discover_inductive_net
discovery::case_centric::inductive_minerSourceDiscovers an AcceptingPetriNet with the Inductive Miner, by converting the
runs
Reports progress as
discover_inductive_tree with ProcessTree::to_petri_net. Without a miner, the bindingruns
IMf at 0.2.Reports progress as
discover_inductive_tree.EventLogActivityProjectionrequired
InductiveMinerAn Inductive Miner with its settings: one that recurses on the log, or the
import r4pm
import r4pm.petri_net
# An EventLog works wherever an EventLogActivityProjection is expected.
log = r4pm.import_item("EventLog", "data/road-traffic-fines.xes.gz")
# Without a miner, the binding runs IMf with a noise threshold of 0.2.
imf = r4pm.bindings.discover_inductive_net(log)
# Variants: IM, IMf, IMa, IMd, IMfd. Presets: default, thesis, prom, pm4py.
imf_prom = r4pm.bindings.discover_inductive_net(
log, r4pm.bindings.inductive_miner_options("IMf", "prom", 0.1)
)
# IMfd recurses on the directly-follows graph instead of the log.
imfd = r4pm.bindings.discover_inductive_net(
log, r4pm.bindings.inductive_miner_options("IMfd")
)
for name, net in [("IMf", imf), ("IMf, ProM, 0.1", imf_prom), ("IMfd", imfd)]:
print(f"{name}: {len(net['places'])} places, {len(net['transitions'])} transitions")
r4pm.petri_net.export_pnml(imf, "model.pnml")process_mining/src/discovery/case_centric/inductive_miner/mod.rs:389