Re: Two suggestions about elldata

[Bill Allombert <Bill.Allombert@math.u-bordeaux1.fr>]

On Fri, Jan 20, 2006 at 08:51:09AM +0000, Prof. J. E. Cremona wrote:
> A small adjustment to Bill's suggestion might be easier to use:
> 
> forell(E,N=N1,N2, seq )
> 
> where simultaneously N runs through conductors from N1 to N2 and E is
> assigned to the successive curves of conductor N in turn;  moreover E
> should be assigned ot the full data in the database, so
> E=[E[1],E[2],E[3]] = [id string,  coeffs,  points].

I am afraid that will be more cumbersome to use than my proposal
(especially if you only interested in say conductors multiple of 3).
May we could keep both the orginal forell and forellcond ?

> I suggest also keeping the single string for the id, but providing a
> little utility function as in my first posting to expand the id code
> into a triple [N,"class",num].

How should we call it ? Should we return the class as a string or an
integer ? (a->0 b->1, c->2, etc.) ?
Should the function also for backward taking a vector '[N,class,num]'
and returning the name ?

In any case, this is easy to implement, that is not the issue!

Cheers,
Bill..