Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
A109385
Maximum number of prime implicants of a symmetric function of n Boolean variables.
2
1, 2, 6, 13, 32, 92, 218, 576, 1698, 4300, 11770, 34914, 91105, 254438, 759488, 2030618, 5746274, 17189858, 46698068, 133334440, 399479982, 1099206284, 3159208516, 9470895658, 26313455375, 76003857800, 227935595004, 638304618462, 1850933165704, 5551816202580
OFFSET
1,2
COMMENTS
Many people have conjectured that this sequence is equal to A003039. Certainly it is a lower bound. An upper bound is given in A109388.
REFERENCES
Yoshihide Igarashi, An improved lower bound on the maximum number of prime implicants, Transactions of the IECE of Japan, E62 (1979), 389-394.
A. P. Vikulin, Otsenka chisla kon"iunktsii v sokrashchennyh DNF [An estimate of the number of conjuncts in reduced disjunctive normal forms], Problemy Kibernetiki 29 (1974), 151-166.
EXAMPLE
a(10) = 4300 because the symmetric function S_{1,2,4,5,6,7,9,10}(x_1,...,x_{10}) has 90+4200+10 prime implicants.
MATHEMATICA
b[m_, n_] := If[m < 0, 0, Multinomial[Floor[m/2], Ceiling[m/2], n - m] + b[Ceiling[m/2] - 2, n]]; a[n_] := Multinomial[Floor[n/3], Floor[(n + 1)/3], Floor[(n + 2)/3]] + b[Floor[(n - 4)/3], n] + b[Floor[(n - 5)/3], n]; Table[a[n], {n, 35}]
CROSSREFS
KEYWORD
easy,nonn
AUTHOR
Don Knuth, Aug 25 2005
EXTENSIONS
Extended by T. D. Noe using the Mma program, Jan 15 2012
STATUS
approved