note.txt

OK, so we really have

Generic
Herm (s* basic)
HermID (s* composite)
Andro (s* basic, p_m only affects N*/N)
AndroID (s* composite, p_m affects s*)
GynoID (s* composite, p_f affects s*)

We can change a to s~.

Change C to R.

Mention that people should use the Generic model (under Gyno) if
they don't have tau.

Q: Why is the generic model different than the Herm model?

The Herm model prints N*/N, but the generic model doesn't.

If you don't have info about the Gyno model, using the Herm model
(which is wrong) and setting tau to 1 (which is wrong) and printing
N*/N=1 (which is wrong) would be weird.

Q: Can you do a Gyno model w/o ID?
A: No.  Because the idea is that you could place a uniform prior on s*,
but then s* would not be composite anymore.

A: Yes, you COULD if you can information on everything that effects N*/N,
if you ONLY want to estimate N*/N.