Year-end appeal: Please make a donation to the OEIS Foundation to support ongoing development and maintenance of the OEIS. We are now in our 61st year, we have over 378,000 sequences, and we’ve reached 11,000 citations (which often say “discovered thanks to the OEIS”).
%I #18 May 04 2019 03:50:26
%S 2,3,3,5,5,7,5,7,7,11,11,13,11,13,13,17,17,19,17,19,13,23,19,19,23,23,
%T 19,29,29,31,23,29,31,29,31,37,29,37,37,41,41,43,41,43,31,47,43,37,47,
%U 43,43,53,47,43,53,53,43,59,59,61,53,59,61,59,61,67,53,67,67,71,71,73
%N Greatest p less than or equal to n with p and q both prime, p+q = 2n.
%C Essentially the same as A002374, which is the main entry for this sequence. - _Franklin T. Adams-Watters_, Jan 25 2010
%C Well defined only under the assumption that the yet unproved Goldbach conjecture holds, which states that any even N = 2n > 2 has a decomposition as sum of two primes. - _M. F. Hasler_, May 03 2019
%F a(n) = n - A047160(n). - _Jason Kimberley_, Aug 31 2011
%F a(n) = n if and only if n is prime, i.e., n in A000040. - _M. F. Hasler_, May 03 2019
%e From _M. F. Hasler_, May 03 2019: (Start)
%e For n = 2, the largest prime p <= n is p = 2, and q := 2n - p = 4 - 2 = 2 is also prime, whence a(2) = 2. We see that whenever n is prime, we will have a(n) = p = q = n.
%e For n = 4, the largest prime p <= n is p = 3, and q := 2n - p = 8 - 3 = 5 is also prime, whence a(4) = p = 3.
%e For n = 8, the largest prime less than n is p' = 7, but 2n - p' = 16 - 7 = 9 is not prime, so we have to go to the next smaller prime p = 5 and now q := 2n - p = 16 - 5 = 11 is also prime, whence a(8) = p = 5. (End)
%t f[n_] := Block[{p = n/2}, While[ !PrimeQ[p] || !PrimeQ[n - p], p-- ]; p]; Table[ f[n], {n, 4, 146, 2}]
%o (PARI) a(n) = {my(p = precprime(n)); while (!isprime(2*n-p), p = precprime(p-1)); p;} \\ _Michel Marcus_, Oct 22 2016
%o (PARI) A112823(n)=forprime(q=n,2*n,isprime(2*n-q)&&return(2*n-q)) \\ _M. F. Hasler_, May 03 2019
%Y Cf. A002374, A020481, A047160.
%K nonn
%O 2,1
%A _Robert G. Wilson v_, Sep 05 2005