Principled handling of growth advantages time units

[pawel-czyz]

Currently growth advantages are modelled on the scale where time is on the interval $[0, 1]$. These are not directly interpretable, so introduce mappings to weekly growth advantage, add settings for generation interval time.

Consider a bundle of settings.
(The scope of this issue is to be made precise.)

[dr-david]

Currently I rescale them using the TimeScaler t_min and t_max to get rates/day. Further settings could include:

• multiply per 7 to get growth/week
• multiply per generation time to get fitness

These rates can also be transformed for a discrete time model with exp(r)-1