Add `N_clust_full` to `e()` scalars

[MRuzzante]

reghdfe stores three types of scalars for the number of observations: e(N), e(num_singletons), and e(N_full).

It would be great to have the same structure for the number of clusters, which currently only incldes e(N_clust), and allow the user to show the _full option for both the number of observations and clusters.

PS: Many thanks for the great command! :)

[sergiocorreia]

Hi Matteo,

Makes sense. I'm in a bit of crunch time but will try to take a look at this soon (and a few of the other issues!)

Best, S