No, I got hung up trying to do the version in the draft of BDA.
Aki rewrote the section and suggested radial basis functions.
I'm happy to code a model and get it working on fake data and
add it to the manual if someone can point me to an appropriate
definition. Seems to be another of these black arts of choosing
bases.
- Bob
> --
> You received this message because you are subscribed to the Google Groups "stan development mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to
stan-dev+u...@googlegroups.com.
> For more options, visit
https://groups.google.com/groups/opt_out.
>
>