mersenneforum.org > Math The constant in Hardy-Littlewood's Conjecture F
 User Name Remember Me? Password
 Register FAQ Search Today's Posts Mark Forums Read

 2010-07-06, 15:44 #1 CRGreathouse     Aug 2006 3·1,987 Posts The constant in Hardy-Littlewood's Conjecture F The Hardy-Littlewood Conjecture F includes calculation of the infinite product $\prod_{\varpi\ge3,\varpi\not|a}\left(1-\frac{1}{\varpi-1}\left(\frac D\varpi\right)\right)$ where (I believe!) $\varpi$ ranges over the primes and $\left(\frac D\varpi\right)$ is the Jacobi symbol. Is there a good way to calculate this? Or, how can one calculate a reasonable number of decimal places of $f(D)=\prod\left(1-\frac{1}{\varpi-1}\left(\frac D\varpi\right)\right)$ where $\varpi$ ranges over the odd primes? (Recovering the original problem for any factorable a is easy.)
 2010-07-06, 16:47 #2 CRGreathouse     Aug 2006 3×1,987 Posts The problem can be solved if there are good ways to accelerate the calculation of sums over primes in congruence classes: $\prod_{\varpi\equiv m\pmod n}1-\frac{1}{\varpi-1}$ and $\prod_{\varpi\equiv m\pmod n}1+\frac{1}{\varpi-1}$ for $\varpi$ prime and integer m,n. Can this be done? Last fiddled with by CRGreathouse on 2010-07-06 at 16:56
2010-07-06, 17:33   #3
ccorn

Apr 2010

9416 Posts

Quote:
 Originally Posted by CRGreathouse The Hardy-Littlewood Conjecture F includes calculation of the infinite product $\prod_{\varpi\ge3,\varpi\not|a}\left(1-\frac{1}{\varpi-1}\left(\frac D\varpi\right)\right)$ where (I believe!) $\varpi$ ranges over the primes and $\left(\frac D\varpi\right)$ is the Jacobi symbol. Is there a good way to calculate this? Or, how can one calculate a reasonable number of decimal places of $f(D)=\prod\left(1-\frac{1}{\varpi-1}\left(\frac D\varpi\right)\right)$ where $\varpi$ ranges over the odd primes? (Recovering the original problem for any factorable a is easy.)
Have you checked Shanks' paper?

2010-07-06, 17:34   #4
ccorn

Apr 2010

100101002 Posts

Quote:
 Originally Posted by CRGreathouse The problem can be solved if there are good ways to accelerate the calculation of sums over primes in congruence classes: $\prod_{\varpi\equiv m\pmod n}1-\frac{1}{\varpi-1}$ and $\prod_{\varpi\equiv m\pmod n}1+\frac{1}{\varpi-1}$ for $\varpi$ prime and integer m,n. Can this be done?
(Logs of) These would diverge, I'm afraid.

2010-07-06, 17:47   #5
CRGreathouse

Aug 2006

10111010010012 Posts

Thanks! I'll look it over.

Quote:
 Originally Posted by ccorn (Logs of) These would diverge, I'm afraid.
(facepalm)

Of course if I had a Mertens' Theorem-like estimate for
$\prod_{p\equiv m\pmod n,p\le x}1-\frac{1}{p-1}$
and
$\prod_{p\equiv m\pmod n,p\le x}1+\frac{1}{p-1}$
that would suffice.

Last fiddled with by CRGreathouse on 2010-07-06 at 17:59

 Similar Threads Thread Thread Starter Forum Replies Last Post badbud65 Software 46 2016-05-02 23:18 Zeta-Flux Math 4 2007-11-30 08:56 kar_bon Riesel Prime Search 45 2007-11-27 19:15 mfgoode Math 10 2004-06-02 04:06 GP2 Math 2 2003-09-14 04:36

All times are UTC. The time now is 13:42.

Fri Jan 22 13:42:27 UTC 2021 up 50 days, 9:53, 0 users, load averages: 3.20, 2.83, 2.71