[Axiom-developer] Re: expand sin (nx) and cos (nx)

[Martin Rubey]

Dear Francois,

Francois Maltey <address@hidden> writes:

> I hope I can compile it without problem later inside the MANIP.spad

yes.

> At the end of this input file you can try : 
> 
> expand1 (cos (3*x))+%i*expand1 (sin (3*x)) - (cos(x)+%i*sin(x))^3
> expand1 (cos (a+b))
> expand1 (sin (a-b))  -- doesn't work
> expand1 (sin (b-a))  -- is right
> 
> So I have several problems :
> 
> // A //
> 
> My first function is a function which gives the name of the function :
> 
> With mupad I done :
> op (3*x+5*y, 1)   -> _plus
> op (3*x, 1)       -> _mult
> op (x^3, 1)       -> _power
> op (sin(3*x), 1)  -> sin
> op (cos(3*x), 1)  -> cos

in axiom there are operations isPlus, isTimes and so on. Maybe this helps. In
fact, it seems to me that you only need to test whether your expression is of
the form sin(something) or cos(something). This can be easily done with the
domain PATTERN and the operation isOp

For example, try:

isOp(sin(x)::PATTERN INT)$PATTERN INT

or 

s := (isOp(sin(x)::PATTERN INT)$PATTERN INT).op

(I just forgot how to get the operator "sin")

isOp(sin(x+y)::PATTERN INT, s)$PATTERN INT

> // B //
> 
> Why must I add some coerces  in this input file

the coercions b := ... : EXPR INT are necessary, since "numer" returns a
SparseMultivariatePolynomial. If you want to get an Expression, try "numerator"
instead. However, in the EXPR domain there is no leadingMonomial. So the
coercion is necessary.

> // C //
> 
> What are the usual Expression ... type.
> 
> Can I be sure that coef in Integer, Fraction Integer, Complex Integer ?
> Is it possible to have Float or DoubleFloat...

No. There is no usual EXPR ... type. Well, Boolean does not work well. Float
does, though.

> How can I test if variable coef in an Integer or not ? 

Looking at your code, coef is of type R if arg is of type EXPR R. Thus if R is
INT, you are done. If R is something else you have to use 

(retractIfCan coef)@Union(Integer,"failed")




Sorry that I'm answering late, but I'm out of time currently...

Martin