[R-sig-eco] package plspm: why rounding to 4 digits?

[Arne Erpenbach]

Dear all,

I am using partial least squares regression on an ecological data set,
comparing different packages. The implementations are quite different,
which is interesting enough.

One minor thing: in package plspm, several functions internally use
rounding to four digits. For an example, take a look at the source of
plsreg1{plspm}.

Q: Does anybody know a reason for this? (Either ecological,
chemometrical or mathematical?)

I tried to contact the author, no response so far.
Assume he's quite busy, so I though I'll ask the community.


Cheers,
arne