Bill Allombert on Thu, 16 Dec 1999 21:21:46 +0100 (MET)


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: poldegree bug?


There are two commonly used definition of the degree of a rational
fraction  R=P/Q

The first is deg P-deg Q, the second is max(d°P,d°Q)

PARI return the degree as in the first definition.
so
 poldegree(1/x^2)==-2 
 poldegree(1/(x^2+1))==-2
poldegree(1/x^2+1)==poldegree((x^2+1)/x^2)==0

So the result is correct.

Did you eventually mistype the parens ?it happens...

But the specifications say that the degree of the zero polynomial is -1,
that is strange:
poldegree(1/x)
-1 
 poldegree(Pol([]))
-1
 poldegree(0)
Well, I think ~VERYBIGINT would be cleaner for 0 polynomials.

In fact this is done to match the behaviour of the library function
"degree" which apply only to POL and simply return lgef(P)-3, so
degree(Pol([]))=2-3=-1
   
Beside, there is no polvaluation fonction, but we can use
valuation(P,x) instead.

Bill.