PM4Py 2.1.3

This is a minor release, consisting of the following changes:

1. commit 41ed572
   • deepcopy of inputs: since the dictionaries/sets are modified, a "deepcopy" is the best option to ensure data
     integrity.
   • keep_all_activities parameter in paths filter: decides if all the activities (also the ones connected by the low
     occurrences edges) should be kept, or only the ones appearing in the edges with more occurrences (default).
2. commit 63ccc05
   • IM and IMf - Removed dependency on "pm4py.algo.filtering" package
3. commit 388348f
   • bugfix remove_flow BPMN function
4. commit 8a5788f
   • more advanced to bpmn conversion in the simplified interface
5. commit 3bd4fe0
   • bug fix - DFG visualization needs deepcopy, otherwise it can remove element from the underlying DFG
6. commit 92fde8c
   • bug fix in Petri net playout in stop criterion
7. commit 66e0c07
   • refactoring conversion parameters log->stream and improved stream compression
8. commit 73054b0
   • improved performance of the line-by-line xes importer
   • increased XES-standard conformity
9. commit e6136ce
   • improved performance of the xes exporters
   • increased XES-standard conformity
   • progress bar for exporting enabled
10. commit 3b692b3
    • added parameter to enable/disable progress bar in importing in ITERPARSE
    • compression from file - XES compression moved from general entrypoint to single variant
11. commit 68ff7d0
    • more efficient importing of .xes.gz files
12. commit 4aad427, commit 67a42d2
    • get predecessors and successors of a dfg node (in dfg utils)
13. commit 6a946fb
    • allow to compute alignmnets directly on the dfg using dijkstra
14. commit 882468e
    • fix compatibility with pulp version 1.6.x
15. commit 50722bb
    • allow for DFG playout including Markovian probability of traces
16. commit 0888ab2
    • added minimum trace length in process tree extensive playout

PM4Py 2.1.2

This is a minor (christmas) release, consisting of the following changes:

1. commit 97cc315
   • add (de)serialization functionality for PM4Py objects
2. commit 96a7681
   • significant memory footprint reduction for iterparse-based event log importing
3. commit 8270a46
   • significantly faster 'line-by-line' xes exporter
4. commit aae4be3
   • advanced DFG filtering in (activities percentage, paths percentage) ensuring reachability from start and end
5. commit 2fb0b80
   • bug fix in BPMN layouting (nodes with same label)
6. commit d38f4c9
   • number of occurrences as a start or end activity is visualized in the DFG visualization

PM4Py 2.1.1

This is a minor release, consisting of the following chages:

1. commit 029d30f
   • allowing for visualizing and exporting BPMN models
2. commit f76dd37
   • added conversion to simplified interface
3. commit 43e7ce9
   • full conversion of WF-nets to BPMN models
4. commit 71c9eee
   • pull-request #185: reformatting of README file
5. commit b189e74
   • pull-request #187: memory-leak fix in xes importing

PM4Py 2.1.0.2

===== PM4Py 2.1.0.2 =====
This is a hotfix release, consisting of the following hotfixes:

1. commit f18c3e1
   • downgrading Numpy to 1.19.3 for Windows 10 2004 compatibility problems
   • skip blocking Scikit-Learn installation for Python 3.9

PM4Py 2.1.0.1

This is a hotfix release, consisting of the following hotfixes:

1. commit f6cdf1a
   • hotfix problem in XES importer parameters read

PM4Py 2.1.0

===== PM4Py 2.1.0 =====
This is a major release, consisting of the following major changes:

1. commit 6afd8ab
   • support for importing/exporting BPMN files;
   • supported elements: tasks, xor, and, or gateways
   • conversion of BPMN to WF-net
   • conversion of Process Tree to BPMN
2. commit a0a7fd0
   • bug fix in the inductive miner: sequence cuts were to maximal, leading to underfitting models (involving too many skips)
3. commit 848a161
   • conversion of dataframes to event streams / event logs now detects the use of XES Extensions

The following minor changes included in this release:

1. commit 8874f7b
   • add statistics on concurrent activities
   • discover eventually follows relations with concurrent activities
2. commit 3b8fcd4
   • apply invisible transition reduction on Petri nets obtained by Heursistic Miner
3. commit 0138f93
   • deprecated 'pm4py.write_csv' -> conversion and subsequent pandas export should be used.
4. commit 5b32725
   • fix use of deepcopy in event log conversion
5. commit 33103b5
   • fix get-variants behavior for event log / df (yielded different results)

PM4Py 2.0.1.3

This is a hotfix release.

PM4Py 2.0.1.2

This is a minor release, fixing the compatibility of the streaming package with Python 3.5.x - 3.7.x

PM4Py 2.0.1

This is a minor release, consisting of the following changes:

1. commit 81579e1
   • relaxing the importing of some dependencies (pandas, pulp, graphviz, intervaltree)
     to make a basic set of functionalities of PM4Py work even without those dependencies
2. commit ed45eaf
   • fixing circular dependencies issues and added partial compatibility with (old) Python 3.4
3. commit c51b1f0
   • parametrizing size of the thread pool in streaming package
4. commit b9d47b8
   • fix the live event stream class to finish all the events that are still in the internal queue when stop() is called
5. commit 8e07d84
   • add (unbounded) dfg discovery on an event stream
6. commit 4c115db
   • fix bug in hash function of traces
7. commit 16e3f7f
   • integration of the wf-net to process tree algorithm described in https://doi.org/10.3390/a13110279
8. commit 0dd8f28
   • fix bug in the alpha+ algorithm (was adding artificial start/end to the underlying event log)
9. commit 1a26f67
   • allow creation of 'live' streaming objects directly from xes and csv files
10. commit b08564e
    • start minor support for interval-based process models
11. commit 36cb713
    • added support for generic dictionaries in streaming conformance checking

PM4Py 2.0.0

This is a major release, consisting of the following major changes:

1. commit f408a18
   • add streaming token-based replay and footprint-based comparison
2. commit a9c5aa3
   • compatibility with Python 3.9 (limited to windows)
3. commit 7a89d4d
   • remove deprecated factories

The following minor changes included in this release:

1. commit 1a8f928
   • bug fix in PANDAS case size filter
2. commit fde08b0
   • output conformance checking diagnostics in a dataframe
3. commit 6280151
   • improved performance of token-based replay with duplicate labels