def apply(ocel: OCEL,
correspondence_dict: Dict[str, Collection[str]],
parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
"""
Filters an object-centric event log keeping only the specified object types
with the specified activity set (filters out the rest).
Parameters
----------------
ocel
Object-centric event log
correspondence_dict
Dictionary containing, for every object type of interest, a
collection of allowed activities. Example:
{"order": ["Create Order"], "element": ["Create Order", "Create Delivery"]}
Keeps only the object types "order" and "element".
For the "order" object type, only the activity "Create Order" is kept.
For the "element" object type, only the activities "Create Order" and "Create Delivery" are kept.
parameters
Parameters of the algorithm, including:
- Parameters.ACTIVITY_KEY => the activity key
- Parameters.OBJECT_TYPE => the object type column
Returns
-----------------
filtered_ocel
Filtered object-centric event log
"""
if parameters is None:
parameters = {}
activity_key = exec_utils.get_param_value(Parameters.ACTIVITY_KEY,
parameters, ocel.event_activity)
object_type_column = exec_utils.get_param_value(Parameters.OBJECT_TYPE,
parameters,
ocel.object_type_column)
temp_column = exec_utils.get_param_value(Parameters.TEMP_COLUMN,
parameters, "@@temp_column")
temp_separator = exec_utils.get_param_value(Parameters.TEMP_SEPARATOR,
parameters, "@#@#")
ocel = copy(ocel)
inv_dict = set()
for ot in correspondence_dict:
for act in correspondence_dict[ot]:
inv_dict.add(act + temp_separator + ot)
ocel.relations[temp_column] = ocel.relations[
activity_key] + temp_separator + ocel.relations[object_type_column]
ocel.relations = ocel.relations[ocel.relations[temp_column].isin(inv_dict)]
del ocel.relations[temp_column]
return filtering_utils.propagate_relations_filtering(ocel,
parameters=parameters)
def apply(ocel: OCEL, parameters: Optional[Dict[Any, Any]] = None):
"""
Creates an iterator over the events of an object-centric event log
Parameters
----------------
ocel
OCEL
parameters
Parameters of the method, including:
- Parameters.OCEL_TYPE_PREFIX => the prefix of the object types in the OCEL (default: ocel:type)
Returns
----------------
yielded event
The events of the OCEL, one by one.
"""
if parameters is None:
parameters = {}
ot_prefix = exec_utils.get_param_value(
Parameters.OCEL_TYPE_PREFIX, parameters,
ocel_constants.DEFAULT_OBJECT_TYPE_PREFIX_EXTENDED)
ext_table = ocel.get_extended_table(ot_prefix)
for k, ev in ext_table.iterrows():
yield {
x: y
for x, y in dict(ev).items()
if isinstance(y, list) or not pd.isna(y)
}
def apply(ocel: OCEL, parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
"""
Explode an OCEL: an event associated to N objects is "split" to N events, each one associated to one object.
Parameters
-----------------
ocel
Object-centric event log
parameters
Possible parameters of the algorithm
Returns
-----------------
ocel
Exploded object-centric event log
"""
if parameters is None:
parameters = {}
ocel = deepcopy(ocel)
ocel.relations[ocel.event_id_column] = ocel.relations[ocel.event_id_column] + "_" + ocel.relations[ocel.object_id_column]
ocel.events = ocel.relations.copy()
del ocel.events[ocel.object_id_column]
del ocel.events[ocel.object_type_column]
return ocel
def from_traditional_pandas(df: pd.DataFrame,
parameters: Optional[Dict[Any,
Any]] = None) -> OCEL:
"""
Transforms a dataframe to an OCEL
Parameters
-----------------
df
Pandas dataframe
parameters
Parameters of the algorithm, including:
- Parameters.TARGET_OBJECT_TYPE => the name of the object type to which the cases should be mapped
- Parameters.ACTIVITY_KEY => the attribute to use as activity
- Parameters.TIMESTAMP_KEY => the attribute to use as timestamp
- Parameters.CASE_ID_KEY => the attribute to use as case identifier
- Parameters.CASE_ATTRIBUTE_PREFIX => the prefix identifying the attributes at the case level
Returns
-----------------
ocel
OCEL (equivalent to the provided event log)
"""
if parameters is None:
parameters = {}
target_object_type = exec_utils.get_param_value(
Parameters.TARGET_OBJECT_TYPE, parameters, "OTYPE")
activity_key = exec_utils.get_param_value(Parameters.ACTIVITY_KEY,
parameters,
xes_constants.DEFAULT_NAME_KEY)
timestamp_key = exec_utils.get_param_value(
Parameters.TIMESTAMP_KEY, parameters,
xes_constants.DEFAULT_TIMESTAMP_KEY)
case_id_key = exec_utils.get_param_value(Parameters.CASE_ID_KEY,
parameters,
constants.CASE_CONCEPT_NAME)
case_attribute_prefix = exec_utils.get_param_value(
Parameters.CASE_ATTRIBUTE_PREFIX, parameters,
constants.CASE_ATTRIBUTE_PREFIX)
events = __get_events_dataframe(df, activity_key, timestamp_key,
case_id_key, case_attribute_prefix)
objects = __get_objects_dataframe(df, case_id_key, case_attribute_prefix,
target_object_type)
relations = __get_relations_from_events(events, target_object_type)
del events[ocel_constants.DEFAULT_OBJECT_ID]
events = events.sort_values([
ocel_constants.DEFAULT_EVENT_TIMESTAMP, ocel_constants.DEFAULT_EVENT_ID
])
relations = relations.sort_values([
ocel_constants.DEFAULT_EVENT_TIMESTAMP, ocel_constants.DEFAULT_EVENT_ID
])
return OCEL(events=events, objects=objects, relations=relations)
def apply(ocel: OCEL,
values: Collection[Any],
parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
"""
Filters the object-centric event log on the provided object attributes values
Parameters
----------------
ocel
Object-centric event log
values
Collection of values
parameters
Parameters of the algorithm, including:
- Parameters.ATTRIBUTE_KEY => the attribute that should be filtered
- Parameters.POSITIVE => decides if the values should be kept (positive=True) or removed (positive=False)
Returns
----------------
ocel
Filtered object-centric event log
"""
if parameters is None:
parameters = {}
attribute_key = exec_utils.get_param_value(Parameters.ATTRIBUTE_KEY,
parameters,
ocel.object_type_column)
positive = exec_utils.get_param_value(Parameters.POSITIVE, parameters,
True)
ocel = copy(ocel)
if positive:
ocel.objects = ocel.objects[ocel.objects[attribute_key].isin(values)]
else:
ocel.objects = ocel.objects[~ocel.objects[attribute_key].isin(values)]
return filtering_utils.propagate_object_filtering(ocel,
parameters=parameters)
def propagate_event_filtering(ocel: OCEL,
parameters: Optional[Dict[Any,
Any]] = None) -> OCEL:
"""
Propagates the filtering at the event level to the remaining parts of the OCEL structure
(objects, relations)
Parameters
----------------
ocel
Object-centric event log
parameters
Parameters of the algorithm, including:
- Parameters.EVENT_ID => the column to be used as case identifier
- Parameters.OBJECT_ID => the column to be used as object identifier
- Parameters.OBJECT_TYPE => the column to be used as object type
Returns
----------------
ocel
Object-centric event log with propagated filter
"""
if parameters is None:
parameters = {}
event_id = exec_utils.get_param_value(Parameters.EVENT_ID, parameters,
ocel.event_id_column)
object_id = exec_utils.get_param_value(Parameters.OBJECT_ID, parameters,
ocel.object_id_column)
selected_event_ids = set(ocel.events[event_id].unique())
ocel.relations = ocel.relations[ocel.relations[event_id].isin(
selected_event_ids)]
selected_object_ids = set(ocel.relations[object_id].unique())
ocel.objects = ocel.objects[ocel.objects[object_id].isin(
selected_object_ids)]
return ocel
def filter_end_events_per_object_type(
ocel: OCEL,
object_type: str,
parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
"""
Filters the events in which an object for the given object type terminates its lifecycle.
(E.g. an event with activity "Pay Order" might terminate an order).
Parameters
------------------
ocel
Object-centric event log
object_type
Object type to consider
parameters
Parameters of the algorithm, including:
- Parameters.EVENT_ID => the attribute working as event identifier
- Parameters.OBJECT_ID => the attribute working as object identifier
- Parameters.OBJECT_TYPE => the attribute working as object type
Returns
------------------
filtered_ocel
Filtered object-centric event log
"""
if parameters is None:
parameters = {}
event_id = exec_utils.get_param_value(Parameters.EVENT_ID, parameters,
ocel.event_id_column)
object_id = exec_utils.get_param_value(Parameters.OBJECT_ID, parameters,
ocel.object_id_column)
object_type_column = exec_utils.get_param_value(Parameters.OBJECT_TYPE,
parameters,
ocel.object_type_column)
evs = set(
ocel.relations[ocel.relations[object_type_column] == object_type].
groupby(object_id).last()[event_id].unique())
ocel = copy(ocel)
ocel.events = ocel.events[ocel.events[event_id].isin(evs)]
return filtering_utils.propagate_event_filtering(ocel,
parameters=parameters)
def sample_ocel_events(ocel: OCEL,
parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
"""
Keeps a sample of the events of an object-centric event log
Parameters
------------------
ocel
Object-centric event log
parameters
Parameters of the algorithm, including:
- Parameters.EVENT_ID => event identifier
- Parameters.NUM_EVENTS => number of events
Returns
------------------
sampled_ocel
Sampled object-centric event log
"""
if parameters is None:
parameters = {}
event_id_column = exec_utils.get_param_value(Parameters.EVENT_ID,
parameters,
ocel.event_id_column)
num_entities = exec_utils.get_param_value(Parameters.NUM_ENTITIES,
parameters, 100)
events = list(ocel.events[event_id_column].unique())
num_events = min(len(events), num_entities)
random.shuffle(events)
picked_events = events[:num_events]
ocel = copy(ocel)
ocel.events = ocel.events[ocel.events[event_id_column].isin(picked_events)]
return filtering_utils.propagate_event_filtering(ocel,
parameters=parameters)
def apply_timestamp(ocel: OCEL,
min_timest: Union[datetime.datetime, str],
max_timest: Union[datetime.datetime, str],
parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
"""
Filters the object-centric event log keeping events in the provided timestamp range
Parameters
-----------------
ocel
Object-centric event log
min_timest
Left extreme of the allowed timestamp interval (provided in the format: YYYY-mm-dd HH:MM:SS)
max_timest
Right extreme of the allowed timestamp interval (provided in the format: YYYY-mm-dd HH:MM:SS)
parameters
Parameters of the algorithm, including:
- Parameters.TIMESTAMP_KEY => the attribute to use as timestamp
Returns
-----------------
filtered_ocel
Filtered object-centric event log
"""
if parameters is None:
parameters = {}
timestamp_key = exec_utils.get_param_value(Parameters.TIMESTAMP_KEY,
parameters,
ocel.event_timestamp)
min_timest = get_dt_from_string(min_timest)
max_timest = get_dt_from_string(max_timest)
ocel = copy(ocel)
ocel.events = ocel.events[ocel.events[timestamp_key] >= min_timest]
ocel.events = ocel.events[ocel.events[timestamp_key] <= max_timest]
return filtering_utils.propagate_event_filtering(ocel,
parameters=parameters)
def apply(file_path: str, parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
"""
Imports an object-centric event log from a JSON-OCEL file, using the default JSON backend of Python
Parameters
-----------------
file_path
Path to the JSON-OCEL file
parameters
Parameters of the algorithm, including:
- Parameters.EVENT_ID
- Parameters.EVENT_ACTIVITY
- Parameters.EVENT_TIMESTAMP
- Parameters.OBJECT_ID
- Parameters.OBJECT_TYPE
- Parameters.INTERNAL_INDEX
Returns
------------------
ocel
Object-centric event log
"""
if parameters is None:
parameters = {}
ocel = json.load(open(file_path, "r"))
events = []
relations = []
objects = []
event_id = exec_utils.get_param_value(Parameters.EVENT_ID, parameters,
constants.DEFAULT_EVENT_ID)
event_activity = exec_utils.get_param_value(
Parameters.EVENT_ACTIVITY, parameters,
constants.DEFAULT_EVENT_ACTIVITY)
event_timestamp = exec_utils.get_param_value(
Parameters.EVENT_TIMESTAMP, parameters,
constants.DEFAULT_EVENT_TIMESTAMP)
object_id = exec_utils.get_param_value(Parameters.OBJECT_ID, parameters,
constants.DEFAULT_OBJECT_ID)
object_type = exec_utils.get_param_value(Parameters.OBJECT_TYPE,
parameters,
constants.DEFAULT_OBJECT_TYPE)
internal_index = exec_utils.get_param_value(
Parameters.INTERNAL_INDEX, parameters,
constants.DEFAULT_INTERNAL_INDEX)
parser = dt_parsing.parser.get()
types_dict = {}
for obj_id in ocel[constants.OCEL_OBJECTS_KEY]:
obj = ocel[constants.OCEL_OBJECTS_KEY][obj_id]
obj_type = obj[constants.DEFAULT_OBJECT_TYPE]
types_dict[obj_id] = obj_type
dct = {object_id: obj_id, object_type: obj_type}
for k, v in obj[constants.OCEL_OVMAP_KEY].items():
dct[k] = v
objects.append(dct)
for ev_id in ocel[constants.OCEL_EVENTS_KEY]:
ev = ocel[constants.OCEL_EVENTS_KEY][ev_id]
dct = {
event_id: ev_id,
event_timestamp:
parser.apply(ev[constants.DEFAULT_EVENT_TIMESTAMP]),
event_activity: ev[constants.DEFAULT_EVENT_ACTIVITY]
}
for k, v in ev[constants.OCEL_VMAP_KEY].items():
dct[k] = v
for obj in ev[constants.OCEL_OMAP_KEY]:
relations.append({
event_id:
ev_id,
event_activity:
ev[constants.DEFAULT_EVENT_ACTIVITY],
event_timestamp:
parser.apply(ev[constants.DEFAULT_EVENT_TIMESTAMP]),
object_id:
obj,
object_type:
types_dict[obj]
})
events.append(dct)
events = pd.DataFrame(events)
objects = pd.DataFrame(objects)
relations = pd.DataFrame(relations)
events[internal_index] = events.index
relations[internal_index] = relations.index
events = events.sort_values([event_timestamp, internal_index])
relations = relations.sort_values([event_timestamp, internal_index])
del events[internal_index]
del relations[internal_index]
globals = {}
globals[constants.OCEL_GLOBAL_LOG] = ocel[constants.OCEL_GLOBAL_LOG]
globals[constants.OCEL_GLOBAL_EVENT] = ocel[constants.OCEL_GLOBAL_EVENT]
globals[constants.OCEL_GLOBAL_OBJECT] = ocel[constants.OCEL_GLOBAL_OBJECT]
return OCEL(events, objects, relations, globals)
def apply(ocel: OCEL,
min_num_obj_type: Dict[str, int],
parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
"""
Filters the events of the object-centric logs which are related to at least
the specified amount of objects per type.
E.g. apply(ocel, {"order": 1, "element": 2})
Would keep the following events:
ocel:eid ocel:timestamp ocel:activity ocel:type:element ocel:type:order
0 e1 1980-01-01 Create Order [i4, i1, i3, i2] [o1]
1 e11 1981-01-01 Create Order [i6, i5] [o2]
2 e14 1981-01-04 Create Order [i8, i7] [o3]
Parameters
------------------
ocel
Object-centric event log
min_num_obj_type
Minimum number of objects per type
parameters
Parameters of the filter, including:
- Parameters.EVENT_ID => the event identifier
- Parameters.OBJECT_ID => the object identifier
- Parameters.OBJECT_TYPE => the object type
Returns
-----------------
filtered_event_log
Filtered object-centric event log
"""
if parameters is None:
parameters = {}
event_id = exec_utils.get_param_value(Parameters.EVENT_ID, parameters,
ocel.event_id_column)
object_id = exec_utils.get_param_value(Parameters.OBJECT_ID, parameters,
ocel.object_id_column)
object_type = exec_utils.get_param_value(Parameters.OBJECT_TYPE,
parameters,
ocel.object_type_column)
num_obj = objects_ot_count.get_objects_ot_count(ocel,
parameters=parameters)
filt_evs = set()
for evid, evobjs in num_obj.items():
is_ok = True
for k, v in min_num_obj_type.items():
if not k in evobjs:
is_ok = False
break
elif evobjs[k] < v:
is_ok = False
break
if is_ok:
filt_evs.add(evid)
ocel = copy(ocel)
ocel.events = ocel.events[ocel.events[event_id].isin(filt_evs)]
return filtering_utils.propagate_event_filtering(ocel,
parameters=parameters)
def get_ocel_from_extended_table(
df: pd.DataFrame,
objects_df: Optional[Dict[Any, Any]] = None,
parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
if parameters is None:
parameters = {}
object_type_prefix = exec_utils.get_param_value(
Parameters.OBJECT_TYPE_PREFIX, parameters,
constants.DEFAULT_OBJECT_TYPE_PREFIX_EXTENDED)
event_activity = exec_utils.get_param_value(
Parameters.EVENT_ACTIVITY, parameters,
constants.DEFAULT_EVENT_ACTIVITY)
event_id = exec_utils.get_param_value(Parameters.EVENT_ID, parameters,
constants.DEFAULT_EVENT_ID)
event_timestamp = exec_utils.get_param_value(
Parameters.EVENT_TIMESTAMP, parameters,
constants.DEFAULT_EVENT_TIMESTAMP)
object_id_column = exec_utils.get_param_value(Parameters.OBJECT_ID,
parameters,
constants.DEFAULT_OBJECT_ID)
object_type_column = exec_utils.get_param_value(
Parameters.OBJECT_TYPE, parameters, constants.DEFAULT_OBJECT_TYPE)
non_object_type_columns = set(x for x in df.columns
if not x.startswith(object_type_prefix))
object_type_columns = set(x for x in df.columns
if x.startswith(object_type_prefix))
meaningful_columns = object_type_columns.union(
{event_activity, event_id, event_timestamp})
internal_index = exec_utils.get_param_value(
Parameters.INTERNAL_INDEX, parameters,
constants.DEFAULT_INTERNAL_INDEX)
df_red = df[meaningful_columns]
stream = df_red.to_dict("records")
relations = []
objects = {x: set() for x in object_type_columns}
for ev in stream:
for ot in object_type_columns:
ot_stri = ot.split(object_type_prefix)[1]
ev[ot] = parse_list(ev[ot])
oot = objects[ot]
for obj in ev[ot]:
oot.add(obj)
relations.append({
event_id: ev[event_id],
event_activity: ev[event_activity],
event_timestamp: ev[event_timestamp],
object_id_column: obj,
object_type_column: ot_stri
})
relations = pd.DataFrame(relations)
if objects_df is None:
objects = [{
object_type_column: x.split(object_type_prefix)[1],
object_id_column: y
} for x in objects for y in objects[x]]
objects_df = pd.DataFrame(objects)
del objects
df = df[non_object_type_columns]
df[event_timestamp] = pd.to_datetime(df[event_timestamp])
df[internal_index] = df.index
relations[internal_index] = relations.index
df = df.sort_values([event_timestamp, internal_index])
relations = relations.sort_values([event_timestamp, internal_index])
del df[internal_index]
del relations[internal_index]
return OCEL(df, objects_df, relations)
def apply(file_path: str, parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
"""
Imports an object-centric event log from a XNK-OCEL file, using LXML
Parameters
-----------------
file_path
Path to the XML-OCEL file
parameters
Parameters of the algorithm, including:
- Parameters.EVENT_ID
- Parameters.EVENT_ACTIVITY
- Parameters.EVENT_TIMESTAMP
- Parameters.OBJECT_ID
- Parameters.OBJECT_TYPE
- Parameters.INTERNAL_INDEX
Returns
------------------
ocel
Object-centric event log
"""
if parameters is None:
parameters = {}
events = []
relations = []
objects = []
obj_type_dict = {}
event_id = exec_utils.get_param_value(Parameters.EVENT_ID, parameters,
constants.DEFAULT_EVENT_ID)
event_activity = exec_utils.get_param_value(
Parameters.EVENT_ACTIVITY, parameters,
constants.DEFAULT_EVENT_ACTIVITY)
event_timestamp = exec_utils.get_param_value(
Parameters.EVENT_TIMESTAMP, parameters,
constants.DEFAULT_EVENT_TIMESTAMP)
object_id = exec_utils.get_param_value(Parameters.OBJECT_ID, parameters,
constants.DEFAULT_OBJECT_ID)
object_type = exec_utils.get_param_value(Parameters.OBJECT_TYPE,
parameters,
constants.DEFAULT_OBJECT_TYPE)
internal_index = exec_utils.get_param_value(
Parameters.INTERNAL_INDEX, parameters,
constants.DEFAULT_INTERNAL_INDEX)
date_parser = dt_parsing.parser.get()
parser = etree.XMLParser(remove_comments=True)
tree = objectify.parse(file_path, parser=parser)
root = tree.getroot()
for child in root:
if child.tag.lower().endswith("events"):
for event in child:
eve_id = None
eve_activity = None
eve_timestamp = None
eve_omap = []
eve_vmap = {}
for child2 in event:
if child2.get("key") == "id":
eve_id = child2.get("value")
elif child2.get("key") == "timestamp":
eve_timestamp = parse_xml(child2.get("value"),
child2.tag.lower(),
date_parser)
elif child2.get("key") == "activity":
eve_activity = child2.get("value")
elif child2.get("key") == "omap":
for child3 in child2:
eve_omap.append(child3.get("value"))
elif child2.get("key") == "vmap":
for child3 in child2:
eve_vmap[child3.get("key")] = parse_xml(
child3.get("value"), child3.tag.lower(),
date_parser)
event_dict = {
event_id: eve_id,
event_activity: eve_activity,
event_timestamp: eve_timestamp
}
for k, v in eve_vmap.items():
event_dict[k] = v
events.append(event_dict)
for obj in eve_omap:
rel_dict = {
event_id: eve_id,
event_activity: eve_activity,
event_timestamp: eve_timestamp,
object_id: obj
}
relations.append(rel_dict)
elif child.tag.lower().endswith("objects"):
for object in child:
obj_id = None
obj_type = None
obj_ovmap = {}
for child2 in object:
if child2.get("key") == "id":
obj_id = child2.get("value")
elif child2.get("key") == "type":
obj_type = child2.get("value")
elif child2.get("key") == "ovmap":
for child3 in child2:
obj_ovmap[child3.get("key")] = parse_xml(
child3.get("value"), child3.tag.lower(),
date_parser)
objects.append({
object_id: obj_id,
object_type: obj_type,
constants.OCEL_OVMAP_KEY: obj_ovmap
})
obj_type_dict[obj_id] = obj_type
for rel in relations:
rel[object_type] = obj_type_dict[rel[object_id]]
events = pd.DataFrame(events)
objects = pd.DataFrame(objects)
relations = pd.DataFrame(relations)
events[internal_index] = events.index
relations[internal_index] = relations.index
events = events.sort_values([event_timestamp, internal_index])
relations = relations.sort_values([event_timestamp, internal_index])
del events[internal_index]
del relations[internal_index]
globals = {}
return OCEL(events, objects, relations, globals)
def from_traditional_log(log: EventLog,
parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
"""
Transforms an EventLog to an OCEL
Parameters
-----------------
log
Event log
parameters
Parameters of the algorithm, including:
- Parameters.TARGET_OBJECT_TYPE => the name of the object type to which the cases should be mapped
- Parameters.ACTIVITY_KEY => the attribute to use as activity
- Parameters.TIMESTAMP_KEY => the attribute to use as timestamp
- Parameters.CASE_ID_KEY => the attribute to use as case identifier
Returns
-----------------
ocel
OCEL (equivalent to the provided event log)
"""
if parameters is None:
parameters = {}
log = log_converter.apply(log,
variant=log_converter.Variants.TO_EVENT_LOG,
parameters=parameters)
target_object_type = exec_utils.get_param_value(
Parameters.TARGET_OBJECT_TYPE, parameters, "OTYPE")
activity_key = exec_utils.get_param_value(Parameters.ACTIVITY_KEY,
parameters,
xes_constants.DEFAULT_NAME_KEY)
timestamp_key = exec_utils.get_param_value(
Parameters.TIMESTAMP_KEY, parameters,
xes_constants.DEFAULT_TIMESTAMP_KEY)
case_id_key = exec_utils.get_param_value(Parameters.CASE_ID_KEY,
parameters,
xes_constants.DEFAULT_TRACEID_KEY)
events = []
objects = []
relations = []
ev_count = 0
for trace in log:
case_id = trace.attributes[case_id_key]
obj = {
ocel_constants.DEFAULT_OBJECT_ID: case_id,
ocel_constants.DEFAULT_OBJECT_TYPE: target_object_type
}
for attr in trace.attributes:
if attr != case_id_key:
obj[attr] = trace.attributes[attr]
objects.append(obj)
for ev in trace:
ev_count = ev_count + 1
activity = ev[activity_key]
timestamp = ev[timestamp_key]
eve = {
ocel_constants.DEFAULT_EVENT_ID: str(ev_count),
ocel_constants.DEFAULT_EVENT_ACTIVITY: activity,
ocel_constants.DEFAULT_EVENT_TIMESTAMP: timestamp
}
for attr in ev:
if attr not in [activity, timestamp]:
eve[attr] = ev[attr]
events.append(eve)
relations.append({
ocel_constants.DEFAULT_EVENT_ID:
str(ev_count),
ocel_constants.DEFAULT_EVENT_ACTIVITY:
activity,
ocel_constants.DEFAULT_EVENT_TIMESTAMP:
timestamp,
ocel_constants.DEFAULT_OBJECT_ID:
case_id,
ocel_constants.DEFAULT_OBJECT_TYPE:
target_object_type
})
events = pd.DataFrame(events)
objects = pd.DataFrame(objects)
relations = pd.DataFrame(relations)
return OCEL(events=events, objects=objects, relations=relations)
def from_interleavings(df1: pd.DataFrame,
df2: pd.DataFrame,
interleavings: pd.DataFrame,
parameters: Optional[Dict[Any, Any]] = None) -> OCEL:
"""
Transforms a couple of dataframes, along with the interleavings between them, to an OCEL
Parameters
-----------------
df1
First of the two dataframes
df2
Second of the two dataframes
interleavings
Interleavings dataframe
parameters
Parameters of the algorithm, including:
- Parameters.ACTIVITY_KEY => the attribute to use as activity
- Parameters.TIMESTAMP_KEY => the attribute to use as timestamp
- Parameters.CASE_ID_KEY => the attribute to use as case identifier
- Parameters.CASE_ATTRIBUTE_PREFIX => the prefix identifying the attributes at the case level
- Parameters.TARGET_OBJECT_TYPE => the name of the object type to which the cases of the first log should be mapped
- Parameters.TARGET_OBJECT_TYPE_2 => the name of the object type to which the cases of the second log should be mapped
- Parameters.LEFT_INDEX => the index column of the events of the first dataframe, in the interleavings dataframe
- Parameters.RIGHT_INDEX => the index column of the events of the second dataframe, in the interleavings
dataframe.
- Parameters.DIRECTION => the direction of the interleavings (LR or RL)
Returns
-----------------
ocel
OCEL (equivalent to the provided event log)
"""
if parameters is None:
parameters = {}
activity_key = exec_utils.get_param_value(Parameters.ACTIVITY_KEY,
parameters,
xes_constants.DEFAULT_NAME_KEY)
timestamp_key = exec_utils.get_param_value(
Parameters.TIMESTAMP_KEY, parameters,
xes_constants.DEFAULT_TIMESTAMP_KEY)
case_id_key = exec_utils.get_param_value(Parameters.CASE_ID_KEY,
parameters,
constants.CASE_CONCEPT_NAME)
case_attribute_prefix = exec_utils.get_param_value(
Parameters.CASE_ATTRIBUTE_PREFIX, parameters,
constants.CASE_ATTRIBUTE_PREFIX)
target_object_type = exec_utils.get_param_value(
Parameters.TARGET_OBJECT_TYPE, parameters, "OTYPE")
target_object_type_2 = exec_utils.get_param_value(
Parameters.TARGET_OBJECT_TYPE_2, parameters, "OTYPE2")
left_index = exec_utils.get_param_value(Parameters.LEFT_INDEX, parameters,
"@@left_index")
right_index = exec_utils.get_param_value(Parameters.RIGHT_INDEX,
parameters, "@@right_index")
direction = exec_utils.get_param_value(Parameters.DIRECTION, parameters,
"@@direction")
events1 = __get_events_dataframe(df1,
activity_key,
timestamp_key,
case_id_key,
case_attribute_prefix,
events_prefix="E1_")
objects1 = __get_objects_dataframe(df1, case_id_key, case_attribute_prefix,
target_object_type)
relations1 = __get_relations_from_events(events1, target_object_type)
relations1_minimal = relations1[{
ocel_constants.DEFAULT_EVENT_ID, ocel_constants.DEFAULT_OBJECT_ID,
ocel_constants.DEFAULT_OBJECT_TYPE
}]
events2 = __get_events_dataframe(df2,
activity_key,
timestamp_key,
case_id_key,
case_attribute_prefix,
events_prefix="E2_")
objects2 = __get_objects_dataframe(df2, case_id_key, case_attribute_prefix,
target_object_type_2)
relations2 = __get_relations_from_events(events2, target_object_type_2)
relations2_minimal = relations2[{
ocel_constants.DEFAULT_EVENT_ID, ocel_constants.DEFAULT_OBJECT_ID,
ocel_constants.DEFAULT_OBJECT_TYPE
}]
interleavings[left_index] = "E1_" + interleavings[left_index].astype(
int).astype(str)
interleavings[right_index] = "E2_" + interleavings[right_index].astype(
int).astype(str)
interleavings_lr = interleavings[interleavings[direction] == "LR"][[
left_index, right_index
]]
interleavings_rl = interleavings[interleavings[direction] == "RL"][[
left_index, right_index
]]
relations3 = events1.merge(interleavings_lr,
left_on=ocel_constants.DEFAULT_EVENT_ID,
right_on=left_index)
relations3 = relations3.merge(relations2_minimal,
left_on=right_index,
right_on=ocel_constants.DEFAULT_EVENT_ID,
suffixes=('', '_@#@#RIGHT'))
relations3[ocel_constants.DEFAULT_OBJECT_ID] = relations3[
ocel_constants.DEFAULT_OBJECT_ID + '_@#@#RIGHT']
relations3[ocel_constants.DEFAULT_OBJECT_TYPE] = target_object_type_2
relations4 = events2.merge(interleavings_rl,
left_on=ocel_constants.DEFAULT_EVENT_ID,
right_on=right_index)
relations4 = relations4.merge(relations1_minimal,
left_on=left_index,
right_on=ocel_constants.DEFAULT_EVENT_ID,
suffixes=('', '_@#@#LEFT'))
relations4[ocel_constants.DEFAULT_OBJECT_ID] = relations4[
ocel_constants.DEFAULT_OBJECT_ID + '_@#@#LEFT']
relations4[ocel_constants.DEFAULT_OBJECT_TYPE] = target_object_type
del events1[ocel_constants.DEFAULT_OBJECT_ID]
del events2[ocel_constants.DEFAULT_OBJECT_ID]
events = pd.concat([events1, events2])
objects = pd.concat([objects1, objects2])
relations = pd.concat([relations1, relations2, relations3, relations4])
events = events.sort_values([
ocel_constants.DEFAULT_EVENT_TIMESTAMP, ocel_constants.DEFAULT_EVENT_ID
])
relations = relations.sort_values([
ocel_constants.DEFAULT_EVENT_TIMESTAMP, ocel_constants.DEFAULT_EVENT_ID
])
return OCEL(events=events, objects=objects, relations=relations)