[go: up one dir, main page]

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”).

Search: a369614 -id:a369614
     Sort: relevance | references | number | modified | created      Format: long | short | data
Size of acyclic domain of size n based on the alternating scheme.
+10
3
1, 2, 4, 9, 20, 45, 100, 222, 488, 1069, 2324, 5034, 10840, 23266, 49704, 105884, 224720, 475773, 1004212, 2115186, 4443896, 9319702, 19503224, 40750884, 84990640, 177017810, 368108680, 764571492, 1585851248, 3285861924, 6800042704, 14059397560, 29037419424
OFFSET
1,2
LINKS
A. Galambos and V. Reiner, Acyclic sets of linear orders via the Bruhat order, Social Choice and Welfare, 30 (No. 2, 2008), 245-264.
Bernard Monjardet, Acyclic domains of linear orders: a survey, in "The Mathematics of Preference, Choice and Order: Essays in Honor of Peter Fishburn", edited by Steven Brams, William V. Gehrlein and Fred S. Roberts, Springer, 2009, pp. 139-160. Available as a preprint halshs-00198635.
FORMULA
Monjardet quotes the following formula from Galambos and Reiner: if n mod 2 = 0 then a(n) = 2^(n-3)*(n+3)-binomial(n-2,n/2-1)*(n-3/2), otherwise a(n) = 2^(n-3)*(n+3)-binomial(n-1,(n-1)/2)*(n-1)/2. [Corrected by Jan Volec (janvolec(AT)jikos.cz), Oct 26 2009]
a(n) ~ n*2^(n-3). - Clayton Thomas, Aug 19 2019
PROG
(SageMath)
def a(n):
return (n+3)*2^(n-3) - (binomial(n-2, n/2-1)*(n-3/2) if is_even(n)
else binomial(n-1, (n-1)/2)*(n-1)/2)
print([a(n) for n in (1..20)]) # Andrey Zabolotskiy, Oct 20 2024
CROSSREFS
Cf. A369614.
KEYWORD
nonn,easy
AUTHOR
N. J. A. Sloane Feb 07 2009
EXTENSIONS
More terms added, using the formula, by Andrey Zabolotskiy, Oct 20 2024
STATUS
approved
Maximal size of a connected acyclic domain of permutations of n elements with diameter n*(n-1)/2.
+10
1
1, 2, 4, 9, 20, 45, 100
OFFSET
1,2
COMMENTS
a(n) is at most 2.487^n and at least 2.076^n for large enough n (see Felsner & Valtr). Originally conjectured to equal A144685, but in fact a(n) is asymptotically larger and exceeds A144685 at least for n >= 34 (see Karpov & Slinko). - Clayton Thomas, Aug 19 2019 [Updated by Andrey Zabolotskiy, Dec 31 2023]
REFERENCES
B. Monjardet, Acyclic domains of linear orders: a survey, in "The Mathematics of Preference, Choice and Order: Essays in Honor of Peter Fishburn", edited by Steven Brams, William V. Gehrlein and Fred S. Roberts, Springer, 2009, pp. 139-160.
LINKS
James Abello, The Weak Bruhat Order of S_Sigma, Consistent Sets, and Catalan Numbers, SIAM Journal on Discrete Mathematics, 4 (1991), 1-16; alternative link.
James Abello, The Majority Rule and Combinatorial Geometry (via the Symmetric Group), Annales Du Lamsade, 3 (2004), 1-13.
Vladimir I. Danilov, Alexander V. Karzanov, and Gleb Koshevoy, Condorcet domains of tiling type, Discrete Applied Mathematics 160.7-8 (2012), pages 933-940.
Stefan Felsner and Pavel Valtr, Coding and counting arrangements of pseudolines, Discrete & Computational Geometry 46.3 (2011), pages 405-416.
Alexander Karpov and Arkadii Slinko, Constructing large peak-pit Condorcet domains, Theory and Decision, 94 (2023), 97-120.
B. Monjardet, Acyclic domains of linear orders: a survey, in "The Mathematics of Preference, Choice and Order: Essays in Honor of Peter Fishburn", edited by Steven Brams, William V. Gehrlein and Fred S. Roberts, Springer, 2009, pp. 139-160 ⟨halshs-00198635⟩.
CROSSREFS
Cf. A090245 (has same initial terms but probably is unrelated), A144685, A144687, A369614.
KEYWORD
nonn,hard,more
AUTHOR
N. J. A. Sloane, Feb 07 2009
EXTENSIONS
a(1)-a(2) added and name edited by Andrey Zabolotskiy, Dec 31 2023
STATUS
approved

Search completed in 0.007 seconds