Re: First non-trivial PARI/GP script from chatgpt session

On 2025-10-12 22:06, Bill Allombert wrote:
> On Sun, Oct 12, 2025 at 05:35:48PM +0200, [email protected] 
> wrote:
>> vertex_count(n) = 12 * qfbclassno(-4 * n);
>> 
>> \\ Function to analyze face data
>> analyze_faces(n, face_counts) =
>> {
>>   V = vertex_count(n); \\ number of vertices
>>   F = 0; E2 = 0;
...
>>   print("Euler characteristic (should be 2): ", euler_check);
>>   print("Average face degree = ", avg_face_degree);
>>   print("Average vertex degree = ", avg_vertex_degree);
>> }
> 
> This is overly verbose and it does not apply GP scoping rules
> correctly. Frankly this does not deserve a passing grade.
> 
It was the script that surprised me, but even more that chatgpt proposed 
to run this script to determine the avererage face length.

> Please do not use qfbclassno! quadclassunit is almost always a better 
> choice.
> 
I take your word, instead
classno(n)=qfbclassno(n);

better this?
? classno(n)=quadclassunit(n)[1];
? classno(41*73)
6
?


> The trick is to lower user expectation to the point they are satisfied 
> with
> anything that is not obviously wrong.
> 
Your negative view was mine not long ago.

At this years Heidelberg Laureate forum they had three events on AI, and 
mathematicians taked positive about using AI to help for complex 
mathematical problems (youtube video of first session in article):
https://scilogs.spektrum.de/hlf/mathematicians-new-best-friend/


Regards,

Hermann.


P.S:
While I use PARI/GP now for quite some time, I learned only recently 
about qhull.org (started last century as PARI/GP). I used some GP helper 
scripts, some C++ code, planar face coloring code and qconvex and 
geomview to nicely display sum of three squares polyhedron with at most 
6 face colors:
https://github.com/Hermann-SW/RSA_numbers_factored/tree/main/qhull#readme