We are beginning to get corrections to typing of images from major datasets. Rather than silently changing these, we have agreed that the old typing should be retained along with metadata. This needs to be outside of the indexing/query system - which can be most easily achieved by using and annotation property.
Function should take a list of tuples as an arg. Tuples have: individual ID, type ID (FBbt), comment. It should then convert the type edges specified to an annotation property 'oio:obsolete_type', annotated with the comment (rdfs:comment).
Adding new, corrected typing should be a second step using existing pipelines.
Limitation: If we get a major version update on a dataset, it would be useful to be able to assess what has changed in order to determine what to obsolete. This could be achieved via a function that wraps this one (which should be the subject of a separate ticket).
We are beginning to get corrections to typing of images from major datasets. Rather than silently changing these, we have agreed that the old typing should be retained along with metadata. This needs to be outside of the indexing/query system - which can be most easily achieved by using and annotation property.
Function should take a list of tuples as an arg. Tuples have: individual ID, type ID (FBbt), comment. It should then convert the type edges specified to an annotation property 'oio:obsolete_type', annotated with the comment (rdfs:comment).
Adding new, corrected typing should be a second step using existing pipelines.
Limitation: If we get a major version update on a dataset, it would be useful to be able to assess what has changed in order to determine what to obsolete. This could be achieved via a function that wraps this one (which should be the subject of a separate ticket).