GNS-Science/nzshm-model

Odds and Ends

Opened this issue · 0 comments

chrisdicaprio commented

Some thoughts on how to improve the classes that have come up while working on them:

  • change member name in source classes from nrml_id to something that is not tied to openquake (be hazard engine agnostic)
  • do we need the type member of the source classes? It's possible to set them to something that doesn't match the class (e.g. "distributed" for a InversionSource)
  • should we even have DistributedSource and InversionSource? Should we just store any meta data (inversion id, etc.) in a class dict?
  • should we at least have a generic base class for sources to use when neither DistributedSource nor InversionSource fits the bill? We may, in the future, have inversions that also include the DSM
  • elevate SLT version2 to 1st class
  • move values to Branch base class (GMCM LTs also have a structure)?