Collections

A collection is a data instance containing a set of (references to) instances and relationships between them and can contain (references to) other collections as well. It allows providing context to the instances and represent the knowledge of a domain and may e.g. describe where data can be found.

DLite collections are implemented as a list of RDF relations using the vocabulary defined in the datamodel ontology, which can be seen from the YAML representation of the collection entity:

    uri: http://onto-ns.com/meta/0.1/Collection
    description: Metadata description a collection.
    dimensions:
      nrelations: Number of relations.
    properties:
      relations:
        type: relation
        shape: [nrelations]
        description: Array of relations (s-p-o triples).

When instances are added to a collection, they are give a label local to the collection.

Semantics of a collection

TODO: add text when issue https://github.com/SINTEF/dlite/issues/174 is concluded…

Example

    >>> import dlite

    # Create some instances
    # Since the content of the instances are unimportant, we simply create a
    # Blob metadata (comes with DLite) and create some instances of it.
    >>> Blob = dlite.get_instance('http://onto-ns.com/meta/0.1/Blob')
    >>> blob1 = Blob(dims={"n": 2}, id="ex:blob1")
    >>> blob2 = Blob(dims={"n": 1}, id="ex:blob2")
    >>> blob3 = Blob(dims={"n": 5}, id="ex:blob3")

    # Create a collection
    >>> coll = dlite.Collection(id="ex:coll")

    # Add instances to the collection giving them a local label.
    >>> coll.add(label="blob1", inst=blob1)
    >>> coll.add(label="blob2", inst=blob2)
    >>> coll.add(label="blob3", inst=blob3)

    # Add relations between the instances
    # Here we express that `blob2` and `blob3` are parts of `blob1` as well
    # as that `blob3` is a proper part of `blob2` (note that collections does
    # no perform any reasoning, so it not able to infer that `blob3` must
    # also be a proper part of `blob1`).
    >>> coll.add_relation("blob1", "emmo:hasPart", "blob2")
    >>> coll.add_relation("blob1", "emmo:hasPart", "blob3")
    >>> coll.add_relation("blob2", "emmo:hasProperPart", "blob3")

    # Print the content of the collection
    >>> print(coll)
    {
      "uuid": "622d1aa3-4233-56bf-bab9-9b9d8c2aeab0",
      "uri": "ex:coll",
      "meta": "http://onto-ns.com/meta/0.1/Collection",
      "dimensions": {
        "nrelations": 12
      },
      "properties": {
        "relations": [
          [
            "blob1",
            "_is-a",
            "Instance"
          ],
          [
            "blob1",
            "_has-uuid",
            "a8645591-cc43-563f-8005-8ec63852ab6f"
          ],
          [
            "blob1",
            "_has-meta",
            "http://onto-ns.com/meta/0.1/Blob"
          ],
          [
            "blob2",
            "_is-a",
            "Instance"
          ],
          [
            "blob2",
            "_has-uuid",
            "3fa5a23e-e744-5dd6-94b5-28de06017bef"
          ],
          [
            "blob2",
            "_has-meta",
            "http://onto-ns.com/meta/0.1/Blob"
          ],
          [
            "blob3",
            "_is-a",
            "Instance"
          ],
          [
            "blob3",
            "_has-uuid",
            "f6cc739b-db7a-5fa1-808e-a6788033077d"
          ],
          [
            "blob3",
            "_has-meta",
            "http://onto-ns.com/meta/0.1/Blob"
          ],
          [
            "blob1",
            "emmo:hasPart",
            "blob2"
          ],
          [
            "blob1",
            "emmo:hasPart",
            "blob3"
          ],
          [
            "blob2",
            "emmo:hasProperPart",
            "blob3"
          ]
        ]
      }
    }