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 #30 Oct 15 2020 16:56:31
%S 0,3,12,10,12,39,36,46,30,63,48,106,24,93,216,148,24,141,60,196,162,
%T 141,120,304,60,111,162,232,42,459,108,394,174,141,372,442,54,183,420,
%U 538,42,489,78,352,540,243,198,904,102,303,294,334,78,513
%N Number of solutions to 4/n = 1/x + 1/y + 1/z in positive integers.
%C Corrected version of A192786.
%C The Erdos-Straus conjecture is that a(n) > 0 for n > 1. Swett verified the conjecture for n < 10^14.
%C Vaughan shows that the number of n < x with a(n) = 0 is at most x exp(-c * (log x)^(2/3)) for some c > 0.
%C After a(2) = 3, the values shown are all composite. [_Jonathan Vos Post_, Jul 17 2011]
%H Hugo Pfoertner, <a href="/A292581/b292581.txt">Table of n, a(n) for n = 1..1000</a>
%H Christian Elsholtz and Terence Tao, <a href="https://arxiv.org/abs/1107.1010">Counting the number of solutions to the Erdos-Straus equation on unit fractions</a>, arXiv:1107.1010 [math.NT], 2011-2015.
%H Allan Swett, <a href="https://web.archive.org/web/20110716110731/http://math.uindy.edu/swett/esc.htm">The Erdos-Strauss Conjecture</a>, 1999.
%H R. C. Vaughan, <a href="http://dx.doi.org/10.1112/S0025579300002886">On a problem of Erdős, Straus and Schinzel</a>, Mathematika 17 (1970), pp. 193-198.
%e a(3)=12 because 4/3 can be expressed in 12 ways:
%e 4/3 = 1/1 + 1/4 + 1/12
%e 4/3 = 1/1 + 1/6 + 1/6
%e 4/3 = 1/1 + 1/12 + 1/4
%e 4/3 = 1/2 + 1/2 + 1/3
%e 4/3 = 1/2 + 1/3 + 1/2
%e 4/3 = 1/3 + 1/2 + 1/2
%e 4/3 = 1/4 + 1/1 + 1/12
%e 4/3 = 1/4 + 1/12 + 1/1
%e 4/3 = 1/6 + 1/1 + 1/6
%e 4/3 = 1/6 + 1/6 + 1/1
%e 4/3 = 1/12 + 1/1 + 1/4
%e 4/3 = 1/12 + 1/4 + 1/1
%e a(4) = 10 because 4/4 = 1 can be expressed in 10 ways:
%e 4/4= 1/2 + 1/3 + 1/6
%e 4/4= 1/2 + 1/4 + 1/4
%e 4/4= 1/2 + 1/6 + 1/3
%e 4/4= 1/3 + 1/2 + 1/6
%e 4/4= 1/3 + 1/3 + 1/3
%e 4/4= 1/3 + 1/6 + 1/2
%e 4/4= 1/4 + 1/2 + 1/4
%e 4/4= 1/4 + 1/4 + 1/2
%e 4/4= 1/6 + 1/2 + 1/3
%e 4/4= 1/6 + 1/3 + 1/2
%t checkmult[a_, b_, c_] := If[Denominator[c] == 1, If[a == b && a == c && b == c, Return[1], If[a != b && a != c && b != c, Return[6], Return[3]]], Return[0]];
%t a292581[n_] := Module[{t, t1, s, a, b, c, q = Quotient}, t = 4/n; s = 0; For[a = q[1, t]+1, a <= q[3, t], a++, t1 = t - 1/a; For[b = Max[q[1, t1] + 1, a], b <= q[2, t1], b++, c = 1/(t1 - 1/b); s += checkmult[a, b, c]]]; Return[s]];
%t Reap[For[n=1, n <= 54, n++, Print[n, " ", an = a292581[n]]; Sow[an]]][[2, 1]] (* _Jean-François Alcover_, Dec 02 2018, adapted from PARI *)
%o (PARI) \\ modified version of code by _Charles R Greathouse IV_ in A192786
%o checkmult (a,b,c) =
%o {
%o if(denominator(c)==1,
%o if(a==b && a==c && b==c,
%o return(1),
%o if(a!=b && a!=c && b!=c,
%o return(6),
%o return(3)
%o )
%o ),
%o return(0)
%o )
%o }
%o a292581(n) =
%o {
%o local(t, t1, s, a, b, c);
%o t = 4/n;
%o s = 0;
%o for (a=1\t+1, 3\t,
%o t1=t-1/a;
%o for (b=max(1\t1+1,a), 2\t1,
%o c=1/(t1-1/b);
%o s+=checkmult(a,b,c);
%o )
%o );
%o return(s);
%o }
%o for (n=1,54,print1(a292581(n),", "))
%Y For more references and links see A192787.
%Y Cf. A073101, A192786, A292624, A337432.
%K nonn
%O 1,2
%A _Hugo Pfoertner_, Sep 20 2017