## The primeness of even n compared to odd n in k*2^n+/-1

Expand Messages
• Recently I announced the discovery of a super-prime power series of the form k*2^n-1, n variable, k fixed, with over 200 primes to the Yahoo Primeform group.
Message 1 of 2 , Aug 19, 2013
• 0 Attachment
Recently I announced the discovery of a super-prime power series of the form k*2^n-1, n variable, k fixed, with over 200 primes to the Yahoo Primeform group.

http://tech.groups.yahoo.com/group/primeform/message/11407

An analysis of the primes discovered to date shows that there are significantly more even n in the list of primes (119) than odd (94). The complete list of the 213 primes to date is at:

I have been fascinated for some while at the ability of k.4^n+/-1 to generate long Cunningham chains, and I am wondering if the predominance of even members of the super-prime power series is connected to the fact that no prime p is p-1 modulo a square base, 4 being 2^2 of course.

I had done some initial work on generalised Cunningham chains back in 2007, and - found good length chains quickly for b=4 and 9, see

Robert
• ... The occurence of more even exponents in the list of primes is a direct result of the distrubution of even and odd n in the input file(s). After sieving to
Message 2 of 2 , Aug 21, 2013
• 0 Attachment
--- In primenumbers@yahoogroups.com, "robert44444uk" <robert_smith44@...> wrote:
>
> Recently I announced the discovery of a super-prime power series of the form k*2^n-1, n variable, k fixed, with over 200 primes to the Yahoo Primeform group.
>
> http://tech.groups.yahoo.com/group/primeform/message/11407
>
> An analysis of the primes discovered to date shows that there are significantly more even n in the list of primes (119) than odd (94). The complete list of the 213 primes to date is at:
>

The occurence of more even exponents in the list of primes is a direct result of the distrubution of even and odd n in the input file(s). After sieving to p=6.34B the input file for the range n=500,001-800,000 contains 30872 even and 26912 odd n candidates.
That's a ratio of 53.4% : 46.6%.
The ratio for the 213 primes is quite similar: 55.9% : 44.1%.

The occurence of a specific n in a sieve file for a given k (and therefore even and/or odd n) is determined by the residue classes (k mod p) and (b^n mod p) and the relations between them (e.g. for eliminating a specific n in k*b^n-1 we need to have (k mod p) == 1/(b^n mod p) ).

For simpler (smaller) k there might be cases with no even or odd exponents at all, for example: 5*2^n-1 (no odd exponents) or 7*2^n-1 (no even ones).

I would therefore expect also some Payam sequences with an opposite distribution, e.g. more odd than even exponents.

-- Thomas
Your message has been successfully submitted and would be delivered to recipients shortly.