Handle the index for a node not being an integer (e.g. "index=NA")

[bredelings]

Fixes #400

[davidcerny]

This all looks good to me. I assume lines 1433–1436 in Tree.cpp are for exceptions that we can't handle using RbException but that can still be handled using the standard library?

Also, that's interesting that dnMultiSpeciesCoalescentInverseGamma is parameterized using shape and scale rather than shape and rate (which is how dnInverseGamma does it). To prevent confusion, it might be worth calling these arguments kappa and theta rather than alpha and beta, maybe?

[bredelings]

This all looks good to me. I assume lines 1433–1436 in Tree.cpp are for exceptions that we can't handle using RbException but that can still be handled using the standard library?

Right.

Also, that's interesting that dnMultiSpeciesCoalescentInverseGamma is parameterized using shape and scale rather than shape and rate (which is how dnInverseGamma does it). To prevent confusion, it might be worth calling these arguments kappa and theta rather than alpha and beta, maybe?

So, that change just updates RbHelpDatabase to match the *.md files. The change in the parameters has already been made (by someone else) -- this just updates the documentation.

I do agree that there is some confusion about the parameter names. Usually I use alpha and beta for the shape and scale, but I see that wikipedia suggests k and theta. I have more opinions, but I suppose the main thing is that we should be consistent. Buuuuut... other people made these changes and I don't have time to make all the parameterizations consistent right now.

[bredelings]

Thanks!