Re: Implementation of functions for "Integer partitions detect the primes" paper
Bill Allombert <[email protected]>
| Newsgroups | gmane.comp.mathematics.pari.user |
|---|---|
| Message-ID | <aG7NkUapcv7xsi5p@seventeen> |
On Wed, Jul 09, 2025 at 10:00:15PM +0200, [email protected] wrote: > On 2025-07-08 22:09, Bill Allombert wrote: > > You should do > > M1(n)=my(s=0);fordiv(n,d,s+=d);s; > > > > But this function is just sigma(n) > > > Ok. > > > > ? M2(n)=s=0;for(m=1,n,forpart(v=m,if(v[1]<v[2],for(d=1,n\v[1],r=n-d*v[1];if(r%v[2]==0,s+=d*(r\v[2])))),[1,m],[2,2]));s; > > > > As I understand this function is > > (sigma(n,3)-(2*n-1)*sigma(n))/8 > > which is much faster to compute. > > > That does not work: > > ? M1(n)=sigma(n); > ? M2(n)=(sigma(n,3)-(2*n-1)*sigma(n))/8; > ? T1(n)=(n^2 - 3^n + 2)*M1(n) - 8*M2(n); ^^^ This should be 3*n. Cheers, Bill