marking = curr.m
continue
me.change_ini_vec(curr.m)
h, x = me.solve()
__update_heu_dict_specific_point(heu_dict, heu_max_ind_dict, mm, index, h, x)
lp_solved += 1
tp = utils.TweakedSearchTuple(curr.g + h, curr.g, h, curr.m, curr.p, curr.t, x, True, True)
curr = heapq.heappushpop(open_set, tp)
current_marking = curr.m
already_closed = current_marking in closed
if already_closed:
continue
if curr.h < 0.01:
if current_marking == fin:
trans_list = __transitions_list_from_state(curr)
return __reconstruct_alignment(trans_list, curr.f, visited, queued, traversed,
ret_tuple_as_trans_desc=ret_tuple_as_trans_desc, lp_solved=lp_solved)
if curr.virgin:
# try to see if the firing sequence is already fine
firing_sequence, reach_fm, explained_events = me.get_firing_sequence(curr.x)
if reach_fm:
trans_list = __transitions_list_from_state(curr) + list(firing_sequence)
return __reconstruct_alignment(trans_list, curr.f, visited, queued, traversed,
ret_tuple_as_trans_desc=ret_tuple_as_trans_desc, lp_solved=lp_solved)
mm, index = __get_model_marking_and_index(curr.m)
__update_heu_dict(heu_dict, heu_max_ind_dict, mm, index, h, x, firing_sequence, incidence_matrix, cost_vec)
closed.add(current_marking)
visited += 1
enabled_trans = copy(trans_empty_preset)
for p in current_marking:
for t in p.ass_trans:
if t.sub_marking <= current_marking:
enabled_trans.add(t)
trans_to_visit_with_cost = [(t, cost_function[t]) for t in enabled_trans if not (
t is not None and utils.__is_log_move(t, skip) and utils.__is_model_move(t, skip))]
for t, cost in trans_to_visit_with_cost:
traversed += 1
new_marking = utils.add_markings(current_marking, t.add_marking)
if new_marking in closed:
continue
g = curr.g + cost
queued += 1
h, x = utils.__derive_heuristic(incidence_matrix, cost_vec, curr.x, t, curr.h)
trust = utils.__trust_solution(x)
mm, index = __get_model_marking_and_index(new_marking)
if not trust:
h2, x2, trust2 = __get_heu_from_dict(heu_dict, heu_max_ind_dict, mm, index)
if h2 is not None and (h2 > h or trust2):
h = h2
x = x2
trust = trust2
else:
__update_heu_dict_specific_point(heu_dict, heu_max_ind_dict, mm, index, h, x)
new_f = g + h
tp = utils.TweakedSearchTuple(new_f, g, h, new_marking, curr, t, x, trust, False)
heapq.heappush(open_set, tp)
def __search_without_synchr(net,
ini,
fin,
log_trace,
skip=utils.SKIP,
ret_tuple_as_trans_desc=True,
max_align_time_trace=sys.maxsize,
expo=2):
'''
In this function that can be called with the following way:
alignment.algorithm.apply(trace, net, marking, fmarking,variant=ali.VERSION_DIJKSTRA_EXPONENTIAL_HEURISTIC,
parameters={ali.Parameters.SYNCHRONOUS:False})
we compute the distance at each marking. However there is a question:
Should we keep the entire trace or cut it?
Notice that the heuristic on marking reachability is ON
'''
trace = [log_trace[i]["concept:name"] for i in range(len(log_trace))]
start_time = time.time()
decorate_transitions_prepostset(net)
decorate_places_preset_trans(net)
closed = {}
# a node in this version contains the marking state.m[Ø] and the position in the trace state.m[1]
ini_state = utils.DijkstraSearchTuple(0, (ini, 0), None, None, 0)
open_set = [ini_state]
heapq.heapify(open_set)
visited = 0
queued = 0
traversed = 0
def cost_function(t, l, expo):
if t == utils.SKIP:
return expo**(-l)
else:
return 0
trans_empty_preset = set(t for t in net.transitions if len(t.in_arcs) == 0)
while not len(open_set) == 0:
if (time.time() - start_time) > max_align_time_trace:
return None
curr = heapq.heappop(open_set)
current_marking = curr.m
already_closed = current_marking in closed.keys()
if already_closed:
# if marking has already been visited, we don't visit it again
# it's a heuristic
continue
if current_marking[0] == fin and current_marking[1] == len(trace):
return utils.__reconstruct_alignment(
curr,
visited,
queued,
traversed,
ret_tuple_as_trans_desc=ret_tuple_as_trans_desc)
closed[current_marking] = curr.l
visited += 1
# ----- either we try to move in model
possible_enabling_transitions = copy(trans_empty_preset)
for p in current_marking[0]:
for t in p.ass_trans:
possible_enabling_transitions.add(t)
enabled_trans = [
t for t in possible_enabling_transitions
if t.sub_marking <= current_marking[0]
]
trans_to_visit_with_cost = [(t, cost_function(utils.SKIP, curr.l,
expo))
for t in enabled_trans if t is not None]
for t, cost in trans_to_visit_with_cost:
traversed += 1
new_marking = (utils.add_markings(current_marking[0],
t.add_marking),
current_marking[1])
if new_marking in closed.keys():
# if marking has already been visited, we don't visit it again
# it's a heuristic
continue
queued += 1
tp = utils.DijkstraSearchTuple(curr.g + cost, new_marking, curr,
(">>", t), curr.l + 1)
heapq.heappush(open_set, tp)
# ------ either we try to move in log
if current_marking[1] < len(trace):
traversed += 1
new_marking = (current_marking[0], current_marking[1] + 1)
if new_marking in closed.keys():
# if marking has already been visited, we don't visit it again
# it's a heuristic
continue
queued += 1
tp = utils.DijkstraSearchTuple(
curr.g + cost_function(utils.SKIP, curr.l, expo), new_marking,
curr, (trace[current_marking[1]], ">>"), curr.l + 1)
heapq.heappush(open_set, tp)
# ------ either we try to move in both (synchronous moves)
trans_to_visit_with_cost = [
(t, cost_function(t, curr.l, expo)) for t in enabled_trans
if t is not None and str(t) == trace[current_marking[1]]
]
for t, cost in trans_to_visit_with_cost:
traversed += 1
new_marking = (utils.add_markings(current_marking[0],
t.add_marking),
current_marking[1] + 1)
if new_marking in closed.keys():
# if marking has already been visited, we don't visit it again
# it's a heuristic
continue
queued += 1
tp = utils.DijkstraSearchTuple(curr.g + cost, new_marking,
curr,
(trace[current_marking[1]], t),
curr.l + 1)
heapq.heappush(open_set, tp)
def __search_with_synchr(sync_net,
ini,
fin,
skip,
ret_tuple_as_trans_desc=False,
max_align_time_trace=sys.maxsize,
expo=2):
'''
In this function that can be called with the following way:
alignment.algorithm.apply(trace, net, marking, fmarking,variant=ali.VERSION_DIJKSTRA_EXPONENTIAL_HEURISTIC,
parameters={ali.Parameters.SYNCHRONOUS:True})
Cost of transition depends on the run of the synchronous product.
Other parameters:
ali.Parameters.EXPONENT:2 (change the base of the log)
'''
start_time = time.time()
decorate_transitions_prepostset(sync_net)
decorate_places_preset_trans(sync_net)
closed = {}
ini_state = utils.DijkstraSearchTuple(0, ini, None, None, 0)
open_set = [ini_state]
heapq.heapify(open_set)
visited = 0
queued = 0
traversed = 0
def cost_function(t, l, expo):
if t.label is None:
return expo**(-l)
if t.label[1] == utils.SKIP or t.label[0] == utils.SKIP:
return expo**(-l)
else:
return 0
trans_empty_preset = set(t for t in sync_net.transitions
if len(t.in_arcs) == 0)
while not len(open_set) == 0:
if (time.time() - start_time) > max_align_time_trace:
return None
curr = heapq.heappop(open_set)
current_marking = curr.m
already_closed = current_marking in closed.keys()
if already_closed:
continue
if current_marking == fin:
return utils.__reconstruct_alignment(
curr,
visited,
queued,
traversed,
ret_tuple_as_trans_desc=ret_tuple_as_trans_desc)
closed[current_marking] = curr.l
visited += 1
possible_enabling_transitions = copy(trans_empty_preset)
for p in current_marking:
for t in p.ass_trans:
possible_enabling_transitions.add(t)
enabled_trans = [
t for t in possible_enabling_transitions
if t.sub_marking <= current_marking
]
trans_to_visit_with_cost = [(t, cost_function(t, curr.l, expo))
for t in enabled_trans if t is not None]
for t, cost in trans_to_visit_with_cost:
traversed += 1
new_marking = utils.add_markings(current_marking, t.add_marking)
already_closed = new_marking in closed.keys()
if already_closed:
continue
queued += 1
tp = utils.DijkstraSearchTuple(curr.g + cost, new_marking, curr, t,
curr.l + 1)
heapq.heappush(open_set, tp)
