cp's OEIS Frontend

This is a front-end for the Online Encyclopedia of Integer Sequences, made by Christian Perfect. The idea is to provide OEIS entries in non-ancient HTML, and then to think about how they're presented visually. The source code is on GitHub.

A229136 Number of solutions to Sum_{i=1..n} x_i^2 == 1 (mod 4) with x_i in 0..3.

Original entry on oeis.org

2, 8, 24, 64, 192, 768, 3584, 16384, 69632, 278528, 1081344, 4194304, 16515072, 66060288, 266338304, 1073741824, 4311744512, 17246978048, 68853694464, 274877906944, 1098437885952, 4393751543808, 17583596109824, 70368744177664, 281543696187392, 1126174784749568
Offset: 1

Views

Author

Keywords

Comments

Conjecture: a(n) = A131885(n)*2^(n-1) for n >= 1. [Corrected by Petros Hadjicostas, Dec 20 2019]
From Petros Hadjicostas, Dec 20 2019: (Start)
Since this sequence is column k = 1 of A330619, we have a(n) = 4*a(n-1) - 8*a(n-2) + 2^(2*n-3) for n >= 3. (This follows from the theory developed in A330619.) If we let b(n) = a(n)/2^(n-1) for n >= 1, we get b(n) = 2*b(n-1) - 2*b(n-2) + 2^(n-2) for n >= 3.
It follows that 2*b(n-1) = 4*b(n-2) - 4*b(n-3) + 2^(n-2) for n >= 4. Subtracting the last equation from the previous one, we get (after some algebra) b(n) = 4*b(n-1) - 6*b(n-2) + 4*b(n-3) for n >= 4. We can easily verify that b(1) = 2, b(2) = 4, and b(3) = 6, and this proves that b(n) = A131885(n) for n >= 1. This proves the above conjecture. (End)

Crossrefs

Programs

  • Mathematica
    a[1] = 2; a[2] = 8; a[3] = 24; a[n_] := a[n-1]*8 + a[n-2]*(-24) + 32*a[n - 3]; Table[a[n], {n, 15}]
  • PARI
    Vec(-2*x*(2*x-1)^2/((4*x-1)*(8*x^2-4*x+1)) + O(x^100)) \\ Colin Barker, Nov 10 2014

Formula

G.f.: 1/(1 - 4*x) + Q(0)/(2 - 4*x), where Q(k) = 1 + 1/(1 - 2*x*(k+1)/(2*x*(k+2) + 1/Q(k+1) )); (continued fraction). - Sergei N. Gladkovskii, Sep 27 2013
G.f.: -2*x*(2*x - 1)^2 / ((4*x - 1)*(8*x^2 - 4*x + 1)). - Colin Barker, Nov 10 2014
a(n) = 4*a(n-1) - 8*a(n-2) + 2^(2*n-3) for n >= 3. - Petros Hadjicostas, Dec 20 2019