discover_path_schemas
analysis::object_centric::path_schemasSourceDiscover path schemas between the query types: enumerate schemas, summarize each schema's
connections, score it by support, coverage, selectivity, reach, exclusivity and throughput, and group connection-equivalent schemas.
The source entities (shared by every enumerated schema) are collected once. Each schema is summarized in a single streaming pass (see summarize_connections) that never materializes the connection list, and connection equivalence is decided by a 64-bit hash of each schema's (source, target) set rather than by retaining the sets. Reports progress to a Control in two stages: "Enumerating path schemas" (see Control::enumerate_schemas), then "Summarizing path schemas", where done counts the source entities scanned over all schemas, out of the number of schemas times the number of source entities.
Streams each schema as soon as it is summarized: one partial value is one DiscoveredSchema, in index order, as in PathSchemaDiscovery::schemas. Only its equivalence_class is provisional: the index of the first schema with the same connection set so far (its own index if none), which the result replaces with an index into PathSchemaDiscovery::equivalence_classes.
A discovery query: source/target types, max schema length, and connection params.
Result of [`discover_path_schemas`].
import r4pm
# Also: .json, .json.gz, .xml.gz, .ocel.csv, .ocel.csv.gz, ...
ocel = r4pm.import_item('SlimLinkedOCEL', 'ocel.xml')
query = {
"allow_cycles": False,
"max_length": 1,
"params": {"dedup_targets": False, "selection": 'All', "temporal": 'None'},
"source": {"Event": 'example'},
}
res = r4pm.bindings.discover_path_schemas(ocel, query)