NGAEastUSGSGMPE not working outside of unit tests

[CB-quakemodel]

The core NGAEastUSGSGMPE does not appear to work outside of the unit tests. Given a gmpe_table is not required anymore for the core gsim (based on the class docstring), specifying the gmpe alone as within the screenshot should work. I have tried in the area classical PSHA and this generates the following error:

Thanks,

Chris

[micheles]

It could be that the docstring is wrong. I see it was written 5 years ago by Graeme. I thought only subclasses of NGAEastUSGSGMPE could be used, the ones listed at the end of the file.

[g-weatherill]

It's been a long time since I did this but I don't think I intended for the NGAEastUSGSGMPE to be used directly as a class in itself. So Michele is right. It didn't make any sense to use it as a standalone GMPE, as the operation of the GMPE is specifically for what I recall was the 2018 US NSHM set of CEUS GMPEs, so it must be associated to one of the tables that are given in the aliases. I guess I hadn't precluded the possibility that one could pass a different table of ground motions than one of those used by the USGS, it just wasn't developed with that in mind.

[CB-quakemodel]

Thanks for the clarification. I misinterpreted from the docstring that we could instantiate the entire 2018 USGS CEUS GMPE set as would be obtained from the gsim logic tree of the stable shallow crust branch of the USA Mosaic model by not specifying an individual gmpe table, rather than this referring to the use of add_alias making specifying a gmpe table unnecessary. I'll open a very small PR clarifying the docstring and link it