login

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”).

Triangle of 3-Narayana numbers, N(n,k), for n >= 1, 0 <= k <= 2n-2.
2

%I #15 Aug 29 2016 02:49:40

%S 1,1,3,1,1,10,20,10,1,1,22,113,190,113,22,1,1,40,400,1456,2212,1456,

%T 400,40,1,1,65,1095,7095,20760,29484,20760,7095,1095,65,1,1,98,2541,

%U 26180,127435,320034,433092,320034,127435,26180,2541,98,1,1,140,5250,79870

%N Triangle of 3-Narayana numbers, N(n,k), for n >= 1, 0 <= k <= 2n-2.

%C N(n,k) counts ballot sequences for three candidates having length 3n, ending in a tie and having k instances of a vote for a weaker candidate being followed immediately by a vote for a stronger one.

%C Equivalently, N(n,k) counts the lattice paths P := p_1p_2...p_{3n} using the steps X := (1,0,0), Y := (0,1,0) and Z := (0,0,1), running from (0,0,0) to (n,n,n), lying in {(x,y,z) : 0<=x<=y<=z } and satisfying #{i : p_ip_{i+1} in {XY,XZ,YZ} } = k.

%H Paul Barry, <a href="https://cs.uwaterloo.ca/journals/JIS/VOL9/Barry/barry91.html">On Integer-Sequence-Based Constructions of Generalized Pascal Triangles</a>, Journal of Integer Sequences, Vol. 9 (2006), Article 06.2.4.

%H R. A. Sulanke, <a href="http://www.combinatorics.org/Volume_7/Abstracts/v7i1r40.html">Counting Lattice Paths by Narayana Polynomials</a> Electronic J. Combinatorics 7, No. 1, R40, 1-9, 2000.

%H R. A. Sulanke, <a href="http://www.combinatorics.org/ojs/index.php/eljc/article/view/v11i1r54">Generalizing Narayana and Schroeder Numbers to Higher Dimensions</a>, Electron. J. Combin. 11 (2004), Research Paper 54, 20 pp.

%H R. A. Sulanke, <a href="http://dx.doi.org/10.1016/j.tcs.2005.08.014">Three-dimensional Narayana and Schröder numbers</a>, Theoretical Computer Science, Volume 346, Issues 2-3, 28 November 2005, Pages 455-468.

%F For 0<=k<=2n-2, N(n, k) := Sum[2*(-1)^(k-j)*C(3*n+1, k-j)*C(n+j, n)*C(n+j+1, n)*C(n+j+2, n)/(n+1)^2/(n+2), {j, 0, k}] = Sum[(-1)^(k-j)*C(3*n+1, k-j)*a(n, j), {j, 0, k}] where a(m, n) is an entry in the triangle of A056939.

%F Recurrence: If N_n(t) := Sum[t^k*N(n, k), {k, 0, 2n-2}] then (3n-4)(n+2)(n+1)^2 N_n(t) = (3n-2)(n+1)( 4(1+t+t^2) - 5(1+7t+t^2)n +3(1+7t+t^2)n^2 ) N_{n-1}(t) - (n-2)( -12 +29n -30n^2 +9n^3)(1-t)^4 N_{n-2}(t) +(3n-1)(n-2)(n-3)(n-4) (1-t)^6 N_{n-3}(t).

%e 1;

%e 1,3,1;

%e 1,10,20,10,1;

%e 1,22,113,190,113,22,1;

%e 1,40,400,1456,2212,1456,400,40,1;

%e 1,65,1095,7095,20760,29484,20760,7095,1095,65,1;

%e 1,98,2541,26180,127435,320034,433092,320034,127435,26180,2541,98,1

%p seq( seq( add(2*(-1)^(k-j)*binomial(3*n+1, k-j)* binomial(n+j,n)*binomial(n+j+1,n)*binomial(n+j+2,n)/(n+1)^2/(n+2), j = 0 .. k), k = 0 .. 2*n-2), n = 1 ..7 );

%Y Cf. A001263 (Narayana numbers), A005789 (= Sum[N(n, k), {k, 0, 2n-2}], 3-dimensional Catalan numbers), A056939 (antichains in the poset 3*m*n).

%K easy,nonn,tabf

%O 1,3

%A Robert A. Sulanke (sulanke(AT)math.boisestate.edu), Sep 23 2003