%I #26 Jan 15 2023 09:28:26
%S 2,2,2,2,2,6,2,2,2,2,2,4,4,2,2,2,4,4,6,6,2,10,2,2,2,2,2,2,10,2,4,4,6,
%T 6,2,2,2,2,6,6,6,2,4,4,2,2,18,6,6,2,2,2,2,6,6,6,2,6,6,14,4,4,2,2,4,4,
%U 2,2,18,6,6,2,6,6,4,4,2,2,14,4,4,14,10,2,2,4,4,2,2,8,8,2
%N Table for period length of periods of primitive reduced binary quadratic forms with discriminants D(n) = A079896(n).
%C The indefinite binary quadratic forms [a,b,c] have discriminant D := b^2 - 4*a*c > 0, not a square, given in A079896.
%C Primitive forms satisfy gcd(a,b,c) = 1. For the definition of reduced binary quadratic forms see a comment under A087048.
%C The number of periods of equivalent primitive reduced forms is given in A087048 (the class number).
%C Here the lengths of these periods is recorded. The computation is based on the book by Scholz and Schoeneberg. The row sums give A082174(n), the number of primitive reduced forms for D(n).
%C Two forms [a,b,c] and [a',b',c'] are properly equivalent if the 2 x 2 coefficient matrices A := [[a,b/2],[b/2,a]] and A' := [[a',b'/2],[b'/2,a']] satisfy A' = S^{-1,T} A S^{-1} with some matrix S, det S = +1 (T stands for transposed). The indeterminates (x,y) and (x',y') which represent the same number k = (x,y) A (x,y)^T = (x',y') A' (x',y')^T are related then by (x',y')^T = S (x,y)^T.
%C For the periods of primitive reduced forms for D(n), n = 0, ..., 100, see the link. See also the Buell reference, with the examples on p. 30, giving the periods for n = 0, ..., 19. They coincide with the ones given in the link up to the cyclic order in the periods.
%C All period lengths are even. See Buell, Proposition 3.6 on p. 24.
%D D. A. Buell, Binary quadratic forms,1989, Springer, especially Ch. 3.
%D A. Scholz and B. Schoeneberg, Einführung in die Zahlentheorie, 5. Aufl., de Gruyter, Berlin, New York, 1973, ch. 31, pp. 112 ff.
%H Wolfdieter Lang, <a href="/A225953/a225953_1.pdf">Table for n = 0..100 and more.</a>
%H Wolfdieter Lang, <a href="/A225953/a225953_3.pdf">Periods of Indefinite Binary Quadratic Forms, Continued Fractions and the Pell +/-4 Equations.</a>
%F a(n,k), n >= 0, k = 1, 2, ..., A087048(n), is the length of the period of the k-th primitive reduced forms for discriminant D(n) = A079896. The order in row n is nonincreasing.
%e The irregular table a(n,k) begins:
%e n/k 1 2 ... D(n) A087048(n) A082174(n)
%e 0: 2 5 1 2
%e 1: 2 8 1 2
%e 2: 2 2 12 2 4
%e 3: 2 13 1 2
%e 4: 6 17 1 6
%e 5: 2 20 1 2
%e 6: 2 2 21 2 4
%e 7: 2 2 24 2 4
%e 8: 4 4 28 2 8
%e 9: 2 29 1 2
%e 10: 2 2 32 2 4
%e 11: 4 4 33 2 8
%e 12: 6 37 1 6
%e 13: 6 2 40 2 8
%e 14: 10 41 1 10
%e 15: 2 2 44 2 4
%e 16: 2 2 45 2 4
%e 17: 2 2 48 2 4
%e 18: 10 52 1 10
%e 19: 2 53 1 2
%e 20: 4 4 56 2 8
%e ... for rows up to n = 100 see the link.
%e a(0) = 2 because there are 2 = A082174(n) primitive reduced forms with discriminant D = 5, namely [-1, 1, 1] and [1, 1, -1], which are equivalent, and they give 1 = A087048(0) period [[[-1, 1, 1],[1, 1, -1]]] of length 2 = a(0,1).
%e Row n=2 is 2 2, because there are 4 primitive reduced forms for D = 12, appearing in 2 periods, each of length 2: [[[-1, 2, 2], [2, 2, -1]], [[1, 2, -2], [-2, 2, 1]]].
%Y Cf. A079896, A082174, A087048.
%K nonn,tabf
%O 0,1
%A _Wolfdieter Lang_, May 27 2013