Daniel Lakeland <dlakelan <at> street-artists.org> writes:
>
> On Fri, Jun 22, 2007 at 07:23:47AM +0000, Andrew Pinsky wrote:
> > I need to assume that f(x) >> r(x) + t(x). I.e. that f(x) + r(x) + t(x) ~=
> > f(x), and f(x) - r(x) - t(x) ~=f(x).
> If r(x) + t(x) never occurs except in conjunction with f(x) then your
> assumption is equivalent to r(x)+t(x) = 0.
> You can do ratsubst(f(x),f(x)+r(x)+t(x),expr) and the same for
> f(x)-r(x)-t(x). If this leaves you with no r or t left, perhaps your
> assumption is too strong?
r(x)+t(x) enters into some other expresions.
The hole expression looks like diff(q(x),t,2) * (f(x) + r(x) + t(x)) + ....,
and I resolved this issue by assuming diff(q(x),t,2) * r(x) = 0, and
diff(q(x),t,2) * t(x) = 0. ( diff(q(x),t,2)*r(x) and diff(q(x),t,2)*r(x) never
occur except this caseses).