def test_prepare_log(self): lg = EventLog() lg.add_trace(1, ['a', 'b', 'g', 'h', 'i', 'j', 'k']) lg.add_trace(2, ['a', 'b', 'g', 'h', 'j', 'i', 'k']) lg.add_trace(3, ['a', 'b', 'g', 'h', 'i', 'k']) lg.add_trace(4, ['a', 'b', 'g', 'h', 'j', 'k']) lg.add_trace(5, ['a', 'b', 'c', 'd', 'e', 'h', 'i', 'k']) lg.add_trace(6, ['a', 'b', 'c', 'd', 'e', 'f', 'c', 'd', 'e', 'h', 'i', 'k']) miner = InductiveMiner(lg) result = [ ['a', 'b', 'g', 'h', 'i', 'j', 'k'], ['a', 'b', 'g', 'h', 'j', 'i', 'k'], ['a', 'b', 'g', 'h', 'i', 'k'], ['a', 'b', 'g', 'h', 'j', 'k'], ['a', 'b', 'c', 'd', 'e', 'h', 'i', 'k'], ['a', 'b', 'c', 'd', 'e', 'f', 'c', 'd', 'e', 'h', 'i', 'k']] result.sort() miner.log.sort() self.assertListEqual(miner.log, result)
def __call__(self, activities_to_keep) -> EventLog:
f = EventLog()
for e in self.log:
if e.get_activity_name() in activities_to_keep:
f.add_event(e)
self.log = f
return self.log
def test_extended_example(self): lg = EventLog() lg.add_trace(1, ['submit application', # a 'review application', # b 'fast forwarding', # g 'final review', # h 'supervisor signature', # i 'sign application', # j 'close application']) # k lg.add_trace(2, ['submit application', # a 'review application', # b 'fast forwarding', # c 'final review', # h 'sign application', # j 'supervisor signature', # i 'close application']) # k lg.add_trace(3, ['submit application', # a 'review application', # b 'fast forwarding', # g 'final review', # h 'supervisor signature', # i 'close application']) # k lg.add_trace(4, ['submit application', # a 'review application', # b 'fast forwarding', # g 'final review', # h 'sign application', # j 'close application']) # k lg.add_trace(5, ['submit application', # a 'review application', # b 'check documents', # c 'check financial status', # d 'write report', # e 'final review', # h 'supervisor signature', # i 'close application']) # k lg.add_trace(6, ['submit application', # a 'review application', # b 'check documents', # c 'check financial status', # d 'write report', # e 'reject report', # f 'check documents', # c 'check financial status', # d 'write report', # e 'final review', # h 'supervisor signature', # i 'close application']) # k miner = InductiveMiner(lg) tree = miner.discover() g = Graph() g = g.from_log(miner.log) d = dot.draw_graph(g) dot.render_dot(d, 'dfg_process_tree') tree.print_tree() dot.draw_process_tree(tree, 'example', format='pdf')
def __call__(self, ending_activity) -> EventLog:
f = EventLog()
for t in self.log.get_traces():
if t[-1].get_activity_name() is ending_activity:
for e in t:
f.add_event(e)
self.log = f
return self.log
def test_find_loop(self):
log = [['d', 'e'], ['d', 'e', 'f', 'g', 'd', 'e']]
tree = ProcessTree(InductiveMiner(EventLog()), log, discover=False)
split = tree.find_loop()
list(map(lambda sub: sub.sort(), split))
print(split)
self.assertListEqual(split, [['d', 'e'], ['f', 'g']])
def test_find_seq(self):
log = [['a', 'x', 'b', 'c'], ['a', 'x', 'c', 'b'],
['a', 'x', 'd', 'e'], ['a', 'x', 'd', 'e', 'f', 'd', 'e']]
tree = ProcessTree(InductiveMiner(EventLog()), log, discover=False)
split = tree.find_seq()
self.assertListEqual(split, [['a', 'x']])
self.fail()
def test_split_loop_cut(self): log = [['d', 'e', 'f', 'g', 'd', 'e'], ['d', 'e']] miner = InductiveMiner(EventLog()) ll, lr = miner.split_log([['d', 'e'], ['f', 'g']], Cut.LOOP, log) self.assertListEqual(ll, [['d', 'e']]) self.assertListEqual(lr, [['f', 'g'], []])
def test_loop_of_one(self): lg = EventLog() lg.add_trace(1, ['a', 'b', 'b', 'c']) lg.add_trace(2, ['a', 'c']) miner = InductiveMiner(lg) tree = miner.discover() print() tree.print_tree() dot.draw_process_tree(tree, 'single_loop', format='pdf')
def __call__(self, name: str, value) -> EventLog:
f = EventLog()
for e in self.log:
if isinstance(value, list):
if e[name] in value:
f.add_event(e)
else:
if e[name] is value:
f.add_event(e)
self.log = f
return self.log
def test_split_para_cut(self): log = [['a', 'b'], ['b', 'a'], ['a']] miner = InductiveMiner(EventLog()) ll, lr = miner.split_log([['a'], ['b']], Cut.PARA, log) self.assertListEqual(ll, [['a']]) self.assertListEqual(lr, [['b'], []])
def test_split_max_seq_cut(self): log = [['a', 'x', 'b', 'c'], ['a', 'x', 'c', 'b'], ['a', 'd', 'e'], ['a', 'x', 'd', 'e', 'f', 'd', 'e']] miner = InductiveMiner(EventLog()) max_seq_split = [['a'], ['x'], [['b', 'c'], ['d', 'e', 'f']]] logs = miner.split_log(max_seq_split, Cut.SEQ, log) print(logs)
def test_split_excl_cut(self): log = [['a', 'b'], ['d', 'e', 'f', 'd','e'], ['d', 'e']] miner = InductiveMiner(EventLog()) ll, lr = miner.split_log([['a', 'b'], ['d', 'e', 'f']], Cut.EXCL, log) self.assertListEqual(ll, [['a', 'b']]) self.assertListEqual(lr, [['d', 'e', 'f', 'd','e'], ['d', 'e']])
def test_find_para(self):
log = [['a', 'b'], ['a', 'b'], ['b', 'a']]
# log = [['supervisor signature', # i
# 'sign application'], # j
# ['sign application', # j
# 'supervisor signature']] # i
tree = ProcessTree(InductiveMiner(EventLog()), log, discover=False)
split = tree.find_para()
split.sort()
print(split)
self.assertListEqual(split, [['a'], ['b']])
def __call__(self, first_activity, second_activity) -> EventLog:
f = EventLog()
for t in self.log.get_traces():
contains_df = False
for e1, e2 in zip(t[:-1], t[1:]):
if e1.get_activity_name(
) is first_activity and e2.get_activity_name(
) is second_activity:
contains_df = True
break
if contains_df:
for e in t:
f.add_event(e)
self.log = f
return self.log
def test_find__max_seq(self):
log = [['a', 'x', 'b', 'c'], ['a', 'x', 'c', 'b'],
['a', 'x', 'd', 'e'], ['a', 'x', 'd', 'e', 'f', 'd', 'e']]
tree = ProcessTree(InductiveMiner(EventLog()), log, discover=False)
split = tree.find_seq()
list(map(lambda sub: sub.sort(), split))
for ele in split:
for sub in ele:
if isinstance(sub, list):
sub.sort()
# split.sort()
self.assertListEqual(split,
[['a'], ['x'], [['b', 'c'], ['d', 'e', 'f']]])
def test_find_cuts(self):
log4 = [['a', 'b', 'g', 'h', 'i', 'j', 'k'],
['a', 'b', 'g', 'h', 'j', 'i', 'k'],
['a', 'b', 'g', 'h', 'i', 'k'], ['a', 'b', 'g', 'h', 'j', 'k'],
['a', 'b', 'c', 'd', 'e', 'h', 'i', 'k'],
['a', 'b', 'c', 'd', 'e', 'f', 'c', 'd', 'e', 'h', 'i', 'k']]
log2 = [['a', 'b', 'c'], ['a', 'c', 'b'], ['a', 'd', 'e'],
['a', 'd', 'e', 'f', 'd', 'e']]
miner = InductiveMiner(EventLog())
tree = ProcessTree(miner, log2)
print()
tree.print_tree()
dot.draw_process_tree(tree)
def test_simple_example(self): lg = EventLog() lg.add_trace(1, ['a', 'b', 'g', 'h', 'i', 'j', 'k']) lg.add_trace(2, ['a', 'b', 'g', 'h', 'j', 'i', 'k']) lg.add_trace(3, ['a', 'b', 'g', 'h', 'i', 'k']) lg.add_trace(4, ['a', 'b', 'g', 'h', 'j', 'k']) lg.add_trace(5, ['a', 'b', 'c', 'd', 'e', 'h', 'i', 'k']) lg.add_trace(6, ['a', 'b', 'c', 'd', 'e', 'f', 'c', 'd', 'e', 'h', 'i', 'k']) miner = InductiveMiner(lg) tree = miner.discover() tree.print_tree() dot.draw_process_tree(tree, format='pdf')
def test_split_seq_cut(self): log = [['a', 'x', 'b', 'c'], ['a', 'x', 'c', 'b'], ['a', 'x', 'd', 'e'], ['d', 'e'], ['a', 'x', 'd', 'e', 'f', 'd', 'e']] lg = EventLog() miner = InductiveMiner(lg) ll, lr = miner.split_log([['a', 'x'], ['b', 'c', 'd', 'e', 'f']], Cut.SEQ, log) self.assertListEqual(ll, [['a', 'x'], []]) self.assertListEqual(lr, [['b', 'c'], ['c', 'b'], ['d', 'e'], ['d', 'e', 'f', 'd', 'e']])
def test_tree_to_net(self): lg = EventLog() lg.add_trace(1, ['a', 'b', 'g', 'h', 'i', 'j', 'k']) lg.add_trace(2, ['a', 'b', 'g', 'h', 'j', 'i', 'k']) lg.add_trace(3, ['a', 'b', 'g', 'h', 'i', 'k']) lg.add_trace(4, ['a', 'b', 'g', 'h', 'j', 'k']) lg.add_trace(5, ['a', 'b', 'c', 'd', 'e', 'h', 'i', 'k']) lg.add_trace(6, ['a', 'b', 'c', 'd', 'e', 'f', 'c', 'd', 'e', 'h', 'i', 'k']) miner = InductiveMiner(lg) tree = miner.discover() tree.print_tree() # dot.draw_process_tree(tree, format='pdf') net = miner.tree_to_petri_net(tree) dot.draw_petri_net(net, 'process_tree_to_net')
def __parse_xes_file(context, tree):
log = None
trace_idx = None
event = None
trace_count = 0 #for trace_id
for action, elem in tqdm(context):
if action == __START: # starting to read
parent = tree[
elem.getparent()] if elem.getparent() in tree else None
if elem.tag.endswith(c.STR):
if parent is not None:
tree = __parse_attribute(elem, parent, elem.get(c.KEY),
elem.get(c.VAL), tree)
continue
elif elem.tag.endswith(c.DTE):
try:
date = __parse_date_ciso(elem.get(c.VAL))
tree = __parse_attribute(elem, parent, elem.get(c.KEY),
date, tree)
except TypeError:
logging.info("failed to parse date: " +
str(elem.get(c.VAL)))
except ValueError:
logging.info("failed to parse date: " +
str(elem.get(c.VAL)))
continue
elif elem.tag.endswith(c.BOL):
if parent is not None:
str_val = str(elem.get(c.VAL)).lower()
if str_val == 'true':
bool_val = True
elif str_val == 'false':
bool_val = False
else:
raise ValueError("failed to parse bool: " +
str(elem.get(c.VAL)))
tree = __parse_attribute(elem, parent, elem.get(c.KEY),
bool_val, tree)
continue
elif elem.tag.endswith(c.INT):
if parent is not None:
int_val = int(elem.get(c.VAL))
tree = __parse_attribute(elem, parent, elem.get(c.KEY),
int_val, tree)
continue
elif elem.tag.endswith(c.FLT):
if parent is not None:
float_val = float(elem.get(c.VAL))
tree = __parse_attribute(elem, parent, elem.get(c.KEY),
float_val, tree)
continue
elif elem.tag.endswith(c.EVE):
if event is not None:
raise SyntaxError(
'file contains <event> in another <event> tag')
if trace_idx is None:
raise SyntaxError(
'file contains a <event> element outside of a <trace> element (trace_id is None)'
)
event = Event()
tree[elem] = event
continue
elif elem.tag.endswith(c.TRC):
if trace_idx is not None:
raise SyntaxError(
'file contains <trace> in another <trace> tag')
trace_idx = trace_count
trace_count += 1
log.add_trace(trace_idx)
tree[elem] = log.traces[trace_idx]
continue
elif elem.tag.endswith(c.EXT):
if log is None:
raise SyntaxError('extension found outside of <log> tag')
if elem.get(c.NME) is not None and elem.get(
c.PRX) is not None and elem.get(c.URI) is not None:
log.extensions[elem.attrib[c.NME]] = {
c.PRX: elem.attrib[c.PRX],
c.URI: elem.attrib[c.URI]
}
continue
elif elem.tag.endswith(c.GLB):
if log is None:
raise SyntaxError('global found outside of <log> tag')
if elem.get(c.SCP) is not None:
log.globals[elem.attrib[c.SCP]] = {}
tree[elem] = log.globals[elem.get(c.SCP)]
continue
elif elem.tag.endswith(c.CLS):
if log is None:
raise SyntaxError('classifier found outside of <log> tag')
if elem.get(c.KYS) is not None:
name, attributes = elem.attrib[c.NME], elem.attrib[c.KYS]
# resolve special case; refactor with __parse_classifier
if "'" in attributes:
attributes = re.findall(r'\'(.*?)\'', attributes)
else:
attributes = attributes.split()
log.classifiers[name] = attributes
continue
elif elem.tag.endswith(c.LOG):
if log is not None:
raise SyntaxError('file contains > 1 <log> tags')
log = EventLog()
tree[elem] = log.attributes
continue
elif action == __END:
__clear_element(elem, tree)
if elem.tag.endswith(c.EVE):
if trace_idx is not None:
log.add_event(event, trace_idx)
event = None
continue
elif elem.tag.endswith(c.TRC):
trace_idx = None
continue
elif elem.tag.endswith(c.LOG):
continue
return log