Closed lambdatronic closed 12 years ago
You made ontario:AllWaterData (and others) an observation:Observation - which in thinklab triggers the creation of the Java peer object linked to observation:Observation, which obviously has several semantic constraints including that of having a functional relationship with an observable. Thinklab creates an instance of whatever observable you pass, and what you see is the result of validating that instance.
One of the main tenets of the approach (as you presented in Vienna) is that observations and observables are orthogonal, so you shouldn't make an observable an observation. Just make it whatever you feel like, this is just the concept that identifies a collection of observations so even a owl:Thing will do.
@fvilla @lambdatronic @bvoigt
AFAICT, all my source, sink, use, and routing models in core.models.sediment-ontario and core.models.water-ontario are running correctly at the Thinklab shell. However, when I build an identification observation containing all of them (except the probabilistic-measurement BN "undiscretizers"), I get an error that appears to be popping out of my local PostGIS database.
Here are the two trouble-makers: