login
Decimal expansion of the real solution to 23*x^5 - 41*x^4 + 10*x^3 - 6*x^2 - x - 1 = 0. Constant occurring in the asymptotic behavior of the number of lattice paths of slope 2/5, observed by D. Knuth.
2

%I #14 May 27 2019 12:35:28

%S 1,6,3,0,2,5,7,6,6,2,9,9,0,3,5,0,1,4,0,4,2,4,8,0,1,8,4,9,3,1,5,9,8,6,

%T 3,0,0,5,1,4,5,8,4,4,2,6,6,9,0,1,4,9,4,0,5,8,4,9,8,5,0,2,6,5,9,5,2,5,

%U 6,8,9,1,2,9,8,6,8,5,0,4,7,9,8,3,4,1,3,2,4,1

%N Decimal expansion of the real solution to 23*x^5 - 41*x^4 + 10*x^3 - 6*x^2 - x - 1 = 0. Constant occurring in the asymptotic behavior of the number of lattice paths of slope 2/5, observed by D. Knuth.

%C In his 2014 lecture in Paris "Problems That Philippe (Flajolet) Would Have Loved" D. Knuth discussed as Problem 4 "Lattice Paths of Slope 2/5" and reported as an empirical observation that A[5*t-1,2*t-1]/B[5*t-1,2*t-1] = a - b/t + O(t^-2), with constants a~=1.63026 and b~=0.159. For the meaning of A and B see A322631. The exact values of a (this sequence) and b (A322633) were found in 2016 by Banderier and Wallner.

%H Cyril Banderier, Michael Wallner, <a href="https://arxiv.org/abs/1605.02967">Lattice paths of slope 2/5</a>, arXiv:1605.02967 [cs.DM], 10 May 2016.

%H D. E. Knuth, <a href="http://www-cs-faculty.stanford.edu/~uno/flaj2014.pdf">Problems That Philippe Would Have Loved</a>, Paris 2014.

%e 1.6302576629903501404248018493159863005145844266901494058498502659525689...

%p evalf[100](solve(23*x^5-41*x^4+10*x^3-6*x^2-x-1=0,x)[1]); # _Muniru A Asiru_, Dec 21 2018

%t RealDigits[Root[23#^5 - 41#^4 + 10#^3 - 6#^2 - # - 1&, 1], 10, 100][[1]] (* _Jean-François Alcover_, Dec 30 2018 *)

%o (PARI) solve(x=1,2,23*x^5-41*x^4+10*x^3-6*x^2-x-1)

%Y Cf. A322631, A322633.

%K nonn,cons

%O 1,2

%A _Hugo Pfoertner_, Dec 21 2018