Function to calculate dominant eigenvector of next generation matrix

As well as `r_eff()` for effective R (i.e. the dominant eigenvalue of the next generation matrix), it can be useful to have a complementary function that calculates the dominant eigenvector, which indicates the expected distribution of new infections across multiple transmission generations (i.e. once any transient distribution of introduced infections has faded).

This would make it possible to produce plots like the below from [Klepac et al (2020)](https://www.medrxiv.org/content/10.1101/2020.02.16.20023754v2), which can be used for planning and situational awareness (e.g. if observed incidence does/doesn't match expectations of what transmission driven by physical contact would produce).

Start by locating r_eff() and the code that constructs the next generation matrix, then determine how its dominant eigenvalue is calculated. Add a complementary function that returns the dominant eigenvector and verify that it represents the expected distribution of new infections across transmission generations.

Written by the indexing model from the issue text.