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 #70 Jun 25 2023 09:39:26
%S 1,4,3,15,20,6,56,105,60,10,210,504,420,140,15,792,2310,2520,1260,280,
%T 21,3003,10296,13860,9240,3150,504,28,11440,45045,72072,60060,27720,
%U 6930,840,36
%N Triangle read by rows, related to clusters of type D.
%C Let C_{n+1} be the cyclic quiver with n+1 vertices. Empirically, the n-th row is related to the green-mutation partial order on clusters for this quiver, restricted to clusters that do not meet the initial seed.
%C Apparently, value of the associated polynomials at -2 is A089849, up to sign.
%C By evaluating the associated polynomials at x-1, one apparently gets A062196.
%C The rows seem to give (up to sign) the coefficients in the expansion of the integer-valued polynomial (x+1)^2*(x+2)^2*(x+3)^2*...*(x+n)^2*(x+n+1)*(x+n+2) / (n! * (n+2)!) in the basis made of the binomial(x+i,i). - _F. Chapoton_, Oct 31 2022
%C Chapoton's observation above is correct: the precise expansion is (x+1)^2*(x+2)^2*(x+3)^2*...*(x+n)^2*(x+n+1)*(x+n+2) / (n! * (n+2)!) = Sum_{k = 0..n} (-1)^k*T(n+1,k)*binomial(x+2*n+2-k, 2*n+2-k), as can be verified using the WZ algorithm. For example, n = 2 gives (x+1)^2*(x+2)^2*(x+3)*(x+4)/(2!*4!) = 15*binomial(x+6,6) - 20*binomial(x+5,5) + 6*binomial(x+4,4). - _Peter Bala_, Jun 24 2023
%F T(n, k) = (n-k)*binomial(n,k)*binomial(2*n-k, n-1)/n, for n >= 1 and 0 <= k < n.
%F From _Peter Bala_, Jun 24 2023: (Start)
%F As conjectured above by Chapoton we have
%F Sum_{k = 0..n-1} T(n,k)*(x - 1)^k = Sum_{k = 0..n-1} A062196(n-1,k)*x^k and
%F Sum_{k = 0..n-1} T(n,k)*(-2)^k = (-1)^floor(n/2)*A089849(n) for n >= 1 (both easily verified using the WZ algorithm). (End)
%e Triangle begins:
%e [1] 1
%e [2] 4, 3
%e [3] 15, 20, 6
%e [4] 56, 105, 60, 10
%e [5] 210, 504, 420, 140, 15
%e [6] 792, 2310, 2520, 1260, 280, 21
%e [7] 3003, 10296, 13860, 9240, 3150, 504, 28
%e ...
%o (Sage)
%o def T_row(n):
%o return [(n-k)*binomial(n,k)*binomial(2*n-k,n-1)//n for k in range(n)]
%o for n in range(1, 8): print(T_row(n))
%o (PARI) row(n) = vector(n, k, k--; (n-k)*binomial(n,k)*binomial(2*n-k, n-1)/n); \\ _Michel Marcus_, Sep 30 2021
%Y Cf. A001791 (T(n,1)), A000217 (T(n,n)), A026002 (row sums), A000012 (alternating row sum), A051924 (number of clusters of type D_n).
%Y Cf. A089849, A062196, A063007, A253283.
%K tabl,nonn
%O 1,2
%A _F. Chapoton_, Sep 30 2021