Re: getting strange ellheight() error for number field point and curve

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

On Mon, Dec 11, 2023 at 03:24:09PM +0100, John Cremona wrote:
> Here is one:
> 
> https://www.lmfdb.org/EllipticCurve/2.2.85.1/180.1/o/1

Thanks ! So with PARI 2.15.4:

K = bnfinit(Polrev([-21, -1, 1]));
E = ellinit([Polrev([0,1]),Polrev([0,0]),Polrev([1,0]),Polrev([-2154,423]),Polrev([45890,-8980])], K);
P=[-13* x + 69 , 104* x - 548 ];
ellisoncurve(E,P)
ellheight(E,P)
%4 = 1
%5 = 0.32253134747871300961804811416751832019

which is in accordance with the height given in the LMFDB, 
however this version gives a wrong result for 20.P:

? ellheight(E,ellmul(E,P,20))/20^2
%5 = 0.31703828603537246116107188798290569166

(the master branch give the correct result).

I will check the other curves later!

Cheers,
Bill