Annotating nhoods by continuous covariates

[emdann]

From discussion with Pe'er lab members: to aid exploratory analysis make functions (and show in vignettes) to relate DA logFC to average value for some continuous covariate in nhoods (e.g. pseudotime, latent factor values, gene signature expression).
Possibly simply extending annotateNhoods?

[MikeDMorgan]

I think we could also have some bespoke plotting functions to aid this for continuous variables, akin to the beeswarm plots we have for discrete variables.