every atom is complex?



Am Dienstag, den 30.11.2010, 11:38 -0800 schrieb Valery Pipin:
> Hi,
> 
> I've perhaps missed something. I observe that now the every atom is complex by 
> default. I think it is not very suitable, e.g., I need to get the realpart of 
> tensor expression. To make the realpart() works I have to declare everything 
> real. It also does not work for special tensors like 'levi_civita because it 
> is with coma ahead. The way around is to make subst([%i=0], expand(expr)).
> As I remember, 2 or 3 years ago there was not default complexification.
> I discovered it because my previous code does not work on the current maxima 
> version. I wonder if there is a switch to make all atoms real by default?
> 
> best wishes
> Valery

I think the behavior that all atoms are by default real has not changed:

(%i1) featurep(a,complex);
(%o1)                                false
(%i2) imagpart(a);
(%o2)                                  0
(%i3) rectform(a);
(%o3)                                  a
(%i4) cabs(a);
(%o4)                               abs(a)

There might be another change, which is responsible for a different
behavior of your code. Can you show more examples.

Dieter Kaiser