From: Erik H. <eh...@gm...> - 2009-03-26 20:59:35
|
> Looks like this behavior is due to the way the imaginary > part is computed for the product of complex numbers. > The following patch changes it so the loop yields zeros. > > I ran into this behavior in the Maxima test suite; > there are several failed tests which seem to be due to it. Fixed. Looks there' |