A universal theory of switching for combinatorial objects, and applications to complex Hadamard matrices
Dean Crnković, Ronan Egan, Andrea Švob · 2025
Read A universal theory of switching for combinatorial objects, and applications to complex Hadamard matrices by Dean Crnković, Ronan Egan, Andrea Švob (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-07020.txt.
Introduction
Dean Crnković, Ronan Egan, and Andrea Švob unify the idea of switching across combinatorics. Switching transforms an object into a non-isomorphic one of the same type while preserving a key property. Godsil-McKay switching preserves a graph's spectrum, design switching preserves parameters, and code switching preserves minimum distance. For Hadamard matrices, Orrick's switching proved a powerful way to generate inequivalent examples.
The authors give a universal definition of switching that adapts to all these known types. In that language they extend Orrick's methods to Butson Hadamard and complex Hadamard matrices, introducing switchings that construct new inequivalent matrices. They also recast the notion of trades in complex Hadamard matrices within the same terminology, and address an open problem about the permissible size of a trade.
Catalog record
| CREATOR | Dean Crnković, Ronan Egan, Andrea Švob |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-07020
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A universal theory of switching for combinatorial objects, and applications to complex Hadamard matrices Dean Crnković, Ronan Egan, Andrea Švob Abstract The concept of switching has arisen in several different areas within combinatorics. The act of switching usually transforms a combinatorial object into a non-isomorphic object of the same type, in a way that some key property is preserved. Godsil-McKay switching of graphs preserves the spectrum, switching of designs preserves their parameters, and switching of binary codes preserves the minimum distance. For Hadamard matrices, the switching techniques introduced by Orrick proved to be an incredibly powerful tool when enumerating the Hadamard matrices of order 32 32 . In this paper, we introduce a universal definition of switching that can be adapted to incorporate these known types of switching. Through this language, we extend Orrick’s methods to Butson Hadamard and complex Hadamard matrices. We introduce switchings of these matrices that can be used to construct new, inequivalent matrices. We also consider the concept of trades in complex Hadamard matrices in this terminology, and address an open problem on the permissible size of a trade. 2020 Mathematics Subject Classification: 05B05, 05B20 Keywords: Switching of combinatorial objects; complex Hadamard matrices. 1 Introduction Many interesting objects in combinatorics are represented by matrices. Examples include graphs, block designs, Hadamard matrices and weighing matrices, to name a few. In these cases there is typically a notion of equivalence or isomorphism between two objects which is most apparent from the relationship between their representative matrices. For example, two graphs with adjacency matrices A and B are isomorphic is there exists a permutation matrix P such that PAP^{\top}=B . In many cases, motivated either by classification problems or by constructing new objects retaining some specified property, a concept known as switching has developed, which is a means to convert one matrix into another with the same required property, by virtue of replacing some subset of the entries of the matrix with other allowable entries. Perhaps the best known example of this is Godsil-McKay switching of graphs [ 7 ] , building on the initial concept of Seidel [ 14 ] , which converts the adjacency matrix of a graph into another with the same spectrum. Another switching method for graphs that preserves the spectrum was introduced in [ 15 ] by Wang, Qiu and Hu. Some of the current authors recently extended the notion of switching to 2 -designs [ 3 ] which preserves the parameters of the design. This builds on similar concepts used in the enumeration of certain symmetric designs by Denniston in [ 6 ] , and on the transformation of new symmetric or quasi-symmetric designs to new ones with the same parameters by Jungnickel and Tonchev in [ 8 ] . In coding theory, Östergård adapted the concept of switching to construct non-isomorphic binary codes with the same parameters as a given code [ 13 ] . Switching is also a well known action on Latin squares, used to construct new ones from existing ones, see e.g., [ 16 ] . A similar concept was applied to Hadamard matrices by Orrick in [ 12 ] as a means to classify the enormous set of Hadamard matrices of order 32 . This was later found to be equivalent to Godsil-McKay switching [ 1 ] . Orrick’s switching of Hadamard matrices also motivated the concept of a trade in a complex Hadamard matrix introduced by Ó Catháin and Wanless in [ 11 ] . We pay particular attention to these cases as the objects of interest in this paper are complex and Butson type Hadamard matrices. With many different but seemingly similar notions of switching equivalence now in the literature for different objects, it is appropriate to establish a unifying definition, capturing the notions that already exist and laying the framework for applications to combinatorial objects more generally. This is the first of two goals of this paper. The second is to develop switching methods for complex and Butson type Hadamard matrices that capture Orrick’s switching of real Hadamard matrices, develop new methods, and resolve some open problems. Orrick’s switching methods have been applied only to real Hadamard matrices. They have not been appropriately generalised to be applicable to Butson or complex Hadamard matrices more widely, but there are good reasons for doing so now. This can potentially produce several new equivalence classes of Butson matrices, and we might plausibly classify them at reasonable orders. Butson and complex Hadamard matrices are gathering increasing interest as they are important objects in quantum information theory, due to their relationship with mutually unbiased bases of d -dimension complex space \mathbb{C}^{d} and their applications to quantum entanglement [ 4 ] . Another reason for this generalisation to complex matrices is that since a type of Orrick’s switching of Hadamard matrices was shown to be equivalent to the Godsil-McKay switching method for producing cospectral graphs [ 1 , 7 ] , it is reasonable to suspect that a generalisation of Hadamard switching could have wider implications, when seen as an example of switching under our universal definition. The paper is outlined as follows. In the next section, we describe the background to Orrick’s switching of Hadamard matrices which is the main motivation for this work, and introduce notation that will be used throughout the paper. In Section 3 , we introduce our general definitions that apply to switching of different objects. In Section 4 , we describe Orrick’s methods more completely and frame them in the language of our general definitions, and then in Section 5 we introduce switchings of Butson Hadamard matrices that incorporate them. Section 6 introduces a set of general conditions that are sufficient for switching complex Hadamard matrices, capturing the key aspects of the methods of Section 5 that make switching possible. Finally, in Section 7 we recall the definition of a trade from [ 11 ] , consider it in terms of our definition, and answer an open question posed by the authors on the permissible size of a trade in a complex Hadamard matrix. 2 Switching of Hadamard matrices Let \mathbb{T} denote the set of all complex numbers of norm 1 , let \zeta_{k}=e^{\frac{2\pi\sqrt{-1}}{k}} be a primitive k^{\rm th} root of unity, and let \langle\zeta_{k}\rangle be the set of all k^{\rm th} roots of unity generated multiplicatively by \zeta_{k} . A n\times n matrix H with entries in \mathbb{T} is a complex Hadamard matrix of order n if HH^{\ast}=nI_{n} , where H^{\ast} is the conjugate transpose of H , and I_{n} is the n\times n identity matrix. Put succinctly, H is complex Hadamard if it is orthogonal and has entries of norm 1 . A complex Hadamard matrix of order n exists for all n ; take for example, the Fourier matrix F_{n}=[\zeta_{n}^{(i-1)(j-1)}]_{1\leq i,j\leq n} . Interesting questions, both from the perspective of mathematical curiosity, and from that of applications to science, engineering, and the wider field of mathematics, arise from placing restrictions on the entries of H . If the entries are all elements of \langle\zeta_{k}\rangle , then we say that H is a Butson Hadamard matrix. There are finitely many of these at any order, and for a fixed k existence is not