- Open Access
Circuit lengths of graphs for the Picard group
© Beşenk et al.; licensee Springer. 2013
- Received: 23 November 2012
- Accepted: 24 February 2013
- Published: 18 March 2013
In this paper, we examine some properties of suborbital graphs for the Picard group. We obtain edge and circuit conditions, then propose a conjecture for the graph to be forest. This paper is an extension of some results in (Jones et al. in The Modular Group and Generalized Farey Graphs, pp. 316-338, 1991).
- Maximal Subgroup
- Permutation Group
- Modular Group
- Residue Class
- Linear Fractional Transformation
is a subgroup of .
that is, .
In this study, we consider the action of the Picard group on the set in the spirit of the theory of permutation groups and a graph arising from this action in hyperbolic geometric terms.
Note that as and , it follows that and so is a reduced fraction.
Theorem 2.1 The action of P on is transitive.
Proof It is enough to prove that the orbit containing ∞ is . If (in reduced form), then as , there exist with . Then the element of P sends ∞ to . □
We now consider the imprimitivity of the action of P on , beginning with a general discussion of the primitivity of permutation groups. Let be a transitive permutation group, consisting of a group G acting on a set Δ transitively. An equivalence relation ≈ on Δ is called G-invariant if, whenever satisfy , then for all . The equivalence classes are called blocks, and the block containing α is denoted by .
the identity relation if and only if ;
the universal relation for all .
Otherwise is called primitive. These two relations are supposed to be trivial relations. Clearly, a primitive group must be transitive, for if not, the orbits would form a system of blocks. The converse is false, but we have the following useful result in .
Lemma 2.2 Let be a transitive permutation group. is primitive if and only if , the stabilizer of , is a maximal subgroup of G for each .
The number of blocks (equivalence classes) is the index and the block containing α is just the orbit .
We can apply these ideas to the case where G is the P and Δ is .
Lemma 2.3 The stabilizer of ∞ in is the set of denoted by .
Definition 2.4 is a subgroup of P.
It is clear that . We will define an equivalence relation ≈ induced on by P. We must point out that this equivalence relation is different from the one in . Here let us take the group instead of . The purpose of our work is related to this choice. We now collect some information on permutation groups (see ). Given a permutation group G on a finite set Δ, some natural questions arise as follows:
Orbit problem: What are the orbits of G?
Block problem: Is G primitive? If not, find a nontrivial block for G.
Actually, it is more important to find the minimal nontrivial blocks for G because many computations dealing with permutation groups work better with it. In this meaning, the choice of decomposition- is substantial.
Hence, our aim is to see how graphs are affected by decomposition .
in P for which , . Now iff , so iff and . Here, the number of blocks is .
Theorem 2.5 The index , where and N is not a unit.
Suppose that . There are residue classes , where denotes Euler’s function. To each of these classes for a, the numbers b and c may be chosen arbitrarily . is uniquely determined. In this case there are altogether solutions.
Suppose that . There are residue classes . Corresponding to each of these, may be chosen arbitrarily since in the case , there are possibilities for and is again uniquely determined. Hence there are additional solutions.
Together we obtain . Consequently, we have since , as required. □
In , Sims introduced the idea of suborbital graphs of a permutation group G acting on a set Δ. These are graphs with a vertex-set Δ, on which G induces automorphisms. We summarize Sims’ theory as follows.
Let be transitive permutation group. Then G acts on by (, ). The orbits of this action are called suborbitals of G. The orbit containing is denoted by . From we can form a suborbital graph : its vertices are the elements of Δ, and there is a directed edge from γ to δ if . A directed edge from γ to δ is denoted by . If , then we will say that there exists an edge in . In this paper our calculation concerns P, so we can draw this edge as a hyperbolic geodesic in the upper half-plane , see .
The orbit is also a suborbital graph and it is either equal to or disjoint from . In the latter case, is just with the arrows reserved and we call, in this case, and paired suborbital graphs. In the former case, and the graph consists of pairs of oppositely directed edges; it is convenient to replace each such pair by a single undirected edge so that we have a undirected graph which we call self-paired.
In this study, G and Δ will be P and , respectively. Since P acts transitively on , each suborbital contains a pair for some ; writing , we denote this suborbital by and the corresponding suborbital graph by .
Definition 3.1 By a directed circuit in , we mean a sequence of different vertices such that , where ; an anti-directed circuit will denote a configuration like the above with at least an arrow (not all) reversed.
If , then the circuit, directed or not, is called a triangle.
If , then we will call the configuration a self-paired edge: it consists of a loop based at each vertex.
We call a graph a forest if it does not contain any circuits.
We now investigate the suborbital graphs for the action P on .
From the determinant, we have . As , , then , . Thus, we obtain that , and . Taking with , we have that , and .
The result is the same. □
Theorem 3.3 is self-paired iff .
Proof We suppose that is self-paired. If , then it must also be . So, there exists such that . From the edge , we have that means that . From the edge , we obtain that . Hence, and then .
Conversely, suppose that . Taking with , we have that . If , then there exists such that . Hence . Let , then , and . The case of is similar. □
If in , then Theorem 3.2 implies that there exists such that , so . Thus each connected component of lies in a single block for ≈, of which there are , so we have the following corollary.
Corollary 3.4 has at least connected components; in particular, is not connected if N is not a unit.
As we saw, each is a disjoint union of subgraphs, the vertices of each subgraph forming a single block with respect to the relation ≈. Since P acts transitively on , it permutes these blocks transitively, so the subgraphs are all isomorphic. We let be the subgraph of whose vertices form the block so that consists of disjoint copies of .
Theorem 3.5 Let . There is an edge in iff there exists a unit such that and , where either or and .
if , then , giving ;
if , then , giving .
Hence in iff either or . Opposite direction can be seen easily. □
Theorem 3.6 permutes the vertices and the edges of transitively.
Proof Suppose that . As P acts on transitively, for some . Since and ≈ is P-invariant equivalence relation, ; that is, . Thus, as , .
Hence as . Furthermore, and ; that is, . □
Theorem 3.7 contains directed triangles iff there exists a unit such that .
Proof Suppose that contains a directed triangle. Because of the transitive action, the form of the directed triangle can be taken as . Since the edge conditions in Theorem 3.5 have to be provided for the edge , there are two cases. In the first, there exists a unit such that and , giving . In the second, there exists a unit such that and , giving . Consequently, we obtain that .
Conversely, let be a unit such that . Theorem 3.2 implies that there is a directed triangle in . □
Theorem 3.8 If , then contains no anti-directed triangles.
Proof Suppose that contains an anti-directed triangle. Because of the transitive action, we may assume that an anti-directed triangle has the form . From the second edge, there exists a unit such that and . Since , then , giving . Hence we have that means that or . But is impossible. Therefore, we obtain that . If , then , which contradicts to . □
3.3 Some results
If , then is satisfied for .
If , then is satisfied for .
Thus, the subgraphs and have directed triangles.
Example 3.10 For , there is no unit which satisfies the congruence . So, the subgraphs and have no directed triangles.
or its square. So, if our calculation is true, ϕ must be conjugate to this transformation. It is well known that the transformations and are conjugates iff there exists a transformation such that . Let us give an example.
Example 3.12 Let . If , then is satisfied for . Hence, is a directed triangle in . The corresponding elliptic element is . Furthermore, it can be easily seen that and are conjugates.
Observation 3.13 It is well known that, because of the abstract group structure of P as a free product amalgamated with a modular group, each finite ordered elliptic element will be either of order 2 or 3. On the other hand, in [2, 11], authors give some results about a connection between the periods of elliptic elements of a chosen permutation group with the circuits in suborbital graphs of it. At this point, it is reasonable to conjecture the following.
Conjecture 3.14 is a forest if and only if it contains no triangles, that is, if and only if is not congruent to zero for modulo N.
Remark 3.15 A similar conjecture is given by Jones, Singerman and Wicks for the modular group in  and then proved by Akbaş . In our case, the assertion which says that no edges cross to each other seems to be a problem.
Dedicated to Professor Hari M Srivastava.
The authors would like to thank Baris Kendirli for reading the first version of the manuscript and his enthusiastic support during XXV National Mathematics Symposium in Nigde. This work is partly supported by the research fund of Karadeniz Technical University (Project No: ARGEBD-9752).
- Bigg NL, White AT London Mathematical Society Lecture Note Series 33. In Permutation Groups and Combinatorial Structures. Cambridge University Press, Cambridge; 1979.View ArticleGoogle Scholar
- Keskin, R: Suborbital graphs for the groups of modular and Picard modular. PhD thesis, Karadeniz Technical University, Trabzon, 1996Google Scholar
- Dixon JD, Mortimer B Graduate Texts in Mathematics 163. In Permutation Groups. Springer, New York; 1996.View ArticleGoogle Scholar
- Schoeneberg B: Elliptic Modular Functions. Springer, Berlin; 1974.View ArticleGoogle Scholar
- Hardy GH, Wright EM: An Introduction to the Theory of Numbers. 5th edition. Oxford University Press, Oxford; 1979.Google Scholar
- Sims CC: Graphs and finite permutation groups. Math. Z. 1967, 95: 76–86. 10.1007/BF01117534MathSciNetView ArticleGoogle Scholar
- Jones GA, Singerman D: Complex Functions: an Algebraic and Geometric Viewpoint. Cambridge University Press, Cambridge; 1987.View ArticleGoogle Scholar
- Neumann PM: Finite permutation groups, edge-coloured graphs and matrices. In Topics in Group Theory and Computation. Edited by: Curran MPJ. Academic Press, London; 1977.Google Scholar
- Tsuzuku T: Finite Groups and Finite Geometries. Cambridge University Press, Cambridge; 1982.Google Scholar
- Yılmaz N, Cangül İN: Conjugacy classes of elliptic elements in the Picard group. Turk. J. Math. 2000, 24: 209–220.Google Scholar
- Akbaş M: On suborbital graphs for the modular group. Bull. Lond. Math. Soc. 2001, 33: 647–652. 10.1112/S0024609301008311View ArticleGoogle Scholar
- Jones GA, Singerman D, Wicks K London Mathematical Society Lecture Note Series 160. In The Modular Group and Generalized Farey Graphs. Cambridge University Press, Cambridge; 1991:316–338.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.