Re: Vectors and integer exponentiation

[Bill Allombert <Bill.Allombert@math.u-bordeaux.fr>]

On Mon, Oct 03, 2022 at 11:52:29AM +0200, Martin Becker wrote:
> Hello,
> 
> In GP (tested with versions 2.15.0 and 2.13.1), the expression
> 
>     2^[4,6,8]
> 
> evaluates to a vector of floating point numbers, while
> 
>     [2^n|n<-[4,6,8]]
> 
> evaluates to a vector of integers.
> 
> I would have expected integers in the first case, too,
> or perhaps a type error if exponentiation was not meant to
> auto-unbox vectors.  Is this a bug?

No, this is the consequence of historical design before [<-|] was introduced
that allows transcendental functions to act component-wise on vectors:

? exp([1,2,3])
%6 = [2.7182818284590452353602874713526624978,7.3890560989306502272304274605750078132,20.085536923187667740928529654581717897]

and it happens that 2^[4,6,8] 
do the same thing as
exp(log(2)*[4,6,8])

Cheers,
Bill