Strange problem with pdiff



-----Rupert Swarbrick <rswarbrick at googlemail.com> wrote: -----

>Is there a "nicer" way of getting pdiff's functionality in Maxima?

Try avoiding pderivop an lambda forms and just do stuff like

 (%i9) diff(u(r * cos(t), r * sin(t)),t);
 (%o9)
 r*cos(t)*u[(0,1)](r*cos(t),r*sin(t))-r*sin(t)*u[(1,0)](r*cos(t),r*sin(t))

 (%i10) ratsubst(y, r * sin(t), ratsubst(x, r*cos(t),%));
 (%o10) x*u[(0,1)](x,y)-u[(1,0)](x,y)*y

Also, avoid using ev --- it can make code hard to understand.

Good luck, let us know if you get stuck. The pdiff code is hard for me
to look at; maybe for the version after 5.16, I'll do a rewrite
(algorithmically,
pdiff is simple).

Barton