Consider exposing mean and sigma calculations in GMMs

There are multiple contexts where gaining direct access to the mean and sigma of a published GMM would be helpful without the wrapping and unwrapping to/from a LogicTree<GroundMotion> obtained via GroundMotionModel.calc(GmmInput):

• PRVI backbone GMMs where only the means of the NGA models are required #251
• Combined GMMs
• Custom GMMs for stable crust to handle distance-dependent logic trees However, this introduces some complexity. calcMean() and calcSigma() methods could be added to the GroundMotionModel interface with default implementations, but would this mean returning published model mean and sigma, or collapsed epistemic uncertainty branches if they exits. NGA-East has no published mean model, one has to collapse the epistemic branches to compute a mean. There could be too much room for confusion.

added GMM label

mentioned in issue #263 (closed)

Placeholder notes; closing for now.

closed