-
Notifications
You must be signed in to change notification settings - Fork 4
Meeting Notes 22.10.2021
Standardize setup and testing in dic2owl.
James – Strategy.
Bottom-up approach.
Instead of starting from defining a "Crystal", let's instead start from specific CIF files and its contents to see what ontological individuals are created from the key/value-pairs in a CIF file, ontologize this, and then move further up.
Matthias: Combining bottom-up and top-down works well in practice. He had an example from the EU-project INTERSECT, using AiiDA and integrating it with the ontology and other software.
Rickard brings the subject back to a previous discussion regarding dimensionality: What is the lowest number of dimensions that still constitutes a crystal?
Saulius: It should have periodicity and have an ideally delta-like function if Fourier transformed.
Maybe name anything less than 3 dimensions as "periodic systems"?
CIF/DDL has no requirement to specify all 3 dimensions. While CIF does not require 3 dimensions, other properties of CIF might assume that 3 dimensions are given. It would be nice to see if the ontology can check whether other CIF properties require 3 dimensions.
Launched into a discussion based on the EMMO definition of "Material" concerning whether a Crystal should or should not be a Continuum as defined by EMMO. It seems EMMO defines these terms with respect to the properties, not scale/size.