Changelog
f8bdc18 - Add CI pipeline (Ihar Suvorau)
998d252 - Update pix-utils to pix-framework (Ihar Suvorau)
668778a - Update poetry github dependencies (David Chapela de la Campa)
200a127 - Update poetry github dependencies (David Chapela de la Campa)
b41c060 - Delete poetry.lock and update gitignore (David Chapela de la Campa)
20c639a - Refactor HeursticsThresholds class to generalize it as ConcurrencyThresholds (David Chapela de la Campa)
b1b0eb2 - Fix broken URL in README.md (David Chapela de la Campa)
ee1ac9c - Configure project with Poetry (David Chapela de la Campa)
6bfceb7 - Rename package (David Chapela de la Campa)
7050f32 - Fix small data in the error of a test log (David Chapela de la Campa)
c6e0c02 - Remove processing scripts and old configuration (David Chapela de la Campa)
3b03e25 - Remove submodules (David Chapela de la Campa)
78e9fe5 - Update OverlappingConcurrencyOracle to measure the concurrency ratio as the proportion of times two activities overlap in time w.r.t. the number of times they co-occur in the same trace (David Chapela de la Campa)
2508bfb - Update readme with instructions on how to get the enabling activity label (David Chapela de la Campa)
0a0a6fa - Add option to also add the enabling activity label (closes 3) (David Chapela de la Campa)
09159bf - Update submodule pix_utils (David Chapela de la Campa)
cf4b306 - Fix bug when estimating enabled times without start_time available (David Chapela de la Campa)
1dca546 - Refactor pd.DataFrame.groupby() call with only one element (David Chapela de la Campa)
67cb83c - Update pix_utils dependency (David Chapela de la Campa)
20ed985 - Add Overlapping Concurrency Oracle to the start time estimator options (David Chapela de la Campa)
ce661db - Implement overlapping concurrency oracle and tests - Consider two activities concurrent when the percentage of them happening at the same time (i.e. overlapping) w.r.t. the number of times they happen in the same trace, is greater than a specified threshold. (David Chapela de la Campa)
46ee295 - Adjust test and implementation to the latest change (David Chapela de la Campa)
47b5c06 - Estimate enabled time of start events to the start of the process case during the main pipeline (David Chapela de la Campa)
33c934d - Merge remote-tracking branch 'origin/main' (David Chapela de la Campa)
90fa0a7 - Refactor calendar resource availability and variable names (David Chapela de la Campa)
e831048 - Fixes mutable default for Configuration.heuristics_thresholds (Ihar Suvorau)
717ed41 - Change default value from variable to factory (David Chapela de la Campa)
e2a6f58 - Update setup.py and remove requirements.txt (David Chapela de la Campa)
1250c78 - Merge remote-tracking branch 'origin/main' (David Chapela de la Campa)
3f37160 - 2 Add working calendars to resource availability (closes 2) - Implement option to consider both resource contention and off-duty hours. - Implement tests. - Export read CSV log to pix_utils package. - Export EventLogIDs to pix_utils package. - Add pix_utils as submodule. - Update version. (David Chapela de la Campa)
8f84607 - Add working calendars to resource availability - Implement option to consider both resource contention and off-duty hours. - Implement tests. - Export read CSV log to pix_utils package. - Export EventLogIDs to pix_utils package. - Add pix_utils as submodule. (David Chapela de la Campa)
5b15eb8 - Update LICENSE holder name (David Chapela de la Campa)
221ae0c - Update LICENSE (Marlon Dumas)
9fdceaf - Update README.md (David Chapela de la Campa)
42639c4 - Upgrade version (David Chapela de la Campa)
5549ea4 - Refactor main (David Chapela de la Campa)
ea0b25f - Remove lifecycle from need IDs and remove write csv log function (David Chapela de la Campa)
0645b2f - Refactor Directly-Follows concurrency oracle (no concurrency) name (David Chapela de la Campa)
9a07ab3 - Refactor re-estimation functions to reduce number of dataframe access (David Chapela de la Campa)
425c8d1 - Refactor to compute RAT and ENT individually - Compute RAT for the full log with one function. - Compute ENT for the full log with one function. - Compute max of both columns for estimated start. - Always use pd.NaT as non estimated time and remove that parameter from the configuration. - Move option to replace current start times with estimated ones to estimate() function. - Add option to the enablement time calculation of a full log, to use trace start time of pd.NaT as enablement time of the first activity instances. - Update tests accordingly. - Minor refactors. (David Chapela de la Campa)
84279c3 - Add process case Idle/Processing analysis (David Chapela de la Campa)
6ae0b0d - Update README.md (David Chapela de la Campa)
ec9884a - Fixing setup.py: entry_points removed, because the package does not have them (Ihar Suvorau)
3b58681 - New version 1.0.0 - The start time estimation is done in a new column 'EventLogIDs.estimated_start_time' - The resource availability time and the enablement time are reported in two new columns. - There is an option to overwrite the start time column with the estimated start times. - There is an option to consider the already recorded start times in the enablement time and resource availability time calculations (ensuring these times are not after the recorded start). - There is an option to re-use the already recorded start times, and estimate only the missing ones (not tested). (David Chapela de la Campa)
66ebd98 - Refactor technique to consider start times during the estimation - If 'consider_start_times' is set to true in the configuration, i) the enablement time is calculated with the largest timestamp being previous to the end time of the current activity instance, and previous or equal to its start; ii) the resource availability is calculated as the largest end timestamp of the same resource being previous to the end time of the current activity instance, and previous or equal to its start. - If it is false, only the end times of the activities are considered (as if there was no start time information). (David Chapela de la Campa)
3d48b00 - Refactor technique to return, by default, an extra column with the estimated start times instead of replacing the recorded start times (David Chapela de la Campa)
ac63c75 - Update README.md (David Chapela de la Campa)
b112cf3 - Upgrade version (David Chapela de la Campa)
7a09b4d - Update setup file (David Chapela de la Campa)
b654e4b - Minor changes (David Chapela de la Campa)
ff2f114 - Update README.md (David Chapela de la Campa)
758d99e - Update README.md (David Chapela de la Campa)
ba1c26a - Update README.md (David Chapela de la Campa)
ecfd046 - Update README.md (David Chapela de la Campa)
2675d85 - Add analysis functions to inspect stats about original event logs (David Chapela de la Campa)
03230c2 - Update default XES IDs (David Chapela de la Campa)
8893830 - Refactor processing directory scripts (David Chapela de la Campa)
f7a8e0e - Fix HUGE bug with Wasserstein Distance (EMD) use - The scipy.stats implementation must receive the list of observations, not the histogram values! (David Chapela de la Campa)
e08e00a - Reset to start of the hours for the EMD calculation (David Chapela de la Campa)
84edec7 - Update metrics adding different EMDs and improving the cycle time EMD binning (David Chapela de la Campa)
12f2d6a - Combine simulated models to obtain EMDs with similar number of samples (David Chapela de la Campa)
142507d - Add metrics calculation for the simulated event logs (David Chapela de la Campa)
a55acfc - Add method to directly add the enabled times to an event log (David Chapela de la Campa)
063900b - Big refactor to remove use of PM4PY (David Chapela de la Campa)
50c8341 - Refactor start_time_config.py name back to config.py (David Chapela de la Campa)
72da72c - Big refactor to have a proper package structure and pytest working (David Chapela de la Campa)