Closed caufieldjh closed 1 year ago
removing the is_a
to node property will fix, because node property is not multivalued
Removing the is_a: node property
here leads to the same TypeError: unhashable type: 'list'
Possibly an issue with how the yaml_loader
works, as linkml-validate
is much more helpful, as long as it's told about the target class of the data:
linkml-validate -s src/standards_schemas/schema/standards_usecase_schema.yaml -C UseCase src/data/examples/UseCase.yaml
DataSubstrate schema example doesn't validate yet because the schema doesn't accept a string as valid.
The metadata_storage
slot will need to accept something other than a URI.
(probably need to define with any_of
and an enum; see https://linkml.io/linkml/schemas/advanced.html)
The DataStandardOrTool
example isn't validating because:
Unknown StandardsCollectionTag enumeration code: Data Model
It needs to be datamodel
.
After that fix, the validation errors are:
Unknown argument: concerns data topic = ['STANDARDSDATATOPIC:4']
Unknown argument: has relevant organization = ['STANDARDSORGANIZATION:76']
(fixed)
Example data doesn't quite parse yet - when hitting the multivalued
relevance_to_dgps
slot, it hits a snag:This could be due to the DGPs being modeled as an enum rather than specific entities, so there may be a solution involving inlining.
This slot definitely needs to be multivalued, though - and that seems to be the core of the error, as
self.relevance_to_dgps = DataGeneratingProject(self.relevance_to_dgps)
doesn't seem like it would allow for that.