PM4Py 2.2.12

Fixed

• a374bad
  • #251
• e88a654
  • #249
• 8451162
  • fix minor bug in the calculation of the handover and subcontracting metrics.

Removed

Deprecated

Changed

• 01fd040
  • The pm4py.view_petri_net() method now uses None as a default initial and final marking.
• 72ed7d0
  • Improved performance of variant discovery of dataframes.

Added

• 9a04357
  • Add rework measurement at the case level in the pm4py.statistics.rework package.
• b725ca0
  • add 'between' filter for pandas dataframes in the pm4py.algo.filtering.pandas package. The filter
    returns subsequences between the two given activities. It creates subtraces for every possible match.
• 211e3c5
  • added local linear embeddings to log_to_features.util.
• 4b59422
  • add support for adding decision points to data petri nets.
• 9261270
  • add support for performance dfg discovery in pm4py.discover_performance_dfg().

Other

PM4Py 2.2.11

Fixed

• 207d69b
  • bug fix in application of the filtering threshold in the IMf algorithm

Removed

Deprecated

Changed

• d98cbb1
  • changed deepcopy and copy functionality of logs for performance improvement
• f3b78a4
  • minor performance optimization in log conversion (log to dataframe)
• 71c0919
  • improved performance for pands -> stream conversion

Added

• f2101a7
  • added various additional features in log-based feature extraction
• 4187365
  • possiblity to directly get all performance metrics of the DFG elements
• 886b44e
  • detection method for trace-level attributes stored at event level
• d5f9f86
  • add transition names to events, based on a given alignment
• 4802e7d
  • add support for importing reset/inhibitor arcs and transition guards
• cc6488f
  • add general support for reset/inhibitor nets
• e805cf5
  • add support for data petri nets
• 1d3a2e7
  • added case termination statistics for pandas data frames

Other

PM4Py 2.2.10.2

Fixed

• 50ad39f
  • Fixed blocking issue with properties of Pandas dataframes when format_dataframe is used (case ID column)
• 3708b98
  • Fixed variants filter, when the output of get_variants_as_tuples is used

Removed

Deprecated

• Deprecated support to Pandas < 0.25 (due to future dropping)
• Deprecated auto-filters (due to future dropping)

Changed

• Different interventions to fix the internal coherency of the project (usage of deprecated functions + missing imports)

PM4Py 2.2.10

Fixed

Removed

Deprecated

Changed

• 4964d6e
  • minor refactoring (rename) in attribute statistics querying; get_attributes() --> get_event_attributes()
• 1148f6c
  • use revised implementation of IM and IMf everywhere, deprecate old implementations

Added

• 6750bf3
  • add support for start time and end-time in timstamp conversion
• e24f5b7
  • computation of event-level overlap
• 8cec5f9
  • add several case/event level statistic functions at the simplified interface level

Other

PM4Py 2.2.9

Fixed

• daf74e8
  • update imports in feature extraction
• 74be3e3
  • minor bug fix in alpha plus (place that was created was not always added to the resulting Petri net)

Removed

Deprecated

Changed

• d97b179
  • drop deepcopy in event log sorting (enhances performance)
• 1d4e625
  • revised IMf implementation (more close to ProM / PhD thesis Sander Leemans)
• 20aabd9
  • calculation of minimum self distance now adheres to the standard invocation structure

Added

• 598c6ec
  • simplified interface now stores properties (using attr attribute) to dataframes
• 1f7a3fa
  • add computation of rework statistic (cases containing the same activity more than once)
• 32c7d33
  • add computation of cycle time (active time of process divided by the number of instances of the process)
• 8187f0e
  • add distribution plots over different time-frames (matplotlib)
• 269d826
  • add batch detection based on Martin, N., Swennen, M., Depaire, B., Jans, M., Caris, A., & Vanhoof, K. (2015,
    December). Batch Processing: Definition and Event Log Identification. In SIMPDA (pp. 137-140).
• d5326d4
  • compute case overlap of a case with all other cases

Other

• 92a7058
  • performance optimization for calculation of performance spectrum
• b0fc57c
  • performance optimization for Pandas datetime conversion non-ISO8601 (regular formats)

PM4Py 2.2.8

Fixed

• c11bab8
  • bug fix in eventually-follows filter
• d3fd1bc
  • bug fix in activity frequency constraints of the log skeleton conformance checking

Removed

Deprecated

Changed

• d96d9d6
  • improved performance of the df-based performance spectrum code
• 499d8a1
  • improved performance of log conversions when (for internal use) deep copy is not required

Added

• 4d67993
  • allow the possibility to filter on a trace attribute that has a type date (e.g., does the planned start date of
    the case fall in a given time window?)
• b7ef36e
  • add properties object to trace attributes (used for internal storage of statistics, will not be exported to disk)
• d702936
  • added some basic ML utilities for event logs, e.g., getting all prefixes of traces, get a train/test split
• 1ec5802
  • new subtrace selection mechanism that gets all events inbetween two given activity labels (uses first match on
    both 1st and 2nd label)
• 9b65bbd
  • allow specification of business hours in sojourn time computation of the DFG
• 4d529d6
  • generic support for feature extraction

Other

PM4Py 2.2.7

Fixed

• 908e06d
  • fix error in loop detection of inductive miner
• b7b63e0
  • add internal log conversion in the flexible heuristics miner
• e9d61bd
  • fix minor bug in bpmn model importing
• 52cc0c7
  • fix minor bug in xes exporting (type of concept:name was not checked)

Removed

Deprecated

• 9c1a961
  • various old utility functions are now deprecated

Changed

• • avoid warnings when visualizing long place names in debug visualization mode

Added

• c2a9633, 52e340b
  • add simple visualization of performance spectrum.
• b6ae4b2
  • add simple dotted chart visualization to the simplified interface.
• 6e3a0ba
  • add properties attribute to event logs and event streams for storage of custom meta-data that is not exported to
    xes.
• fb14235
  • add version of dfg discovery that adds case-level attributes to nodes and edges
• d902609
  • add basic visualization of events per time and cas distribution graphs

Other

PM4Py 2.2.6

Fixed

Removed

Deprecated

Changed

• 766fafa
  • minor refactoring and more generic invocation style for eventually follows-based filtering

Added

• 353c7d6
  • Heuristics miner is now able to filter on edges connecting to/from start/end activity
• d641233
  • Parallel alignment computation can be directly invoked
    using pm4py.conformance_diagnostics_alignments(..., multi_processing=True)
• de84e5f
  • add pm4py.discover_bpmn_inductive(log)

Other

PM4Py 2.2.5

Fixed

• 9854f62
  • minor bug fix in etree xes exporter avoiding faulty None values
• bfe8fb3
  • support non-standard attribute symbols in line-by-line event log exporter

Removed

Deprecated

Changed

• 3631fe5
  • default xes importer is set back to iterparse
• a7ff695
  • large-scale restructuring of the underlying pm4py architecture
• 201879a
  • changed the default maximum number of edges to be visualized in the DFG visualization to 100000 (was: 75)

Added

• 6628396
  • sojourn-time-based coloring for dfgs
• 6639d3f
  • organizational mining, e.g., pm4py.discover_handover_of_work_network(log)
• 9c9ca14
  • allow multiprocessing in alignment computation
• 279fd31
  • add prefix tree vizualiation
• 748c768
  • add 'old' pm4py visualization of the process tree as an alternative visualziation
• 408b37a
  • add filter to check multiple ocurrences of certain attribute values in a case.

Other

PM4Py 2.2.4

Fixed

Removed

Deprecated

Changed

• 56317d8
  • process tree based alignments no longer use trace-based process tree reduction (can still be used through utils)
• c1c1ffc
  • minor optimizations to state-equation based alignment computation
• c95d45c
  • large (internal) refactoring of PM4Py object files and algorithms

Added

• d14d3d2
  • added resource profiles to PM4Py taken from Pika, Anastasiia, et al. "Mining resource profiles from event logs."
    ACM Transactions on Management Information Systems (TMIS) 8.1 (2017): 1-30.
• ab56d89
  • organizational mining according to https://arxiv.org/abs/2011.12445; contains several organizational group-based
    metrics
• 6a77a94
  • add serialization and deserialization to various pm4py objects; available through pm4py.serialize()
    and pm4py.deserialize()

Other