Pressure strain interaction

[plauri]

Added tentative data reducers for calculating pressure dilatation, Pi-D and pressure strain interactions. Very slow for large cell counts due to costly velocity Jacobian estimation. Future fixes should include boosting the efficiency and making the data reducers work for multi-population data (not implemented, as multi-pop data reducers do not support readers as of now)

[alhom]

Couple of comments:

• Can you refactor the reducers to request the [pop/]vg_v_jacobian instead? This would then enable 1) substituting in for a sidecar-jacobian when available 2) alternatively, using a short-term memoization decorator to avoid recalculations.
• add references for the reducers (like with (a)gyrotropy)
• non-pop reducer referring to population-specific values within the reducer is not great. Add v4-type calculation of center-of-mass velocity for that, for completeness?

This starts to feel like a bit of a rabbithole, and I kid of want to start adding kwargs-style dicts to pass to the readers to encapsulate the interp variable calls... but that I start to be pretty hesitant to add to try and keep reducers somewhat simple...

[alhom]

Plan: Souhail checks the results, then we merge with:

• add references for the reducers (like with (a)gyrotropy)
  implemented (@plauri), and fix the rest of it later, in the interest of having the feature available.