Yamamoto again

Aaargh after sending out last message realised that the R code is in fact
on the Cran site.  However, if anyone has an algebraic form that would
still be v useful as it's hard to get an overview of the process by looking
just at the code   Thanks Nick Wray