[Maxima-discuss] complex numbers was Re: A weird bug

[Igor P. <ipe...@gm...>]

On 13/04/25 17:06, Stavros Macrakis wrote:
> On Sun, Apr 13, 2025 at 8:40 AM Igor Pesando <ipe...@gm...> wrote:
>
>     ...(I have cited it in my papers).
>
>
> Thank you for citing Maxima!
>
>     Obviously I have found a number of small bugs (how should I report
>     them).
>
>
> We welcome bug reports for both small and for big problems -- please 
> create an account on sourceforge.net <http://sourceforge.net> and then 
> you can reports bugs at https://sourceforge.net/p/maxima/bugs/new/ .
>
>     In particular I had many issues in plotting complex functions (issues
>     with complex numbers numerics) which obliged me to use maxima in
>     doing
>     symbolic simplifications and gnuplot to do the real plotting.
>
>
> Can you give us a small example of the problem?
>
>         -

I was trying to plot the real part of a complex function but without 
success.

I traced back the issue to (maybe I missed one of the many flags maxima 
has but it is not what I would expect)

qq:ev(0.9900829565854907*(10.123456 - 0.001*%i) + 
3.920663652683926,numer); => not a single complex number

ww:expand(0.9900829565854907 (10.123456 - 0.001 %i) + 
3.920663652683926); => it works

similary

qq:log_gamma(0.9900829565854907*(10.123456 - 0.001*%i) + 3.920663652683926);
ww:ev(qq,numer); => not working
ee:map(expand, qq); => work

but if you try to plot log_gamma of an algebraic function with some real 
numbers which are not rational is a mess.

Obviously you can make a list of the values and use plot2d discrete.

But then for the function I was trying to compute the computation takes 
hours while in gnuplot few seconds.

I can send the maxima file which I batch if it is necessary.

Thanks in advance,

Igor P


Igor