Euphonic Calculations: Refactor Abins/Mantid to use Euphonic

[ajjackson]

Blocking #24

• Refactor Abins to use Euphonic for incoherent powder averaging in the analytic "almost-anisotropic" approximation. This shouldn't be too problematic if the refactoring work to support Euphonic Calculations: Implement analytic powder-averaged incoherent spectra in Euphonic #67 has already been done.
• Arrange for an appropriate release of Euphonic to be included-in/distributed-with Mantid. My understanding is that this will require somebody at ILL to manually do a pip upgrade whenever the packaged Euphonic needs updating, so we should avoid doing that too often. Once per Mantid release, say?

[ajjackson]

The current roadmap is to

• integrate Euphonic with the Mantid SimulatedDensityOfStates algorithm (and settle packaging/distribution issues)
• Integrate Euphonic with Abins to import force constants data and generate frequencies/eigenvectors on-the-fly

Then it will become possible to develop #67 , which is the first stage of this issue.

So the original check-list should be completed in reverse order!

[ajjackson]

Update: the Euphonic/SimulatedDensityOfStates development has been implemented and is just awaiting a bit of snagging re: reliable installation of Euphonic on Mantid systems. mantidproject/mantid#30760 This could be greatly aided by wheel builds of Euphonic.

I have made a start on the next step, importing force constants into Abins using Euphonic. Packaging issues aside, it seems to already work fine: mantidproject/mantid#32220