opengeospatial / om-swg

10 stars 6 forks source link

Observation Collection #2

Closed KathiSchleidt closed 3 months ago

KathiSchleidt commented 5 years ago

The need for a class for collections of observations, with optional values for all of the observation properties – to support sets of observations that have common values for one or more of procedure, feature-of-interest, observedproperty, phenomenon-time, result-time

Source: TC211 Comments AU001, PMG009

KathiSchleidt commented 1 year ago

9.10.2 AbstractObservationCollectionType moved to 9.10.2, modifed req from obs-basic to obs-core

10.4 ObservationCollection content moved to 9.9 AbstractObservationCollection, modifed reqs from obs-basic to obs-core, modified ObservationCollection to AbstractObservationCollection. Added Note that described in Figure 16 (under 9.9.1)

9.1.1 Abstract Observation Core Package Requirements Class: added Imports /req/obs-core/AbstractObservationCollection

10.4 ObservationCollection updated, only requirement in Req Class is /req/obs-core/AbstractObservationCollection, removed subclauses (all under 9.9)

What did I miss?

ilkkarinne commented 1 year ago

Figures impacted by the change:

ilkkarinne commented 1 year ago

Suggested fixes for requirements text due to changing the collection - memberCharacteristics association from between ObservationCollection and ObservationCharacteristics (in Basic) to between AbstractObservationCollection and AbstractObservationCharacteristics (in Core):

KathiSchleidt commented 1 year ago

Done! Also fixed the Req Class tables to account for shifting collection-sem to abstract

sgrellet commented 1 year ago

@KathiSchleidt , @ilkkarinne : while cross-checking there is something that bothers me (pretty sure that was discussed before but I prefer asking)

Why are we traversing abstraction levels ?

sgrellet commented 1 year ago

Also missing:

KathiSchleidt commented 1 year ago

@sgrellet on the traversing levels, think this has to do with the fact that we don't have all OMS FT at the conceptual level (no interface for Collection or Characteristics)

However, your note on C.8 just created a whole new nightmare for me - we now have an AbstractObservationCollection, but no AbstractSampleCollection (once again proving that we're the first to quietly ignore Samples!!!)

sgrellet commented 1 year ago

@KathiSchleidt : on the traversing levels, your analysis is right. I just fear that other (and esp. newcomers to OMS) may be puzzled by this.

On C.8 : sorry :(

dr-shorthair commented 1 year ago

we now have an AbstractObservationCollection, but no AbstractSampleCollection (once again proving that we're the first to quietly ignore Samples

Part of the challenge here is that we know about ObservationCollection - which is a collection of activities, while SampleCollection is a collection of entities (using Prov-o terminology), and we think less about SamplingCollections (the matching collection of sampling activities).

In practice I have usually found the need for a generic ActivityCollection, which is an umbrella for a set of Observations and Samplings and Treatments (which are a kind of Actuation). Project and SiteVisit are often sub-classes of ActivityCollection.

Probably too generic and too late for OMS, but a bit of perspective.

sgrellet commented 1 year ago

`In practice I have usually found the need for a generic ActivityCollection, which is an umbrella for a set of Observations and Samplings and Treatments (which are a kind of Actuation). Project and SiteVisit are often sub-classes of ActivityCollection. Probably too generic and too late for OMS, but a bit of perspective.'

Thanks Simon for this. I also tried to push for MonitoringAcitivity during the revision process but failed. It was more a collection of Observations & Samplings without Treatments. But when reading your description I now feel adding Treatment/Actuation in the package would make a lot of sense as we also do have equivalent such Activities here. I'll create an issue about to avoid forgetting this part.

ilkkarinne commented 1 year ago

Raising this as to not forget to do it:

  • updating Table 2 — Abstract Observation core Core conformance classes and missing the corresponding entry in A.2 Abstract tests for Abstract Observation core Core package : NOT Done

Proposed fixes:

ilkkarinne commented 1 year ago

Proposed fixes:

  • Add new clause A.2.10 Abstract Observation Core – AbstractObservationCollection with a table for Conformance Class /conf/obs-core/AbstractObservationCollection
  • Add a new row in Table 2 — Abstract Observation Core conformance classes for /conf/obs-core/AbstractObservationCollection

These have been added in the Word document with commit 565e9bcda60f02cfafa24ea85bc302828c2e999d

ilkkarinne commented 1 year ago

Cross-checked by @KathiSchleidt, moving into Implementation cross-checked