Tim Taylor

This may no longer be needed since the 2.0.0 refactor. Need to check.

Added in https://github.com/reconverse/incidence2/commit/43d75e6034440453dde70c35a5ad7b3839395993 but still requires tests.

Still need to add example to vignette but function is at least now mentioned (https://github.com/reconverse/incidence2/commit/5ab71af0b10a45fdcb6140028bada4d681cb2356)

Function renamed and updated via https://github.com/reconverse/incidence2/commit/1ef0420a031462eb252d69046fd6fc8f8f1b6c82

closed by https://github.com/reconverse/incidence2/commit/085a715100fb1d37a821948061660b4988cdfacf

Cool. Will try and PR something over the next couple of weeks.

@avallecam - This is now changed in the development version and associated docs. Relevant section rendered in dev docs at https://www.reconverse.org/incidence2/dev/articles/incidence2.html#computing-incidence-from-pre-aggregated-data if you want a gander. I've implemented the warning...

If this is still required I've started packaging it up over at https://github.com/tjtnew/bedoc. A little hectic today but can probably get something reasonable finished by c.o.p tomorrow if it's still...

A few questions: * What are you trying to avoid with the sanity check `n_admissions[1] >= 1`. A user may want to illustrate the ramp up from zero cases over...

I've thought some more in regards to the second and third bullets above. I think it would make sense to have `simulate_occupancy` as an internal function and only export `bed_forecasts`....