This is not an issue. The numerical approximation function N simply
evaluates the expression at a given precision. There is no guarantee
on the overall result as cancellation may occur. If you want accurate
results, use interval or ball arithmetic. If you note your symbolic
expression "expr" you c
When I run the following code on sagecell.sagemath.org, there seem to be
problems in evaluating an expression I have generated in the course of my
investigations:
for k in range(1,101):
print k,N(1/6583885550*(1212730912710731739903840*e^2 -
1209871310179356975377887)*e^(2/3) -
1/658388555
> On May 8, 2019, at 06:14 , Santanu Sarkar
> wrote:
>
> I know how to define variables over BooleanPolynomialRing.
> This is as follows.
>
> n=4
> V=BooleanPolynomialRing(n+1,['z%d'%(i) for i in range(n+1)] )
> V.inject_variables()
>
> Can we define similar code over integers (ZZ) or ratio
