|
|
A085656
|
|
Number of positive-definite real {0,1} n X n matrices.
|
|
9
|
|
|
|
OFFSET
|
1,2
|
|
COMMENTS
|
A real matrix M is positive-definite if x M x' > 0 for all nonzero real vectors x. Equivalently, all eigenvalues of M + M' are positive.
M need not be symmetric. For the number of different values of M + M' see A085657. - Max Alekseyev, Dec 13 2005
|
|
LINKS
|
|
|
EXAMPLE
|
For n = 2 the three matrices are {{{1, 0}, {0, 1}}, {{1, 0}, {1, 1}}, {{1, 1}, {0, 1}}}.
|
|
MATHEMATICA
|
Table[Count[Tuples[{0, 1}, {n, n}], _?PositiveDefiniteMatrixQ], {n, 4}] (* Eric W. Weisstein, Jan 03 2021 *)
|
|
PROG
|
(PARI) { a(n) = M=matrix(n, n, i, j, 2*(i==j)); r=0; b(1); r } { b(k) = local(t); if(k>n, t=0; for(i=1, n, for(j=1, i-1, if(M[i, j]==1, t++); )); r+=2^t; return; ); forvec(x=vector(k-1, i, [0, 1]), for(i=1, k-1, M[k, i]=M[i, k]=x[i]); if( matdet(vecextract(M, 2^k-1, 2^k-1), 1)>0, b(k+1) ) ) } (Alekseyev)
|
|
CROSSREFS
|
Cf. A055165, which counts nonsingular {0, 1} matrices and A085506, which counts {-1, 0, 1} matrices with positive eigenvalues.
|
|
KEYWORD
|
nonn,nice
|
|
AUTHOR
|
|
|
EXTENSIONS
|
|
|
STATUS
|
approved
|
|
|
|