From: John Peterson <peterson@cf...> - 2004-05-07 19:56:02
Jose Geraldo A. Brito Neto writes:
> The way the minimum method is implemented int the latest CVS version of
> the ErrorVector class, it returns 1.0e30 for a vector full of zeros. Is
> this behavior ok?
Er, no it doesn't sound very ok :) I will look into it, thanks.