Hi,
On Thursday 27 March 2003 18:43, Andrei Zorine wrote:
> Hi,
> No, it's not. I have *equations* defining calM(v,r) for
> r=1,2,..,m.
This is not a problem here. You can generate your lists and then
redistribute "=" via map("=",List1,List2)
>In fact, I have one special eqiation for
> calM(v,m+1) (like calm(v,r)=sum(w[r](v)*calM(v,r),r,1,m)).
So, you probably should reformulate the problem via using
recurrences. See packages in share/algebra. If m is unbound you
can not use sum
>To
> solve different problems in my research I have to
> differentiate once or twice the equations, evaluate at some
> v=v0,
No problem,
(D14) q[r](v):=l[r]*v^2+tt[r]*v
(C15) diff(SU[2],v);
(D15) p[2]*PI[n]*q[2](v)*'DIFF(RQ[2](v),v,1)
+p[2]*PI[n]*RQ[2](v)*'DIFF(q[2](v),v,1)+PHI[2](v)*'DIFF(q[2](v),v,1)
+q[2](v)*'DIFF(PHI[2](v),v,1)+p[1]*PI[n]*q[1](v)*'DIFF(RQ[1](v),v,1)
+p[1]*PI[n]*RQ[1](v)*'DIFF(q[1](v),v,1)+PHI[1](v)*'DIFF(q[1](v),v,1)
+q[1](v)*'DIFF(PHI[1](v),v,1)
(C16) ev(%,diff);
(D16) p[2]*PI[n]*(tt[2]*v^2+l[2]*v)*'DIFF(RQ[2](v),v,1)
+(tt[2]*v^2+l[2]*v)*'DIFF(PHI[2](v),v,1)
+p[1]*PI[n]*(tt[1]*v^2+l[1]*v)*'DIFF(RQ[1](v),v,1)
+(tt[1]*v^2+l[1]*v)*'DIFF(PHI[1](v),v,1)
+p[2]*PI[n]*(2*tt[2]*v+l[2])*RQ[2](v)+(2*tt[2]*v+l[2])*PHI[2](v)
+p[1]*PI[n]*(2*tt[1]*v+l[1])*RQ[1](v)+(2*tt[1]*v+l[1])*PHI[1](v)
>add them up, then do exclusions and so on. I need to
> operate with finite sums with unknown but constant upper
> limit.I don't *define* any functions/procedures (calM, Phi
> are not procedures or maxima functions, they're expressions)
Well, in your notations calM(v,m+1) is clearly a kind of
function. Actually it is better to write it as calM[m+1](v) as
a vector function.
rgds,
v