Fact-checked by Grok 2 weeks ago

Graph automorphism

In , a graph automorphism is a bijective from the vertex set of a to itself that preserves adjacency, meaning that two vertices are adjacent their images under the function are adjacent. This mapping represents a structural symmetry of the graph, and the identity mapping (which leaves every vertex fixed) is always a trivial automorphism. The set of all automorphisms of a graph G, denoted \operatorname{Aut}(G), forms a group under , known as the , which fully encodes the symmetries of G. This group is a subgroup of the on the set and plays a central role in distinguishing labelled and unlabelled graphs, as the number of distinct labelings of G is n! / |\operatorname{Aut}(G)|, where n is the number of . Notably, almost all finite graphs are asymmetric, meaning their automorphism group is trivial (containing only the ), with the proportion of such graphs approaching 1 as the number of vertices grows. Furthermore, every finite group arises as the automorphism group of some , a result that connects and deeply. Graph automorphisms have significant applications across disciplines. In , they are essential for testing, canonization (producing a labeling under automorphisms), and enumeration of non-isomorphic graphs. In , automorphism perception algorithms leverage these symmetries to aid in molecular structure elucidation, enabling the identification and comparison of chemical graphs representing molecular configurations. Additional uses include in optimization problems, such as where orbits under the reduce variable redundancy, and in visualizing symmetric structures in .

Fundamentals

Definition

In graph theory, a graph is typically denoted as G = (V, E), where V is the set of vertices and E \subseteq \binom{V}{2} is the set of edges representing unordered pairs of distinct vertices. An automorphism of an undirected graph G is a bijective mapping f: V \to V such that for all distinct vertices u, v \in V, the pair \{u, v\} belongs to E if and only if \{f(u), f(v)\} belongs to E. This permutation of the vertices preserves the adjacency relation, meaning it maps the graph onto itself while maintaining its structural properties. A graph automorphism is a special case of a , where the mapping is from the graph to itself rather than between two potentially distinct graphs. Specifically, while a graph isomorphism \phi: G \to H requires that \{u, v\} \in E_G if and only if \{\phi(u), \phi(v)\} \in E_H for graphs G = (V_G, E_G) and H = (V_H, E_H), an automorphism restricts this to G = H. The concept extends naturally to directed graphs, denoted G = (V, E) where E \subseteq V \times V consists of ordered pairs (arcs). An automorphism here is a f: V \to V such that (u, v) \in E (f(u), f(v)) \in E, preserving the direction of edges. For weighted graphs, where each edge \{u, v\} \in E or (u, v) \in E is assigned a weight w(u, v) \in \mathbb{R}, an f must additionally satisfy w(u, v) = w(f(u), f(v)) for all such pairs, ensuring weights are preserved under the . The set of all automorphisms of a graph G, denoted \operatorname{Aut}(G), forms a group under the operation of function composition, with the identity mapping as the neutral element and inverses given by the inverse permutations. This automorphism group captures the symmetries of G and is a subgroup of the symmetric group on V.

Basic Properties

Graph automorphisms preserve fundamental structural invariants of the graph, including the degree sequence, the number of edges, and connectivity properties, as they maintain adjacency relations between vertices. Specifically, since an automorphism maps adjacent vertices to adjacent vertices and non-adjacent to non-adjacent, the degrees of vertices remain unchanged, ensuring the sorted list of degrees is invariant. The total number of edges is preserved because the mapping is a bijection on the edge set, and connectivity—such as the graph being connected or the number of connected components—is unaltered under this symmetry. In matrix terms, an corresponds to a P such that if A is the of the , then P^T A P = A. This equation reflects the preservation of adjacency: the (i,j)-entry of A equals 1 the permuted entries match, confirming the structure is unchanged. Automorphisms act as permutations on the vertex set, partitioning vertices into orbits under the group action, where an orbit consists of all vertices reachable from a given vertex via repeated applications of automorphisms. Fixed points are vertices in singleton orbits, remaining unchanged by every automorphism in the group. In the cycle decomposition of the permutation representation, cycles correspond to the symmetric mappings within orbits, illustrating how the automorphism rearranges vertices while preserving the graph's edges. Since automorphisms preserve adjacency, they also preserve graph , defined as the length of shortest paths between ; this holds generally but is particularly structured in distance-regular graphs, where the number of at a fixed from any is constant.

Examples

A simple example of a graph automorphism is provided by the K_3, also known as the , which consists of three all connected to each other. The automorphisms of K_3 include permutations of the that preserve adjacency, such as cyclic rotations (mapping 1 to 2, 2 to 3, and 3 to 1) and reflections (such as swapping 2 and 3 while fixing 1). The full \Aut(K_3) is isomorphic to the S_3 of order 6. Cycle graphs C_n for n \geq 3 exhibit rotational and reflectional symmetry. An here corresponds to rotating the cycle or reflecting it over an axis through a and the of the opposite . The \Aut(C_n) is isomorphic to the D_n of order $2n, consisting of n rotations and n reflections. Complete graphs K_n on n vertices, where every pair of distinct vertices is adjacent, have the maximum possible symmetry among graphs with n vertices. Any of the vertices preserves the complete set of edges, so \Aut(K_n) is isomorphic to the S_n of order n!. Path graphs P_n on n vertices, formed by connecting vertices in a linear (e.g., vertices 1-2-...-n), have limited . For n \geq 2, the only non-trivial is the that reverses the , mapping vertex i to n+1-i. Thus, \Aut(P_n) \cong \mathbb{Z}_2, the of order 2. The , a 3-regular graph on 10 vertices often drawn as an outer , an inner star, and connections between them, is a more complex example with high despite its non-planar structure. Its group has order 120 and is isomorphic to the S_5.

Automorphism Groups

Group Structure

The \operatorname{Aut}(G) of a G = (V, E) consists of all bijections \phi: V \to V that preserve adjacency and non-adjacency, forming a group under . This group embeds naturally as a of the S_{|V|}, where each corresponds to a of the vertex set that maintains the graph's structure. The action of \operatorname{Aut}(G) on the vertex set V is faithful, meaning the kernel of the action is trivial: no non-identity automorphism fixes every vertex, ensuring that \operatorname{Aut}(G) is isomorphic to its image in S_{|V|}. This faithful permutation representation highlights how symmetries of the graph translate directly into permutations without redundancy. \operatorname{Aut}(G) is generated by a set of permutations that reflect the graph's symmetries, often expressible in terms of cycles or transpositions corresponding to basic transformations like rotations or reflections in highly symmetric graphs. For instance, in vertex-transitive graphs, generators can be chosen to include elements that cycle vertices within orbits, with the minimal number of such generators bounded by O(\log |V|) in many cases. Key subgroups of \operatorname{Aut}(G) include the stabilizers of individual vertices or edges; the vertex stabilizer \operatorname{Stab}(v) = \{\phi \in \operatorname{Aut}(G) \mid \phi(v) = v\} consists of automorphisms fixing a particular v, and similarly for edges, these stabilizers capture local symmetries. Normal subgroups arise in the structure of \operatorname{Aut}(G) when considering quotients that preserve the action, such as minimal normal subgroups in primitive actions, which often take abelian forms like elementary abelian p-groups. Early foundational work linking group actions to graphical enumerations dates to in 1878, who introduced representations of abstract groups via labeled digraphs, paving the way for modern understandings of as permutation groups acting on .

Order and Orbit-Stabilizer Theorem

The order of the automorphism group \Aut(G) of a graph G with vertex set V of size n = |V| can be determined using tools from , particularly the orbit-stabilizer theorem, which relates the size of the group to the sizes of orbits and stabilizers under its natural action on the vertices. The orbit-stabilizer theorem states that for any vertex v \in V, |\Aut(G)| = |\orbit(v)| \cdot |\Stab(v)|, where \orbit(v) = \{ \phi(v) \mid \phi \in \Aut(G) \} is the of v and \Stab(v) = \{ \phi \in \Aut(G) \mid \phi(v) = v \} is the subgroup fixing v. This relation holds because \Aut(G) acts as a on V, and the theorem applies to any . If G is vertex-transitive (i.e., \Aut(G) acts transitively on V, so |\orbit(v)| = n for all v), then |\Aut(G)| = n \cdot |\Stab(v)|. The size of the stabilizer can often be computed by examining the action on the neighbors of v or other structural features. For non-transitive graphs, the order |\Aut(G)| is the product of the sizes of the orbits times the orders of their corresponding stabilizers, obtained recursively by applying the theorem to the orbits and their point stabilizers. This recursive application leverages the structure of the permutation representation of \Aut(G) to build up the full order without enumerating all elements. Burnside's lemma provides a complementary tool for analyzing the action of \Aut(G) by counting the number of orbits in related sets, such as proper vertex colorings or subgraphs, via the average number of fixed points: the number of orbits is \frac{1}{|\Aut(G)|} \sum_{\phi \in \Aut(G)} \fix(\phi), where \fix(\phi) is the number of elements fixed by \phi. Although this formula requires knowledge of |\Aut(G)| to apply directly, it can verify computed orders by checking consistency with known orbit counts in symmetric structures. For example, consider the cycle graph C_4 on 4 vertices, which is vertex-transitive. The stabilizer of any vertex consists of the identity and the reflection through that vertex and its opposite, so |\Stab(v)| = 2. Thus, |\Aut(C_4)| = 4 \cdot 2 = 8, corresponding to the dihedral group of order 8. Similarly, for the complete graph K_n, |\Stab(v)| = (n-1)! (permutations of the remaining vertices), yielding |\Aut(K_n)| = n \cdot (n-1)! = n!.

Computational Aspects

Complexity of Recognition

The decision problem of determining whether a given graph has a non-identity automorphism, often denoted as the GRAPH AUTOMORPHISM problem, lies in the complexity class NP. A nondeterministic Turing machine can solve it by guessing a permutation of the vertices and verifying in polynomial time whether it preserves the edge set, thereby confirming a nontrivial automorphism if one exists. However, the problem is not known to be NP-complete; while restricted variants, such as deciding the existence of a fixed-point-free automorphism, are NP-complete, the general case is widely believed to be NP-intermediate. The GRAPH AUTOMORPHISM problem is intimately related to the , with both residing in and sharing equivalent reductions in many settings. In 2015, announced a quasi-polynomial-time for GRAPH ISOMORPHISM, running in time \exp(O((\log n)^c)) for some constant c > 0, which implies the same complexity for recognizing nontrivial automorphisms since the can be computed using a polynomial number of isomorphism tests. This breakthrough was confirmed through subsequent refinements and verifications in the , establishing quasi-polynomial solvability for both problems without resolving whether they are in P. In the worst case, naive for GRAPH AUTOMORPHISM require exponential time, such as O(n!) for exhaustive search over all permutations of n vertices, reflecting the inherent difficulty for dense or highly symmetric graphs. However, for graphs of bounded maximum d, Luks developed a polynomial-time in 1982, leveraging group-theoretic techniques to reduce the search effectively. As of 2025, no polynomial-time exists for the general GRAPH AUTOMORPHISM problem, maintaining its as a cornerstone of . Recent developments have focused on algebraic approaches, including linear algebra methods over finite fields, which enhance decomposition techniques in Babai's framework and yield improved bounds for structured graph classes, though the general case remains quasi-polynomial.

Connection to Graph Isomorphism

Graph automorphisms play a central role in addressing the , which asks whether two graphs are structurally identical up to relabeling of vertices. The Aut(G) of a graph G encodes the symmetries of G, and these symmetries can be leveraged to reduce isomorphism testing to more tractable computations. Specifically, by identifying invariants under Aut(G), one can map non-isomorphic labelings to a common representative form, allowing direct comparison of graphs for . A key technique linking automorphisms to isomorphism is canonical labeling, which assigns to each a unique labeling invariant under its . This process reduces the isomorphism problem to checking: two graphs are isomorphic their canonical labelings are identical. Computing a involves exploring the orbits of Aut(G) to select a "minimal" labeling among all possible automorphic equivalents, often using backtrack search refined by symmetry detection. For instance, Brendan McKay's algorithm in the nauty software package computes such labelings by generating automorphisms to prune redundant branches, achieving practical efficiency for graphs up to thousands of vertices. This approach exploits the structure of Aut(G) to avoid enumerating all n! labelings, particularly when |Aut(G)| is large, as symmetries collapse many equivalents into fewer orbits. The method provides another bridge between and through color refinement, which iteratively partitions vertices into color classes based on neighborhood structures. Stabilizers in this refinement process correspond to under Aut(G), as equivalent vertices under receive the same color. The k-dimensional WL algorithm stabilizes to reveal these , enabling testing by comparing refined color partitions; if the partitions differ, the graphs are non-. This connection highlights how Aut(G) influences the method's power: full detection via WL often suffices for practical , though higher dimensions may be needed for graphs with rich symmetries. The original formulation by Weisfeiler and Lehman in , extended in Weisfeiler's 1976 work, ties the method to cellular algebras where groups act as centralizers preserving structures. The size and structure of Aut(G) further modulate the difficulty of isomorphism testing. If Aut(G) is trivial (i.e., |Aut(G)| = 1, containing only the ), isomorphism becomes relatively easier, as there are no non-trivial symmetries to account for in matching vertices, allowing straightforward refinement without orbit collapsing. Conversely, a large Aut(G) can aid efficiency by reducing the search space through symmetry exploitation in canonical forms, but it may hinder if the group is computationally hard to generate, as in highly symmetric graphs like strongly regular ones. This duality underscores the interplay: trivial automorphisms simplify direct comparisons, while expansive ones demand robust group computation to leverage for faster resolution. Historically, the connection traces to George Pólya's 1937 enumeration theorem, which uses group actions—precursors to modern automorphism computations—to count distinct graphs up to . Pólya's method applies over the action of the on potential edge sets, yielding generating functions whose coefficients give isomorphism class sizes. This framework influenced early isomorphism algorithms by emphasizing counting under group actions, providing a combinatorial foundation for later algebraic approaches to Aut(G)-invariant forms.

Algorithms for Computation

Computing the automorphism group of a graph typically involves backtrack search algorithms that explore possible permutations while pruning the search space using symmetries detected during the process. The Nauty algorithm, introduced by Brendan D. McKay in 1981, is a seminal backtrack-based method that computes generators for the by constructing a canonical labeling of the . It employs partition refinement to group vertices into equitable classes based on their structural roles, iteratively splitting partitions until they distinguish non-equivalent vertices or reveal automorphisms. A core technique in Nauty is the individualization-refinement procedure, where a from a non-trivial in the current is selected and "individualized" by assigning it a unique color, followed by refining the to propagate this distinction through the graph's . This process is repeated, building a nest that guides the backtrack search until a base of the group is found, allowing of automorphisms. The search prunes branches using the to avoid redundant explorations of equivalent configurations. Once candidate generators are identified through , the Schreier-Sims is adapted to represent and process the efficiently, computing a base and strong generating set (BSGS) to determine the group's order and orbits. In implementations like Nauty and its successor Traces, a randomized variant of the Schreier method is used to manage the group elements, ensuring probabilistic completeness while handling large groups. Traces, developed as an extension in the by and Adolfo Piperno, enhances Nauty's with breadth-first strategies for certain subproblems, improving performance on dense graphs. In the worst case, these backtrack algorithms exhibit bounded by O(|V|! / |\mathrm{Aut}(G)|), reflecting the number of potential permutations divided by the group order, though practical heuristics like early and refined invariants make them efficient for graphs with up to several thousand vertices.

Applications and Tools

Uses in Graph Algorithms

automorphisms play a key role in symmetry reduction techniques for algorithms, particularly in and problems, where they enable partitioning to significantly reduce the explored state space. In probabilistic model checking, automorphisms define permutations that preserve transition probabilities in models like discrete-time Markov chains, allowing states to be grouped into —equivalence classes under the —such that only one representative per needs to be analyzed, reducing the state space from size M^N (for N symmetric components each with M local states) to a fraction like \binom{M+N-1}{N}. This approach has been implemented using multi-terminal decision diagrams to efficiently compute and apply the reduction without loss of accuracy. Similarly, in problems modeled as , automorphisms facilitate partitioning variables into symmetric , redundant search branches and accelerating solutions for symmetric instances like scheduling or circuit . Graph , which produces a unique labeled representation invariant under , is essential for efficient querying in graph databases and comparing . In database querying, normalizes graphs to enable fast checks during matching or similarity searches, avoiding redundant computations over symmetric labelings. For comparison, partitioning identifies symmetric atoms and bonds in molecular graphs, enabling polynomial-time canonical labeling via iterative refinement of invariants based on extended and planar embedding properties, which has been applied to large structures like fullerenes with up to thousands of atoms. This ensures unique string representations (e.g., canonical SMILES) for database indexing and substructure retrieval in cheminformatics. In network analysis, graph automorphisms aid in detecting symmetric motifs—recurrent subgraphs with high automorphism groups—in social and biological networks, revealing structural redundancies that inform functional insights. For biological networks, such as protein interaction graphs, the symmetry compression method exploits automorphisms to compress symmetric subgraphs before enumeration, eliminating isomorphic duplicates and speeding up motif discovery by orders of magnitude in highly symmetric topologies, while preserving exact results through lossless decompression. In social networks, automorphisms help identify symmetric communities or roles by partitioning vertices into orbits, enabling scalable detection of motifs like balanced triads that indicate stable relationships. The Aut(G) is central to algorithms via Pólya's theorem, which counts distinct graphs up to by averaging the number of fixed colorings over group elements, applied to edge or vertex labelings to generate non-isomorphic structures. This method, originally developed for counting chemical compounds and graphs, uses the of Aut(G) to compute the number of unlabeled graphs with given properties, such as trees or regular graphs, avoiding exhaustive of the $2^{\binom{n}{2}} labeled graphs on n vertices. Recent developments in leverage graph automorphisms for constructing secure functions based on symmetric graph structures, enhancing post-quantum resistance. In group-subgroup pair graphs, automorphisms induced by actions ensure vertex-transitive properties, allowing outputs invariant to symmetric traversals and reducing collision risks through normalized walks analyzed via group presentations. Similarly, structural hashing of directed s employs orbits for node representations, guaranteeing that isomorphic subgraphs yield identical hashes while maintaining , as proven for s under .

Software Implementations

Nauty and Traces form a widely used command-line suite for computing automorphisms and canonical labelings, particularly effective for both dense and sparse . Developed by Brendan McKay and Adolfo Piperno, the tools employ refinement and backtrack search to determine the full as a set of generators, supporting with up to millions of vertices depending on availability, though practical limits around 100,000 vertices are common for dense instances on standard hardware. SageMath integrates graph automorphism computation within its broader mathematical framework, returning the automorphism group as a permutation group object that leverages GAP for subsequent group-theoretic analysis. The automorphism_group() method refines an optional vertex partition to compute the largest equitable subgroup, supporting both undirected and directed graphs, and optionally uses external libraries like Bliss for faster execution on large inputs. This integration allows seamless exploration of group properties such as order, orbits, and generators directly in a Python-like environment. NetworkX, a library for analysis, provides automorphism support through its VF2++ implementation, which enables checking whether a proposed preserves structure and can thus verify individual or generate partial mappings for self-. While it does not compute the full natively, the matchers facilitate detection by treating the against itself, making it suitable for smaller graphs or integration with custom backtrack routines. The system, via its package, specializes in computations tailored to graph automorphisms, representing Aut(G) as a of the on vertices. GRAPE interfaces with Nauty or Bliss to efficiently calculate generators of the , even for colored graphs, and supports testing between graphs; for example, it handles the automorphism group of the Johnson graph J(4,2) with order 48. This makes GAP ideal for applications where group structure is central.

Symmetry Visualization

Methods for Displaying Symmetry

One common technique for displaying the symmetries of a graph induced by its is the use of diagrams, which partition the set into under the and represent these as , sets, or connected components to illustrate equivalence classes and their interrelations. For instance, in distance-regular graphs, diagrams depict how the acts on and edges, often showing fixed points, structures, and adjacency between to convey the overall without rendering the full . This method is particularly effective for abstracting large symmetries, as seen in classifications of highly graphs where are enumerated and diagrammed to highlight transitive actions. Graph drawing methods, particularly force-directed layouts, can be designed to respect automorphism actions by positioning vertices in configurations that mirror the group's operations, such as circular arrangements for cyclic automorphisms or radial symmetries for dihedral groups. These layouts enforce geometric isometries that correspond to automorphisms, ensuring that rotations or reflections in the drawing induce valid graph mappings; for example, algorithms using circular grids place orbit representatives at symmetric points and propagate positions via group elements to achieve balanced, aesthetically symmetric renderings. Linear-time algorithms exist for planar graphs, leveraging connectivity decompositions like SPQR-trees to embed symmetries while preserving planarity. Matrix representations offer a algebraic visualization of automorphisms by depicting the adjacency matrix and the corresponding permutation matrices that conjugate it to itself, illustrating how vertex relabelings preserve the graph's structure. An automorphism corresponds to a P such that P A P^T = A, where A is the ; these can be visualized as overlaid matrices or sequences, with heatmaps or overlays showing permuted entries to demonstrate invariance under . This method is especially useful for computational verification and understanding effects on sparse or dense graphs. Interactive methods enable dynamic exploration of symmetries through web-based interfaces that apply automorphisms step-by-step, often animating and mappings to reveal the group's in real-time. Users can select group elements to observe transformations, such as cycling through orbits or reflecting subgraphs, providing an intuitive grasp of how symmetries preserve ; these approaches typically integrate with tools to update layouts instantaneously, facilitating educational and analytical insights into complex groups.

Tools and Techniques

, a widely used open-source software, supports symmetry-aware layouts through its language, where users can specify node groups and attributes to align symmetric structures, enhancing the visual representation of automorphisms. By assigning the same group ID to vertices in the same under the , the layout engines like neato or fdp can produce more balanced and symmetric drawings that reflect the underlying symmetries. Tools such as and offer plugins and built-in features for orbit-based clustering and symmetry overlays in graph visualization. In , automatic layout algorithms can be configured to cluster nodes based on custom properties derived from orbit partitions, allowing users to overlay symmetry indicators like color-coding for equivalent vertices under automorphisms. Similarly, 's modular architecture enables the use of community detection plugins adapted for orbit clustering, with visual overlays such as edge highlighting to depict automorphism actions. As of 2025, (VR) and (AR) techniques have advanced for immersive visualization of high-dimensional automorphism actions, particularly in molecular graphs where automorphisms correspond to symmetries. The (VMD) software, with its Symmetry Tool plugin, allows users to analyze and display symmetry operations in 3D molecular structures, integrated with VR interfaces like for interactive exploration of automorphism-induced transformations in complex biomolecules. Animation tools facilitate the generation of GIFs or videos demonstrating automorphism applications, such as rotations in 3D embeddings of symmetric . Python's graph-tool supports animating layouts and can be extended to illustrate mappings by sequentially applying permutations to positions, exporting frames for GIF creation via libraries like imageio. For instance, rotations in cubic or polyhedral can be visualized as smooth transitions preserving the . Benchmarking tools like AutoGraphiX aid in generating symmetric graphs, such as regular or distance-regular graphs, using . It displays their structures in interactive XY-plots, enabling users to explore properties and generate conjectures on symmetries.

Notable Graph Families

Vertex-Transitive Graphs

A is vertex-transitive if its acts transitively on the set, meaning that for any two vertices u and v, there exists an mapping u to v. This condition is equivalent to the vertices forming a single under the . Vertex-transitive graphs are necessarily , with all vertices having the same , as the automorphism group permutes vertices while preserving adjacency. This symmetry implies that the graph is distance regular: for any fixed distance k, every vertex has the same number of vertices at k from it, providing a foundation for stronger regularity properties such as distance-regularity in many cases. For instance, the intersection numbers in distance-regular vertex-transitive graphs are independent of the starting vertex, facilitating analysis of their and combinatorial structure. Prominent examples include complete graphs K_n for n \geq 1, where the full symmetric group acts transitively on vertices; cycle graphs C_n for n \geq 3, with the dihedral group providing the transitivity; and Cayley graphs, which are constructed from a group G and a generating set S \subseteq G, inherently vertex-transitive via left multiplication by group elements. Hypercube graphs Q_n, as a subclass of Hamming graphs, also exemplify vertex-transitivity through the action of the hyperoctahedral group. Vertex-transitive graphs can be constructed via group actions, such as defined by a group and symmetric connection set, ensuring the yields . Voltage graphs provide another method, where assignments of group elements (voltages) to edges of a base produce graphs that inherit vertex-transitivity under suitable conditions, often yielding infinite families of symmetric structures. Similarly, , arising from the action of a group on the of a , are vertex-transitive by construction, as the coset space admits a transitive . Enumeration efforts reveal finitely many vertex-transitive graphs for small vertex orders, with complete censuses available for small orders, including up to 47 vertices as of 2019, identifying specific counts such as 64 connected graphs on 12 vertices. However, infinite families abound, including the cycles C_n, complete graphs K_n, and hypercubes Q_n, demonstrating the abundance of such graphs across all orders.

Arc-Transitive and Symmetric Graphs

An is a in which the acts transitively on the set of , where an is an of adjacent vertices. This property implies both vertex-transitivity and edge-transitivity, as the action on preserves the underlying symmetries of vertices and undirected edges. Symmetric graphs generalize this notion, defined as graphs that are s-arc-transitive for some integer s ≥ 1, meaning the acts transitively on the set of s-arcs—a sequence of s+1 vertices where consecutive vertices are adjacent and no three consecutive vertices repeat (no immediate ). For s=1, s-arc-transitivity coincides with arc-transitivity. Graphs that are s-arc-transitive for all finite s are called highly arc-transitive. Complete graphs K_n for n \geq 2 are \infty-arc-transitive, as their automorphism group, the S_n, permits arbitrary permutations of vertices while preserving adjacency. The , a on 10 vertices, is 3-arc-transitive but not 4-arc-transitive, exemplifying a finite-degree . For small s, classifications exist: 1-arc-transitive graphs are precisely the arc-transitive ones, while for cubic (3-regular) graphs, proved that every finite connected symmetric cubic graph is s-arc-transitive for some s ≤ 5. More broadly, R. Weiss established that the only finite connected s-arc-transitive graphs for s ≥ 8 are cycles, providing an upper bound on the possible s for non-cyclic finite graphs. As of 2025, research on related symmetries has advanced with the discovery of new infinite families of half-arc-transitive graphs—graphs that are - and edge-transitive but not arc-transitive—including tetravalent examples where vertex-stabilizers are fully classified.

References

  1. [1]
    [PDF] Automorphisms of graphs - vlsicad page
    An automorphism of a graph G is a permutation g of the vertex set of G with the property that, for any vertices u and v, we have ug ∼ vg if and only.
  2. [2]
    Graph Automorphism -- from Wolfram MathWorld
    An automorphism of a graph is a graph isomorphism with itself, i.e., a mapping from the vertices of the given graph · The group of graph automorphisms of a graph ...
  3. [3]
    Graph automorphism perception algorithms in computer-enhanced ...
    Mar 1, 1993 · Graph automorphism perception algorithms in computer-enhanced structure elucidation | Journal of Chemical Information and Modeling.
  4. [4]
    Applications of Graph Automorphisms
    Nov 1, 2016 · We can use graph automorphisms to compute the orbits of variables in the linear programming problem, and then treat parts with the same orbit as ...
  5. [5]
    Definitions - Discrete Mathematics
    Instead, here is the (now) standard definition of a graph. Graph Definition. A graph is an ordered pair G=(V,E) G = ( V , E ) consisting of a nonempty set V V ...
  6. [6]
    [PDF] 2.1 Graph Isomorphism 2.2 Automorphisms and Symmetry 2.3 ...
    Def 2.1. An isomorphism from a graph G to itself is called an automor- phism. Thus, an automorphism π of graph G is a structure-preserving permutation πV on VG ...
  7. [7]
    [PDF] Divide and Conquer - Series-Parallel Graphs - KIT - ITI Algorithmik
    Definition: Automorphism of a digraph. An automorphism of a directed graph G = (V,E) is a permutation of the vertex set which preserves adjacency of the ...
  8. [8]
    [PDF] Automorphisms, Equitable Partitions, and Spectral Graph Theory
    Oct 7, 2016 · We also allow weighted graphs. (both directed and undirected), that is, graphs in which each edge ( {i, j} or (i, j)) is assigned a numerical.
  9. [9]
    [PDF] Graph Automorphism Groups - East Tennessee State University
    Feb 23, 2018 · A graph automorphism is simply an isomorphism from a graph to itself. In other words, an automorphism on a graph G is a bijection φ : V(G) → V( ...
  10. [10]
    [PDF] IJMC On Symmetry of Some Nano Structures
    For a given adjacency matrix A, we can write a simple GAP program to calculate all the permutation matrices with PtAP = A. Using this program and a similar.
  11. [11]
    [PDF] arXiv:2404.14976v1 [math.QA] 23 Apr 2024
    Apr 23, 2024 · The next lemma is a quantum version of the fact that any automorphism of a graph has to preserve the distance of any pair of vertices.
  12. [12]
    [PDF] Automorphism Group of Graphs (Supplemental Material for Intro to ...
    Jan 15, 2018 · In this supplement, we will assume that all graphs are undirected graphs with no loops or multiple edges. In graph theory, we talk about ...
  13. [13]
    [PDF] Graph Automorphism Group Equivariant Neural Networks - arXiv
    Example 2.11. The automorphism group of the cycle graph on n vertices, Aut(Cn), is isomorphic to the dihedral group. Dn of order 2n ...
  14. [14]
    Complete Graph -- from Wolfram MathWorld
    K_n are n . The automorphism group of the complete graph Aut(K_n) is the symmetric group S_n (Holton and Sheehan 1993, p. 27). CompleteGraphCycles. The ...
  15. [15]
    [PDF] Sensing and Control in Symmetric Networks arXiv:1507.08044v1 ...
    Jul 29, 2015 · The automorphism group of the Petersen graph is iso- morphic to S5, i.e. Aut(P). ∼. = S5. That is, there are 5! = 120 automorphisms. The group ...Missing: S_5 | Show results with:S_5
  16. [16]
    [PDF] Automorphism groups, isomorphism, reconstruction (Chapter 27 of ...
    Jun 12, 1994 · Automorphisms of the graph X = (V,E) are X → X isomorphisms; they form the subgroup Aut(X) of the symmetric group Sym(V ). Automorphisms of ...
  17. [17]
    (PDF) Some problems on Cayley graphs - ResearchGate
    Aug 6, 2025 · This survey paper presents the historical development of some problems on Cayley graphs which are interesting to graph and group theorists.
  18. [18]
    [PDF] On Finding the Number of Graph Automorphisms
    Nov 14, 1997 · 6. Page 7. enumerability of #GA and the complexity of GI might help us obtain a better classification of the Graph Isomorphism problem.<|control11|><|separator|>
  19. [19]
    Some NP-Complete Problems Similar to Graph Isomorphism
    The paper presents altered versions of the Graph Isomorphism problem, such as determining if a graph has a fixed-point-free automorphism, that are NP-complete.
  20. [20]
    [1512.03547] Graph Isomorphism in Quasipolynomial Time - arXiv
    Dec 11, 2015 · Authors:László Babai. View a PDF of the paper titled Graph Isomorphism in Quasipolynomial Time, by L\'aszl\'o Babai. View PDF. Abstract:We show ...
  21. [21]
    [1710.04574] Graph isomorphisms in quasi-polynomial time - arXiv
    Oct 12, 2017 · Babai has recently shown how to solve these problems -- and others linked to them -- in quasi-polynomial time, i.e. in time \exp\left(O(\log n) ...
  22. [22]
    [PDF] On the Relative Power of Linear Algebraic Approximations of Graph ...
    In this paper, we consider two distinct methods for incorporating algorithms for solving linear systems into graph isomorphism solvers and compare them. The ...
  23. [23]
    Practical graph isomorphism, II - ScienceDirect.com
    When we have an efficient canonical labelling procedure, we can use a sorting, hashing or balanced tree algorithm for removing isomorphs from a large collection ...
  24. [24]
    [PDF] PRACTICAL GRAPH ISOMORPHISM - Brendan D. McKay
    In this paper we discuss the design of an algorithm for canoni- cally labelling a vertex-coloured graph and for finding generators for its automorphism group.
  25. [25]
    Introduction - Nauty Traces
    Sep 7, 2025 · The process of finding the canonical member of the isomorphism class containing a given object is called canonical labeling. Two labeled ...
  26. [26]
    [PDF] Lecture Notes in Mathematics - Weisfeiler.Com
    5 to find some orbits of the automorphism group of the graph under consideration. Another essential feature is the procedure designed to deal with correct.
  27. [27]
    Isomorphism Testing and Symmetry of Graphs - ScienceDirect.com
    We survey some aspects of the complexity of graph isomorphism testing and its relation to the size and structure of the automorphism group.
  28. [28]
    The nauty Traces page
    The original design of nauty is in B. D. McKay, Practical Graph Isomorphism, Congressus Numerantium, 30 (1981) 45-87. A scan is available. The original ...
  29. [29]
  30. [30]
    [PDF] nauty and Traces User's Guide (Version 2.8.9)
    The canonically labelled graph produced by nauty or Traces is ... nauty and Traces use the Random Schreier Method to process the automorphism group.
  31. [31]
    Nauty Traces – Home
    Nauty and Traces are programs for computing automorphism groups of graphs and digraphs, and can produce a canonical label.Automorphism Group · Introduction · Traces · Search Tree
  32. [32]
    Generic graphs (common to directed/undirected) - Graph Theory
    Automorphism group:​​ Return the coarsest partition which is finer than the input partition, and equitable with respect to self . Return the largest subgroup of ...
  33. [33]
    Isomorphism — NetworkX 3.5 documentation - Algorithms
    vf2pp_isomorphism : to obtain the node mapping between two graphs, in case they are isomorphic. vf2pp_all_isomorphisms : to generate all possible mappings ...
  34. [34]
    [grape] 8 Automorphism groups and isomorphism testing for graphs
    Please also note that a canonical labelling for a GRAPE graph is the inverse of what a canononical labelling for a graph is usually defined as (such as in bliss) ...
  35. [35]
    The GraphBLAS | Welcome to the GraphBLAS Forum
    The GraphBLAS Forum is an open effort to define standard building blocks for graph algorithms in the language of linear algebra.Missing: automorphism 2025
  36. [36]
    [PDF] On the 486-vertex distance-regular graphs of Koolen–Riebeek and ...
    Aug 19, 2019 · Also, in Figures 2, 4 and 6 we give the distance distribution diagrams for each of the three graphs. (The orbit diagram of ∆ is taken from [9].).
  37. [37]
    [PDF] Symmetric Graph Drawing - Brown CS
    An isometry is a mapping of the plane onto itself that preserves distances. ... Lemma 3.2 is folklore in graph automorphism theory; a proof is in [HME06].
  38. [38]
    [PDF] Automorphisms, Equitable Partitions, and Spectral Graph Theory
    Aug 1, 2017 · Intuitively, a graph automorphism describes how parts of a graph can be interchanged in a way that preserves the graph's overall structure.
  39. [39]
    [PDF] GraphShop: An Interactive Software Environment for Graph Theory ...
    The graph drawing panel contains two methods of controlling the layout or position of the vertices, edges, and arcs being drawn. The first, custom drag-and-drop ...
  40. [40]
    Layout Engines | Graphviz
    Oct 4, 2022 · Various algorithms for projecting abstract graphs into a space for visualization. dot hierarchical or layered drawings of directed graphs.Dot · Neato · Fdp · Nop
  41. [41]
    Tighten the dot graph making it more symmetric - Stack Overflow
    Jul 15, 2011 · To get the layout to be more symmetric, you may try to align the nodes Waiting and Terminated as well as Timed Waiting and Blocked by setting their group ...Missing: aware automorphisms
  42. [42]
    yEd - Graph Editor - yWorks
    yEd is a free desktop application to quickly create, import, edit, and automatically arrange diagrams. It runs on Windows, macOS, and Unix/Linux.Download · yEd Live · Gallery · Diagram editorsMissing: automorphism | Show results with:automorphism
  43. [43]
    Gephi plugins
    Gephi is the leading visualization and exploration software for all kinds of graphs and networks. Gephi is open-source and free.
  44. [44]
    A Resource for Chemical Education with VMD and SYVA Programs
    Sep 6, 2024 · To enhance the visualization of molecular symmetry, modifications were made to both the symmetry tool plugins and the SYVA program. Figure 1 ...
  45. [45]
    VMD as a Platform for Interactive Small Molecule Preparation and ...
    The Symmetry Tool plugin in VMD is a powerful resource for analyzing point group symmetry in small molecules. Point group symmetry is a fundamental concept in ...
  46. [46]
    Animations with graph-tool
    The drawing capabilities of graph-tool (see draw module) can be harnessed to perform animations in a straightforward manner. Here we show some examples which ...
  47. [47]
    AutoGraphiX: Home
    AutoGraphiX (AGX) is a computer system designed to help researchers in graph theory. The main purpose of AGX is to search for extremal graphs, i.e., graphs ...
  48. [48]
    [PDF] Getting started with AutoGraphiX-III (version 3.1.X) - GERAD
    AutoGraphiX-III is a computer aided graph theory system. This docu- ment proposes a set of progressive examples using AGX with step by step explanations. It ...
  49. [49]
    [PDF] Vertex-transitive graphs and their arc-types - arXiv
    May 8, 2015 · Vertex-transitive graphs hold a significant place in mathematics, dating back to the time of first recognition of the Platonic solids, and also ...
  50. [50]
    [PDF] Webs of Complete Graphs - D
    Oct 3, 2022 · Let G be a graph. If G is vertex transitive, then G is distance degree regular. Example 2.23. Three distance degree regular graphs with ...
  51. [51]
    [PDF] Distance mean-regular graphs - UPCommons
    Mar 31, 2016 · The motivation for studying and characterizing distance mean-regular graphs is that they generalize both the vertex-transitive and the distance ...Missing: implications | Show results with:implications
  52. [52]
    [PDF] Structural characterization of Cayley graphs - arXiv
    Sep 27, 2016 · Every Cayley graph is a graph with high symmetry: it is vertex-transitive meaning that the action of its automorphism group is transitive, or ...
  53. [53]
    Common graphs - Graph Theory
    All Hamming graphs are regular, vertex-transitive and distance-regular. Hamming graphs with parameters ( 1 , q ) represent the complete graph with q ...
  54. [54]
    [PDF] arXiv:2206.05583v2 [math.CO] 4 Jul 2024
    Jul 4, 2024 · Covering graphs are not vertex-transitive graphs in general, but they are still highly symmetric, and hence they share many properties with ...
  55. [55]
    [PDF] Up to a double cover, every regular connected graph is isomorphic ...
    Jan 19, 2022 · We prove that every connected locally finite regular graph is either isomorphic to a Schreier graph, or has a double cover which is isomorphic.
  56. [56]
    [PDF] A Census of Vertex-Transitive Graphs - Northern Arizona University
    We provide the census of all vertex-transitive graphs with no more than 12 vertices. While the computer generated the vertex-transitive graphs, we prove some ...Missing: enumeration | Show results with:enumeration
  57. [57]
    Arc-Transitive Graph -- from Wolfram MathWorld
    An arc-transitive graph, sometimes also called a flag-transitive graph, is a graph whose graph automorphism group acts transitively on its graph arcs.
  58. [58]
    Arc-Transitive Graphs - SpringerLink
    An arc in a graph is an ordered pair of adjacent vertices, and so a graph is arc-transitive if its automorphism group acts transitively on the set of arcs.
  59. [59]
    Symmetric graphs (Chapter 17) - Algebraic Graph Theory
    A vertex-transitive graph is symmetric if and only if each vertex-stabilizer Gv acts transitively on the set of vertices adjacent to v. For example, there are ...
  60. [60]
    On s-arc transitive hypergraphs - ScienceDirect.com
    Weiss [10] proved several years later that the only finite connected s -arc transitive graphs with s ≥ 8 are the cycles. Given ...
  61. [61]
    [2508.21336] On tetravalent half-arc-transitive graphs - arXiv
    Aug 29, 2025 · In this paper, we solve this problem by proving that a group is the vertex-stabilizer of a connected tetravalent half-arc-transitive graph if ...Missing: results | Show results with:results