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 #56 Aug 22 2021 09:25:31
%S 1,-1,-2,0,-4,2,-6,0,0,4,-10,0,-12,6,8,0,-16,0,-18,0,12,10,-22,0,0,12,
%T 0,0,-28,-8,-30,0,20,16,24,0,-36,18,24,0,-40,-12,-42,0,0,22,-46,0,0,0,
%U 32,0,-52,0,40,0,36,28,-58,0,-60,30,0,0,48,-20,-66,0,44,-24,-70,0,-72,36,0,0,60,-24,-78,0,0,40,-82,0
%N a(n) = mu(n)*phi(n) where mu(n) is the Mobius function (A008683) and phi(n) is the Euler totient function (A000010).
%C Also, a(n) = mu(n)*uphi(n) where mu(n) is the Mobius function (A008683) and uphi(n) is the unitary totient function (A047994), since phi(n) = uphi(n) when n is squarefree, while mu(n) = 0 when n is not squarefree. - _Franklin T. Adams-Watters_, May 14 2006
%C Conjecture: Sum_{n>=1} mu(n)/phi(n) = Sum_{n>=1} a(n)/phi(n)^2 = 0. It is true that Sum_{n>=1} mu(n)/phi(n)^s = 0 at least for s > 1 since: phi(2)=1, phi is multiplicative, so for n's that are squarefree, the phi(n) values can be partitioned in pairs where phi(m)=phi(2m) and mu(m) = -mu(2m). So Sum_{i=1..n} mu(i)/phi(i)^s < Sum_{j=floor(n/2)..n} 1/phi(j)^s, which approaches 0 as n increases since (1) n^(1-e) < phi(n) < n for any e > 0 and n > N(e) and (2) Sum_{i..n} 1/n^s converges for s > 1. Conjecture: Sum_{n>=1} mu(n)/phi(n)^z = 0 for Re(z) > 1.
%C Multiplicative with a(p^1) = 1-p, a(p^e) = 0, e > 1. - _Mitch Harris_, May 24 2005
%C Row sums of triangle A143153 = a signed version of the sequence such that parity = (-) iff A008683(n) = (+); 0 or (+): (1, 1, 2, 0, 4, -2, 6, 0, 0, -4, 10, 0, 12, -6, 0, 0, 0, ...). - _Gary W. Adamson_, Jul 27 2008
%C Dirichlet inverse of A003958. - _R. J. Mathar_, Jul 08 2011
%H Alois P. Heinz, <a href="/A097945/b097945.txt">Table of n, a(n) for n = 1..10000</a>
%H Wikipedia, <a href="https://en.wikipedia.org/wiki/Euler%27s_totient_function">Euler's totient function</a>
%F Dirichlet g.f.: Product_{primes p} (1-p^(1-s)+p^(-s)). - _R. J. Mathar_, Aug 29 2011
%F Sum_{d|n} abs(a(d)) = rad(n) = A007947(n). - _Rémy Sigrist_, Nov 05 2017
%F Sum_{k=1..n} abs(a(k)) ~ c * n^2, where c = A065464/2 = (1/2) * Product_{primes p} (1 - 2/p^2 + 1/p^3) = 0.21412475283854722... Equivalently, c = A065463 * 3 / Pi^2. - _Vaclav Kotesovec_, Jun 14 2020
%F From _Antti Karttunen_, Aug 20 2021: (Start)
%F a(n) = mu(n)*A000010(n) = mu(n)*A003958(n) = mu(n)*A047994(n) = mu(n)*A173557(n), where mu is Möbius mu function (A008683).
%F a(n) = A008966(n) * A023900(n) = abs(mu(n)) * A023900(n).
%F a(n) = A322581(n) - A003958(n).
%F (End)
%p with(numtheory):
%p a:= n-> mobius(n)*phi(n):
%p seq(a(n), n=1..100); # _Alois P. Heinz_, Aug 06 2012
%t Table[ MoebiusMu[n]EulerPhi[n], {n, 85}] (* _Robert G. Wilson v_, Sep 06 2004 *)
%o (PARI) a(n)=moebius(n)*eulerphi(n) \\ _Charles R Greathouse IV_, Feb 21 2013
%o (PARI) for(n=1, 100, print1(direuler(p=2, n, (1 - p*X + X))[n], ", ")) \\ _Vaclav Kotesovec_, Jun 14 2020
%Y Cf. A000010, A003958, A007947, A008683, A008966, A023900, A047994, A143153, A173557, A322581.
%K sign,mult
%O 1,3
%A _Gerald McGarvey_, Sep 04 2004
%E More terms from _Robert G. Wilson v_, Sep 06 2004
%E Edited by _N. J. A. Sloane_, May 20 2006