Absolute error

Take a value and an approximation to this value, . the absolute error is then given by .E.g.

> p:=exp(0.6);

> pa:=subs(x=0.6,sum(x^i/(i!),i=0..3));

> abserr:=abs(p-pa);