OFFSET
0,4
COMMENTS
See the comments in A000186 for further discussion.
Neven Juric alerted me to the fact that Riordan's formula is misleading.
It is not error of Riordan, since, according to the rook theory, he considered U(1) to be -1. [Vladimir Shevelev, Apr 02 2010]
A combinatorial argument, valid for n >= 2, leads to Touchard's formula for the n-th menage number, U(n), a formula which involves the coefficients of Chebyshev polynomials of the first kind. It is combinatorially reasonable to take U(0) = 1 and U(1) = 0, leading to A335700, but taking the connection with Chebyshev polynomials seriously instead gives U(0) = 2 and U(1) = -1, leading to A102761. Riordan derives equation (30) on page 205 for the number of reduced three-line Latin rectangles (A000186) by making use of product identities on Chebyshev polynomials, and therefore requires the second definition; it also requires extending the definition of menage numbers to negative index. Riordan then obtains equation (30a) on page 206 by eliminating the negative indices and redefining U(0) to be 1 (which leads to A000179). A170904 (this sequence) is what is obtained by mistakenly using A335700 instead of A000179 in Riordan's equation (30a). - William P. Orrick, Aug 11 2020
REFERENCES
J. Riordan, An Introduction to Combinatorial Analysis, Wiley, 1958, pp. 206, 209.
FORMULA
MAPLE
# A000166
unprotect(D);
D := proc(n) option remember; if n<=1 then 1-n else (n-1)*(D(n-1)+D(n-2)); fi; end;
[seq(D(n), n=0..30)];
U := proc(n) if n<=1 then 1-n else add ((-1)^k*(2*n)*binomial(2*n-k, k)*(n-k)!/(2*n-k), k=0..n); fi; end;
[seq(U(n), n=0..30)];
Kbad:=proc(n) local k; global D, U; add( binomial(n, k)*D(n-k)*D(k)*U(n-2*k), k=0..floor(n/2) ); end;
[seq(Kbad(n), n=0..30)];
CROSSREFS
KEYWORD
nonn
AUTHOR
N. J. A. Sloane, Jan 21 2010
EXTENSIONS
Edited by N. J. A. Sloane, Apr 04 2010 following a suggestion from Vladimir Shevelev
STATUS
approved