This paper was converted on www.awesomepapers.org from LaTeX by an anonymous user.
Want to know more? Visit the Converter page.

Set-theoretical solutions to the Hom-Yang-Baxter equation and Hom-cycle sets

Kaiqiang Zhang  and  Xiankun Du K. Zhang: School of Mathematics, Jilin University, Changchun 130012, China [email protected] X. Du: School of Mathematics, Jilin University, Changchun 130012, China [email protected]
Abstract.

Set-theoretic solutions to the Yang-Baxter equation have been studied extensively by means of related algebraic systems such as cycle sets and braces, dynamical versions of which have also been developed. No work focuses on set-theoretic solutions to the Hom-Yang-Baxter equation (HYBE for short). This paper investigates set-theoretic solutions to HYBE and associated algebraic system, called Hom-cycle sets. We characterize left non-degenerate involutive set-theoretic solutions to HYBE and Hom-cycle sets, and establish their relations. We discuss connections among Hom-cycle sets, cycle sets, left non-degenerate involutive set-theoretic solutions to HYBE and the Yang-Baxter equation.

Key words and phrases:
cycle set, Hom-cycle set, Hom-Yang-Baxter equation, left quasigroup, set-theoretic solution
1991 Mathematics Subject Classification:
16T25, 20N02, 20N05
Corresponding author: X Du

1. Introduction

Let VV be a vector space. A solution to the Yang-Baxter equation (YBE shortly) is a linear map R:VVVVR:V\otimes V\rightarrow V\otimes V such that

(RidV)(idVR)(RidV)=(idVR)(RidV)(idVR).(R\otimes\operatorname{id}_{V})(\operatorname{id}_{V}\otimes R)(R\otimes\operatorname{id}_{V})=(\operatorname{id}_{V}\otimes R)(R\otimes\operatorname{id}_{V})(\operatorname{id}_{V}\otimes R).

YBE first appeared in the work of Yang [39] and Baxter [1]. This equation is fundamental to quantum groups. It is a central task to determine all solutions to YBE. However, this is difficult to accomplish. In order to find new solutions to YBE, Drinfeld [12] in 1992 suggested considering set-theoretic solutions to YBE, that is, a map r:X×XX×Xr:X\times X\rightarrow X\times X, where XX is a nonempty set, satisfying

(r×idX)(idX×r)(r×idX)=(idX×r)(r×idX)(idX×r).(r\times\operatorname{id}_{X})(\operatorname{id}_{X}\times r)(r\times\operatorname{id}_{X})=(\operatorname{id}_{X}\times r)(r\times\operatorname{id}_{X})(\operatorname{id}_{X}\times r). (1.1)

Etingof, Schedler and Soloviev [16], Gateva-Ivanova and Van den Bergh
 [20], and Lu, Yan and Zhu [26] initially conducted a systematic study on this subject. They studied set-theoretic solutions with invertibility, nondegeneracy and involutivity by using group theory. Gateva-Ivanova in [17] introduced a combinatorial approach to discuss set-theoretic solutions and she conjectured that every square-free, non-degenerate involutive set-theoretic solution (X,r)(X,r) is decomposable whenever XX is finite. This has been proved by Rump in [30].

Left cycle sets were introduced by Rump [30] to study left non-degenerate involutive set-theoretic solutions to YBE. Rump showed that there is a bijective correspondence between left non-degenerate involutive set-theoretic solutions to YBE and left cycle sets, and non-degenerate solutions correspond to non-degenerate left cycle sets. He also prove that all finite left cycle sets are non-degenerate. The theory of cycle sets has been proved to be very useful in understanding the structure of solutions to YBE (see for example [2, 3, 4, 5, 7, 6, 10, 25]). This theory has been greatly developed, and inspires theory of braces [8, 18, 21, 31, 36].

Another version of YBE, dynamical quantum Yang-Baxter equation, has been studied [13, 14], which is closely related to dynamical quantum groups [15]. Their set-theoretic solution, called DYB maps, were proposed by Shibukawa in [34] and received a lot of attention (see for example [24, 28, 35, 37]). Dynamical braces and dynamical cycle sets were introduced and related to right non-degenerate unitary DYB maps [27, 32].

The Hom-Yang-Baxter equation (HYBE shortly) was introduced by Yau [40] motivated by Hom-Lie algebras, which is related to Hom-quantum groups [42]. Many researchers have devoted considerable attention to HYBE (see for example [9, 23, 29, 38, 41, 43]). However, to our knowledge, no work concentrates on set-theoretical solutions to GYBE.

The aim of this paper is to investigate left non-degenerate involutive set-theoretic solutions to HYBE, corresponding algebraic systems, called Hom-cycle sets, and their relationship.

The paper is organized as follows. In Section 2, we review some basic definitions and results, and provide a general categorical framework for the following discussion. In Section 3, we characterize left non-degenerate involutive set-theoretic solutions to HYBE. In Section 4, we introduce the notion of a Hom-cycle set, and prove that there exists a one to one correspondence between left Hom-cycle sets and left non-degenerate involutive set-theoretic solutions to HYBE. Section 5 is devoted to relationship among Hom-cycle sets, cycle sets, left non-degenerate involutive solutions to HYBE and YBE.

2. Preliminaries

Let XX be a nonempty set and let r:X×XX×Xr:X\times X\rightarrow X\times X be a map. We will write r(x,y)=(λx(y),ρy(x))r(x,y)=(\lambda_{x}(y),\rho_{y}(x)), where λx\lambda_{x} and ρy\rho_{y} are maps from XX to itself for all x,yXx,y\in X. The pair (X,r)(X,r) is referred to a quadratic set in [17].

A quadratic set (X,r)(X,r) (or a map rr) is called

  1. (1)

    left (respectively, right) non-degenerate if the map λx\lambda_{x} (respectively, ρx\rho_{x}) is bijective for all xXx\in X;

  2. (2)

    non-degenerate if rr is both left and right non-degenerate;

  3. (3)

    involutive if r2=idr^{2}=\operatorname{id}, the identify map;

  4. (4)

    a set-theoretic solution to YBE if rr satisfies (1.1).

The following lemma comes from [16, Proposition 1.6] (see also [19, Lemma 2.4]).

Lemma 2.1.
  1. (1)

    A quadratic set (X,r)(X,r) is involutive if and only if

    λλx(y)ρy(x)=x,\displaystyle\lambda_{\lambda_{x}(y)}\rho_{y}(x)=x, (2.1)
    ρρx(y)λy(x)=x,\displaystyle\rho_{\rho_{x}(y)}\lambda_{y}(x)=x, (2.2)

    for all x,yXx,y\in X.

  2. (2)

    A quadratic set (X,r)(X,r) is left non-degenerate and involutive if and only if λx\lambda_{x} is bijective for all xXx\in X and

    ρy(x)=λλx(y)1(x),\rho_{y}(x)=\lambda_{\lambda_{x}(y)}^{-1}(x), (2.3)

    for all x,yXx,y\in X.

Theorem 2.2.

[16, Proposition 1.6] A quadratic set (X,r)(X,r) is a set-theoretic solution to YBE if and only if

  1. (1)

    λλx(y)λρy(x)(z)=λxλy(z)\lambda_{\lambda_{x}(y)}\lambda_{\rho_{y}(x)}(z)=\lambda_{x}\lambda_{y}(z),

  2. (2)

    ρλρy(x)(z)λx(y)=λρλy(z)(x)ρz(y)\rho_{\lambda_{\rho_{y}(x)}(z)}\lambda_{x}(y)=\lambda_{\rho_{\lambda_{y}(z)}(x)}\rho_{z}(y),

  3. (3)

    ρzρy(x)=ρρz(y)ρλy(z)(x)\rho_{z}\rho_{y}(x)=\rho_{\rho_{z}(y)}\rho_{\lambda_{y}(z)}(x),

for all x,y,zXx,y,z\in X.

The following Theorem comes from [8, Proposition 2] (see also [22, Theorem 9.3.10]).

Theorem 2.3.

A quadratic set (X,r)(X,r) is a left non-degenerate involutive set-theoretic solution to YBE if and only if the following hold.

  1. (1)

    λx\lambda_{x} is bijective for all xXx\in X;

  2. (2)

    r2=idr^{2}=\operatorname{id};

  3. (3)

    λxλλx1(y)=λyλλy1(x)\lambda_{x}\lambda_{\lambda_{x}^{-1}(y)}=\lambda_{y}\lambda_{\lambda_{y}^{-1}(x)} for all x,yXx,y\in X.

Let (X,r)(X,r) and (X,r)(X^{\prime},r^{\prime}) be quadratic sets. By a morphism from (X,r)(X,r) to (X,r)(X^{\prime},r^{\prime}) we mean a map f:XXf:X\rightarrow X^{\prime} satisfying (f×f)r=r(f×f)(f\times f)r=r^{\prime}(f\times f).

Lemma 2.4.

Given two quadratic sets (X,r)(X,r) and (X,r)(X^{\prime},r^{\prime}), and a map f:XXf:X\rightarrow X^{\prime}, the following are equivalent:

  1. (1)

    ff is a morphism of quadratic sets;

  2. (2)

    fλx=λf(x)fandfρx=ρf(x)ffor allxXf\lambda_{x}=\lambda_{f(x)}^{\prime}f~{}\text{and}~{}f\rho_{x}=\rho_{f(x)}^{\prime}f~{}\text{for all}~{}x\in X.

If rr and rr^{\prime} are both left non-degenerate and involutive, then both conditions above are equivalent to one of the following conditions:

  1. (3)

    fλx=λf(x)ff\lambda_{x}=\lambda_{f(x)}^{\prime}f for all xXx\in X;

  2. (4)

    fλx1=λf(x)1ff\lambda_{x}^{-1}=\lambda_{f(x)}^{\prime-1}f for all xXx\in X.

By a Hom-quadratic set we mean a triple (X,r,α)(X,r,\alpha) of a nonempty set XX with two maps r:X×XX×Xr:X\times X\to X\times X and α:XX\alpha:X\to X such that r(α×α)=(α×α)rr(\alpha\times\alpha)=(\alpha\times\alpha)r. Thus a Hom-quadratic set is exactly a quadratic set with an endomorphism.

We will identify the quadratic set (X,r)(X,r) with the Hom-quadratic set (X,r,id)(X,r,\operatorname{id}).

Given two Hom-quadratic sets (X,r,α)(X,r,\alpha) and (X,r,α)(X^{\prime},r^{\prime},\alpha^{\prime}), a map f:XXf:X\rightarrow X^{\prime} is called a morphism of Hom-quadratic sets if

(f×f)r=r(f×f)andfα=αf.(f\times f)r=r^{\prime}(f\times f)~{}\text{and}~{}f\alpha=\alpha^{\prime}f.

Thus a morphism of Hom-quadratic sets is exactly a morphism of quadratic sets satisfying fα=αff\alpha=\alpha^{\prime}f.

Corollary 2.5.

Given a quadratic set (X,r)(X,r) and a map α:XX\alpha:X\to X, the triple (X,r,α)(X,r,\alpha) is a Hom-quadratic set if and only if

αλx=λα(x)αandαρx=ρα(x)αfor allxX.\alpha\lambda_{x}=\lambda_{\alpha(x)}\alpha~{}\text{and}~{}\alpha\rho_{x}=\rho_{\alpha(x)}\alpha~{}\text{for all}~{}x\in X.

A Hom-quadratic set (X,r,α)(X,r,\alpha) is called left non-degenerate, non-degenerate, and involutive, respectively, if rr has the same properties.

Denote by 𝖰𝖲\mathsf{QS} and 𝖧𝖰𝖲\mathsf{HQS} the categories of left non-degenerate involutive quadratic sets and left non-degenerate involutive Hom-quadratic sets, respectively. Then 𝖰𝖲\mathsf{QS} is a full subcategory of 𝖧𝖰𝖲\mathsf{HQS} by identifying a quadratic set (X,r)(X,r) with a Hom-quadratic set (X,r,id)(X,r,\operatorname{id}).

By a groupoid we mean a set with a binary operation. For a groupoid XX, denote by σx\sigma_{x} the left multiplication map by xXx\in X defined by

σx:XX,yxy.\sigma_{x}:X\to X,~{}~{}y\mapsto xy.

By a left quasigroup we mean a groupoid XX such that the left multiplication maps σx\sigma_{x} are bijective for all xXx\in X (see [33, Page 9]).

It should be pointed out that the image of an endomorphism of a left quasigroup is a left quasigroup, though the image of a homomorphism from a left quasigroup to a groupoid need not to be a left quasigroup (see [33, Page 15] and [33, Corollary 1.298]).

By a left Hom-quasigroup we mean a pair (X,α)(X,\alpha) of a left quasigroup XX with an endomorphism α\alpha.

We also write a left Hom-quasigroup (X,α)(X,\alpha) as (X,,α)(X,\cdot,\alpha) to indicate the operation \cdot of left quasigroup XX.

We can identify a left quasigroup XX with the left Hom-quasigroup (X,id)(X,\operatorname{id}).

Let (X,α)(X,\alpha) and (X,α)(X^{\prime},\alpha^{\prime}) be two left Hom-quasigroups. A map f:XXf:X\to X^{\prime} is called a morphism of left Hom-quasigroups if αf=fα\alpha f=f\alpha^{\prime} and f(xy)=f(x)f(y)f(xy)=f(x)f(y) for all x,yXx,y\in X.

From Lemma 2.4, we have the following lemma.

Lemma 2.6.

Let (X,r,α)(X,r,\alpha) and (X,r,α)(X^{\prime},r^{\prime},\alpha^{\prime}) be left non-degenerate involutive Hom-quadratic sets, and let (X,,α)(X,\cdot,\alpha) and (X,,α)(X^{\prime},*,\alpha^{\prime}) be left Hom-quasigroups such that xy=λx1(y)x\cdot y=\lambda_{x}^{-1}(y) for all x,yXx,y\in X and xy=λx1(y)x^{\prime}*y^{\prime}=\lambda^{\prime-1}_{x^{\prime}}(y^{\prime}) for all x,yXx^{\prime},y^{\prime}\in X^{\prime}. Then a map f:XXf:X\rightarrow X^{\prime} is a morphism of Hom-quadratic sets if and only if it is a morphism of left Hom-quasigroups.

Denote by 𝖧𝖰𝖦\mathsf{HQG} and 𝖰𝖦\mathsf{QG} the categories of left quasigroups and left Hom-quasigroups, respectively. Then 𝖧𝖰𝖦\mathsf{HQG} is a full subcategory of 𝖰𝖦\mathsf{QG} by identifying a left quasigroup XX with a left Hom-quasigroup (X,id)(X,\operatorname{id}).

Given a left non-degenerate involutive Hom-quadratic set (X,r,α)(X,r,\alpha), we get a left Hom-quasigroups (X,,α)(X,\cdot,\alpha), denoted by G(X,r,α)G(X,r,\alpha), with the operation defined by xy=λx1(y)x\cdot y=\lambda_{x}^{-1}(y) for all x,yXx,y\in X. Then we have a functor G:𝖧𝖰𝖲𝖧𝖰𝖦G:\mathsf{HQS}\to\mathsf{HQG} by associating (X,r,α)(X,r,\alpha) with G(X,r,α)G(X,r,\alpha), and a morphism ff in 𝖧𝖰𝖲\mathsf{HQS} with ff.

Conversely, given a left Hom-quasigroup (X,,α)(X,\cdot,\alpha), we get a Hom-quadratic set (X,r,α)(X,r,\alpha), denoted by S(X,,α)S(X,\cdot,\alpha), with λx(y)=σx1(y)\lambda_{x}(y)=\sigma_{x}^{-1}(y) and ρy(x)=σx1(y)x\rho_{y}(x)=\sigma_{x}^{-1}(y)\cdot x for all x,yXx,y\in X. It is routine to verify that (X,r,α)(X,r,\alpha) is left non-degenerate and involutive. Then we have a functor S:𝖧𝖰𝖦𝖧𝖰𝖲S:\mathsf{HQG}\to\mathsf{HQS} by associating (X,,α)(X,\cdot,\alpha) with S(X,,α)S(X,\cdot,\alpha) and a morphism ff in 𝖧𝖰𝖦\mathsf{HQG} with ff.

Theorem 2.7.

The functors GG and SS are mutually inverse, and so the categories 𝖧𝖰𝖲\mathsf{HQS} and 𝖧𝖰𝖦\mathsf{HQG} are isomorphic.

Proof.

It is straightforward. ∎

The functor GG induces a functor from 𝖰𝖲\mathsf{QS} to 𝖰𝖦\mathsf{QG} and SS induces a functor from 𝖰𝖦\mathsf{QG} to 𝖰𝖲\mathsf{QS}. We still denote the induced functors by GG and SS, respectively. By Theorem 2.7, we have the following corollary.

Corollary 2.8.

The functors G:𝖰𝖲𝖰𝖦G:\mathsf{QS}\to\mathsf{QG} and S:𝖰𝖦𝖰𝖲S:\mathsf{QG}\to\mathsf{QS} are mutually inverse, and so the categories 𝖰𝖲\mathsf{QS} and 𝖰𝖦\mathsf{QG} are isomorphic.

Denote by 𝖲𝗒𝖻𝖾\mathsf{S_{ybe}} the category of left non-degenerate involutive solutions to YBE. Then 𝖲𝗒𝖻𝖾\mathsf{S_{ybe}} is a full subcategory of 𝖰𝖲\mathsf{QS}.

A left quasigroup XX is called a left cycle set if (xy)(xz)=(yx)(yz)(xy)(xz)=(yx)(yz) for all x,y,zXx,y,z\in X [30].

Denote by 𝖢𝖲\mathsf{CS} the category of left cycle sets. Then 𝖢𝖲\mathsf{CS} is a full subcategory of 𝖰𝖦\mathsf{QG}.

By [30, Proposition 1], the functor GG induces a functor from 𝖲𝗒𝖻𝖾\mathsf{S_{ybe}} to 𝖢𝖲\mathsf{CS} and SS induces a functor from 𝖢𝖲\mathsf{CS} to 𝖲𝗒𝖻𝖾\mathsf{S_{ybe}}. We still denote the induced functors by GG and SS, respectively. By Corollary  2.8[30, Proposition 1] can be restated as follows.

Theorem 2.9.

The functors G:𝖲𝗒𝖻𝖾𝖢𝖲G:\mathsf{S_{ybe}}\to\mathsf{CS} and S:𝖢𝖲𝖲𝗒𝖻𝖾S:\mathsf{CS}\to\mathsf{S_{ybe}} are mutually inverse, and so the categories 𝖲𝗒𝖻𝖾\mathsf{S_{ybe}} and 𝖢𝖲\mathsf{CS} are isomorphic.

A groupoid is called Δ\Delta-bijective if the map Δ\Delta is bijective, where Δ:X×XX×X,(x,y)(xy,yx)\Delta:X\times X\to X\times X,~{}~{}(x,y)\mapsto(xy,yx) [2].

The following lemma comes from [2, Lemma 2.10] (see also Lemma 1.28 in Chapter XIII of [11]).

Lemma 2.10.

A groupoid XX is Δ\Delta-bijective if and only if there exists an operation \circ on XX (called the dual operation) such that

(xy)(yx)=x,\displaystyle(x\cdot y)\circ(y\cdot x)=x, (2.4)
(xy)(yx)=x,\displaystyle(x\circ y)\cdot(y\circ x)=x, (2.5)

for all x,yXx,y\in X. Furthermore, if conditions hold, then the operation \circ is unique and the inverse of Δ\Delta is given by (x,y)(xy,yx)(x,y)\mapsto(x\circ y,y\circ x).

Lemma 2.11.

(see [2, Lemma 2.11]) If a groupoid XX is Δ\Delta-bijective, then the square map q:XX,xx2q:X\to X,~{}~{}x\mapsto x^{2} is invertible.

We will say that a groupoid with extra structure is non-degenerate if the underlying groupoid is Δ\Delta-bijective.

Denote by 𝗇𝖽𝖢𝖲\mathsf{ndCS} the category of non-degenerate left cycle sets. Then 𝗇𝖽𝖢𝖲\mathsf{ndCS} is a full subcategory of 𝖰𝖦\mathsf{QG}.

Denote by 𝗇𝖽𝗂𝖲𝗒𝖻𝖾\mathsf{ndiS_{ybe}} the category of non-degenerate involutive solution to YBE. Then 𝗇𝖽𝗂𝖲𝗒𝖻𝖾\mathsf{ndiS_{ybe}} is a full subcategory of 𝖰𝖲\mathsf{QS}.

By Corollary 2.8 and Theorem 2.9, Proposition 2 in [30] can be restated as follows.

Theorem 2.12.

The functors G:𝗇𝖽𝗂𝖲𝗒𝖻𝖾𝗇𝖽𝖢𝖲G:\mathsf{ndiS_{ybe}}\to\mathsf{ndCS} and S:𝗇𝖽𝖢𝖲𝗇𝖽𝗂𝖲𝗒𝖻𝖾S:\mathsf{ndCS}\to\mathsf{ndiS_{ybe}} are mutually inverse, and so the categories 𝗇𝖽𝗂𝖲𝗒𝖻𝖾\mathsf{ndiS_{ybe}} and 𝗇𝖽𝖢𝖲\mathsf{ndCS} are isomorphic.

3. Set-theoretic solutions to the Hom-Yang-Baxter equation

The Hom-Yang-Baxter equation was proposed by Yau [40] motivated by Hom-Lie algebras.

Definition 3.1.

Given a vector space VV and two linear maps R:VVVVR:V\otimes V\rightarrow V\otimes V and α:VV\alpha:V\rightarrow V, the triple (V,R,α)(V,R,\alpha) is called a solution to the Hom-Yang-Baxter equation, if

  1. (1)

    R(αα)=(αα)RR(\alpha\otimes\alpha)=(\alpha\otimes\alpha)R, and

  2. (2)

    (αR)(Rα)(αR)=(Rα)(αR)(Rα)(\alpha\otimes R)(R\otimes\alpha)(\alpha\otimes R)=(R\otimes\alpha)(\alpha\otimes R)(R\otimes\alpha).

By analogy with set-theoretic solutions to YBE, we introduce set-theoretic solutions to HYBE.

Definition 3.2.

Given a nonempty set XX and two maps r:X×XX×Xr:X\times X\rightarrow X\times X and α:XX\alpha:X\rightarrow X, the triple (X,r,α)(X,r,\alpha) is called a set-theoretic solution to HYBE, if

  1. (1)

    r(α×α)=(α×α)rr\circ(\alpha\times\alpha)=(\alpha\times\alpha)\circ r, and

  2. (2)

    (α×r)(r×α)(α×r)=(r×α)(α×r)(r×α)(\alpha\times r)(r\times\alpha)(\alpha\times r)=(r\times\alpha)(\alpha\times r)(r\times\alpha).

Clearly, (X,r)(X,r) is a set-theoretic solution to YBE if and only if (X,r,id)(X,r,\operatorname{id}) is a set-theoretic solution to HYBE.

Let f:XXf:X\to X be a map and YY a subset of XX. Denote by f|Yf|_{Y} the restriction of ff to YY. When there is no ambiguity, we will write ff for the restriction f|Yf|_{Y}.

By analogy with the relation between set-theoretic solutions to YBE and solutions to YBE, we have the following theorem, and the proof is immediate.

Theorem 3.3.

Let VV be a vector space with a basis XX.

  1. (1)

    If (V,R,α)(V,R,\alpha) is a solution to HYBE such that R(XX)XXR(X\otimes X)\subseteq X\otimes X and α(X)X\alpha(X)\subseteq X, then (X,R|XX,α|X)(X,R|_{X\otimes X},\alpha|_{X}) is a set-theoretic solution to HYBE.

  2. (2)

    Conversely, if (X,r,α)(X,r,\alpha) is a set-theoretic solution to HYBE, and R:VVR:V\rightarrow V and α¯:VV\bar{\alpha}:V\to V are linear extensions of rr and α\alpha, respectively, then (V,R,α¯)(V,R,\bar{\alpha}) is a solution to HYBE.

In what follows, a set-theoretic solution is simply called a solution.

Lemma 3.4.

A triple (X,r,α)(X,r,\alpha) with r:X×XX×Xr:X\times X\rightarrow X\times X and α:XX\alpha:X\rightarrow X is a solution to HYBE if and only if the following conditions hold for all x,y,zXx,y,z\in X,

  1. (1)

    αλx=λα(x)α,andαρx=ρα(x)α\alpha\lambda_{x}=\lambda_{\alpha(x)}\alpha,\text{and}~{}\alpha\rho_{x}=\rho_{\alpha(x)}\alpha;

  2. (2)

    αλα(x)λy=λαλx(y)λρy(x)α\alpha\lambda_{\alpha(x)}\lambda_{y}=\lambda_{\alpha\lambda_{x}(y)}\lambda_{\rho_{y}(x)}\alpha;

  3. (3)

    ρλρy(x)α(z)αλx(y)=λρλy(z)α(x)αρz(y)\rho_{\lambda_{\rho_{y}(x)}\alpha(z)}\alpha\lambda_{x}(y)=\lambda_{\rho_{\lambda_{y}(z)}\alpha(x)}\alpha\rho_{z}(y);

  4. (4)

    αρα(y)ρx=ραρy(x)ρλx(y)α\alpha\rho_{\alpha(y)}\rho_{x}=\rho_{\alpha\rho_{y}(x)}\rho_{\lambda_{x}(y)}\alpha.

Proof.

It is straightforward. ∎

Theorem 2.2 is a special case of Lemma 3.4.

Example 3.5.

A triple (X,idX×X,α)(X,\operatorname{id}_{X\times X},\alpha) is a solution to HYBE if and only if α2=α\alpha^{2}=\alpha. In this case, it is involutive, but neither left nor right non-degenerate.

Example 3.6.

Let (X,r,α)(X,r,\alpha) be Hom-quadratic set. If (X,r)(X,r) is a solution to YBE, then (X,(α×α)r,α)(X,(\alpha\times\alpha)r,\alpha) is a solution to HYBE, and the converse holds if additionally α\alpha is injective or surjective.

Example 3.7.

A triple (X,τ,α)(X,\tau,\alpha) with τ(x,y)=(y,x)\tau(x,y)=(y,x) and arbitrary map α:XX\alpha:X\to X is a non-degenerate involutive solution to HYBE, called a trivial solution.

Example 3.8.

A triple (X,r,α)(X,r,\alpha) with r(x,y)=(f(y),g(x))r(x,y)=(f(y),g(x)), where f,gf,g are maps from XX to itself, is a solution to HYBE if and only if α,f,g\alpha,f,g commute. Furthermore, the solution (X,r,α)(X,r,\alpha) is left non-degenerate and involutive if and only if ff is bijective, and g=f1g=f^{-1}.

We are now in a position to characterize left non-degenerate involutive solutions to HYBE.

Theorem 3.9.

A triple (X,r,α)(X,r,\alpha) with r:X×XX×Xr:X\times X\rightarrow X\times X and α:XX\alpha:X\rightarrow X is a left non-degenerate involutive solution to HYBE if and only if the following conditions hold for all x,yXx,y\in X,

  1. (1)

    λx\lambda_{x} is bijective;

  2. (2)

    ρy(x)=λλx(y)1(x)\rho_{y}(x)=\lambda_{\lambda_{x}(y)}^{-1}(x);

  3. (3)

    αλx=λα(x)α\alpha\lambda_{x}=\lambda_{\alpha(x)}\alpha;

  4. (4)

    αλα(x)λλx1(y)=λα(y)λλy1(x)α\alpha\lambda_{\alpha(x)}\lambda_{\lambda_{x}^{-1}(y)}=\lambda_{\alpha(y)}\lambda_{\lambda_{y}^{-1}(x)}\alpha;

  5. (5)

    αλxλλα(x)1(y)=λα(y)λλy1α(x)α\alpha\lambda_{x}\lambda_{\lambda_{\alpha(x)}^{-1}(y)}=\lambda_{\alpha(y)}\lambda_{\lambda_{y}^{-1}\alpha(x)}\alpha;

  6. (6)

    αλxλλα(x)1α(y)=λyλλα(y)1α(x)α\alpha\lambda_{x}\lambda_{\lambda_{\alpha(x)}^{-1}\alpha(y)}=\lambda_{y}\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}\alpha.

Proof.

(\Rightarrow) (1) and (2) follows from Lemma 2.1(2). (3) follows from Lemma 3.4(1).

To prove (4), replacing yy by λx1(y)\lambda_{x}^{-1}(y) in (2) and Lemma 3.4(2), we get

ρλx1(y)(x)=λy1(x)andαλα(x)λλx1(y)=λα(y)λρλx1(y)(x)α.\rho_{\lambda_{x}^{-1}(y)}(x)=\lambda_{y}^{-1}(x)~{}~{}\text{and}~{}~{}\alpha\lambda_{\alpha(x)}\lambda_{\lambda_{x}^{-1}(y)}=\lambda_{\alpha(y)}\lambda_{\rho_{\lambda_{x}^{-1}(y)}(x)}\alpha.

Then (4) follows.

To prove (5), using (2) we can write Lemma 3.4(3) as

λλαλx(y)λρy(x)α(z)1αλx(y)=λρλy(z)α(x)αλλy(z)1(y).\lambda_{\lambda_{\alpha\lambda_{x}(y)}\lambda_{\rho_{y}(x)}\alpha(z)}^{-1}\alpha\lambda_{x}(y)=\lambda_{\rho_{\lambda_{y}(z)}\alpha(x)}\alpha\lambda_{\lambda_{y}(z)}^{-1}(y).

It follows by Lemma 3.4(2) that

λαλα(x)λy(z)1αλx(y)=λρλy(z)α(x)αλλy(z)1(y),\lambda_{\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\alpha\lambda_{x}(y)=\lambda_{\rho_{\lambda_{y}(z)}\alpha(x)}\alpha\lambda_{\lambda_{y}(z)}^{-1}(y),

which implies that αλx(y)=λαλα(x)λy(z)λρλy(z)α(x)αλλy(z)1(y)\alpha\lambda_{x}(y)=\lambda_{\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}\lambda_{\rho_{\lambda_{y}(z)}\alpha(x)}\alpha\lambda_{\lambda_{y}(z)}^{-1}(y). Since λy\lambda_{y} is bijective and zz is arbitrary, we can replace λy(z)\lambda_{y}(z) by zz in the last equation to obtain αλx(y)=λαλα(x)(z)λρzα(x)αλz1(y)\alpha\lambda_{x}(y)=\lambda_{\alpha\lambda_{\alpha(x)}(z)}\lambda_{\rho_{z}\alpha(x)}\alpha\lambda_{z}^{-1}(y). Thus

αλx=λαλα(x)(z)λρzα(x)αλz1,\alpha\lambda_{x}=\lambda_{\alpha\lambda_{\alpha(x)}(z)}\lambda_{\rho_{z}\alpha(x)}\alpha\lambda_{z}^{-1},

and so αλxλz=λαλα(x)(z)λρzα(x)α\alpha\lambda_{x}\lambda_{z}=\lambda_{\alpha\lambda_{\alpha(x)}(z)}\lambda_{\rho_{z}\alpha(x)}\alpha. By (2) we have

αλxλz=λαλα(x)(z)λλλα(x)(z)1α(x)α,\alpha\lambda_{x}\lambda_{z}=\lambda_{\alpha\lambda_{\alpha(x)}(z)}\lambda_{\lambda_{\lambda_{\alpha(x)}(z)}^{-1}\alpha(x)}\alpha,

in which replacing zz by λα(x)1(y)\lambda_{\alpha(x)}^{-1}(y) gives (5).

Now we prove (6). We first claim that the conditions (1) through (5) imply that

αλx=αλα2(x).\alpha\lambda_{x}=\alpha\lambda_{\alpha^{2}(x)}. (3.1)

Indeed, replacing xx by α(x)\alpha(x) in (4), we have

αλα2(x)λλα(x)1(y)=λα(y)λλy1α(x)α,\alpha\lambda_{\alpha^{2}(x)}\lambda_{\lambda_{\alpha(x)}^{-1}(y)}=\lambda_{\alpha(y)}\lambda_{\lambda_{y}^{-1}\alpha(x)}\alpha,

which together with (5) gives αλxλλα(x)1(y)=αλα2(x)λλα(x)1(y)\alpha\lambda_{x}\lambda_{\lambda_{\alpha(x)}^{-1}(y)}=\alpha\lambda_{\alpha^{2}(x)}\lambda_{\lambda_{\alpha(x)}^{-1}(y)}. By (1) we get αλx=αλα2(x)\alpha\lambda_{x}=\alpha\lambda_{\alpha^{2}(x)}.

Note that Lemma 3.4(2) implies that

λρx(z)α=λαλz(x)1αλα(z)λx.\lambda_{\rho_{x}(z)}\alpha=\lambda_{\alpha\lambda_{z}(x)}^{-1}\alpha\lambda_{\alpha(z)}\lambda_{x}. (3.2)

By (2), we can write 3.4(4) as

αλλρx(z)α(y)1ρx(z)=λλρλx(y)α(z)αρy(x)1ρλx(y)α(z),\displaystyle\alpha\lambda_{\lambda_{\rho_{x}(z)}\alpha(y)}^{-1}\rho_{x}(z)=\lambda_{\lambda_{\rho_{\lambda_{x}(y)}\alpha(z)}\alpha\rho_{y}(x)}^{-1}\rho_{\lambda_{x}(y)}\alpha(z),

for any zXz\in X. It follows by (2) that

αλλρx(z)α(y)1λλz(x)1(z)=λλρλx(y)α(z)αλλx(y)1(x)1λλα(z)λx(y)1α(z).\displaystyle\alpha\lambda_{\lambda_{\rho_{x}(z)}\alpha(y)}^{-1}\lambda_{\lambda_{z}(x)}^{-1}(z)=\lambda_{\lambda_{\rho_{\lambda_{x}(y)}\alpha(z)}\alpha\lambda_{\lambda_{x}(y)}^{-1}(x)}^{-1}\lambda_{\lambda_{\alpha(z)}\lambda_{x}(y)}^{-1}\alpha(z).

By using (3.2) and substituting λρx(z)α\lambda_{\rho_{x}(z)}\alpha and λρλx(y)α(z)α\lambda_{\rho_{\lambda_{x}(y)}\alpha(z)}\alpha into the last equation, we obtain

αλλαλz(x)1αλα(z)λx(y)1λλz(x)1(z)=λλαλα(z)λx(y)1αλα2(z)(x)1λλα(z)λx(y)1α(z).\alpha\lambda_{\lambda_{\alpha\lambda_{z}(x)}^{-1}\alpha\lambda_{\alpha(z)}\lambda_{x}(y)}^{-1}\lambda_{\lambda_{z}(x)}^{-1}(z)=\lambda_{\lambda_{\alpha\lambda_{\alpha(z)}\lambda_{x}(y)}^{-1}\alpha\lambda_{\alpha^{2}(z)}(x)}^{-1}\lambda_{\lambda_{\alpha(z)}\lambda_{x}(y)}^{-1}\alpha(z).

Thus by (3.1), we have

αλλαλz(x)1αλα(z)λx(y)1λλz(x)1(z)=λλαλα(z)λx(y)1αλz(x)1λλα(z)λx(y)1α(z).\alpha\lambda_{\lambda_{\alpha\lambda_{z}(x)}^{-1}\alpha\lambda_{\alpha(z)}\lambda_{x}(y)}^{-1}\lambda_{\lambda_{z}(x)}^{-1}(z)=\lambda_{\lambda_{\alpha\lambda_{\alpha(z)}\lambda_{x}(y)}^{-1}\alpha\lambda_{z}(x)}^{-1}\lambda_{\lambda_{\alpha(z)}\lambda_{x}(y)}^{-1}\alpha(z).

Replacing xx by λz1(x)\lambda_{z}^{-1}(x) in the previous equation, we have

αλλα(x)1αλα(z)λλz1(x)(y)1λx1(z)=λλαλα(z)λλz1(x)(y)1α(x)1λλα(z)λλz1(x)(y)1α(z).\alpha\lambda_{\lambda_{\alpha(x)}^{-1}\alpha\lambda_{\alpha(z)}\lambda_{\lambda_{z}^{-1}(x)}(y)}^{-1}\lambda_{x}^{-1}(z)=\lambda_{\lambda_{\alpha\lambda_{\alpha(z)}\lambda_{\lambda_{z}^{-1}(x)}(y)}^{-1}\alpha(x)}^{-1}\lambda_{\lambda_{\alpha(z)}\lambda_{\lambda_{z}^{-1}(x)}(y)}^{-1}\alpha(z).

Noting that λα(z)λλz1(x)1(y)\lambda_{\alpha(z)}\lambda_{\lambda_{z}^{-1}(x)}^{-1}(y) is an arbitrary element of XX, we may simply denote it by yy. Then the last equation can be written as

αλλα(x)1α(y)1λx1(z)=λλα(y)1α(x)1λy1α(z),\alpha\lambda_{\lambda_{\alpha(x)}^{-1}\alpha(y)}^{-1}\lambda_{x}^{-1}(z)=\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}^{-1}\lambda_{y}^{-1}\alpha(z),

that is, αλλα(x)1α(y)1λx1=λλα(y)1α(x)1λy1α\alpha\lambda_{\lambda_{\alpha(x)}^{-1}\alpha(y)}^{-1}\lambda_{x}^{-1}=\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}^{-1}\lambda_{y}^{-1}\alpha, which implies

αλxλλα(x)1α(y)=λyλλα(y)1α(x)α.\alpha\lambda_{x}\lambda_{\lambda_{\alpha(x)}^{-1}\alpha(y)}=\lambda_{y}\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}\alpha.

This proves (6).

(\Leftarrow) The nondegeneracy and involutivity of (X,r,α)(X,r,\alpha) follows from (1) and (2) by Lemma 2.1(2). We now prove that (X,r,α)(X,r,\alpha) satisfies the four conditions in Lemma 3.4.

Lemma 3.4(1) follows from Lemma 2.4.

By (2) and (4), we have

λαλx(y)λρy(x)α=λαλx(y)λλλx(y)1(x)α=αλα(x)λλx1λx(y)=αλα(x)λy,\lambda_{\alpha\lambda_{x}(y)}\lambda_{\rho_{y}(x)}\alpha=\lambda_{\alpha\lambda_{x}(y)}\lambda_{\lambda_{\lambda_{x}(y)}^{-1}(x)}\alpha=\alpha\lambda_{\alpha(x)}\lambda_{\lambda_{x}^{-1}\lambda_{x}(y)}=\alpha\lambda_{\alpha(x)}\lambda_{y},

which proves Lemma 3.4(2).

To prove Lemma 3.4(3), replacing xx by α(x)\alpha(x) and yy by λy(z)\lambda_{y}(z) in Lemma 3.4(2) and using (3.1) we get

λαλα(x)λy(z)λρλy(z)α(x)α=αλα2(x)λλy(z)=αλxλλy(z).\lambda_{\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}\lambda_{\rho_{\lambda_{y}(z)}\alpha(x)}\alpha=\alpha\lambda_{\alpha^{2}(x)}\lambda_{\lambda_{y}(z)}=\alpha\lambda_{x}\lambda_{\lambda_{y}(z)}.

It follows that

λαλα(x)λy(z)1αλx=λρλy(z)α(x)αλλy(z)1.\lambda_{\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\alpha\lambda_{x}=\lambda_{\rho_{\lambda_{y}(z)}\alpha(x)}\alpha\lambda_{\lambda_{y}(z)}^{-1}. (3.3)

Thus by Lemma 3.4(2), (3.3) and  (2), we have

ρλρy(x)α(z)αλx(y)\displaystyle\rho_{\lambda_{\rho_{y}(x)}\alpha(z)}\alpha\lambda_{x}(y) =λλαλx(y)λρy(x)α(z)1αλx(y)\displaystyle=\lambda_{\lambda_{\alpha\lambda_{x}(y)}\lambda_{\rho_{y}(x)}\alpha(z)}^{-1}\alpha\lambda_{x}(y)
=λαλα(x)λy(z)1αλx(y)\displaystyle=\lambda_{\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\alpha\lambda_{x}(y)
=λρλy(z)α(x)αλλy(z)1(y)\displaystyle=\lambda_{\rho_{\lambda_{y}(z)}\alpha(x)}\alpha\lambda_{\lambda_{y}(z)}^{-1}(y)
=λρλy(z)(α(x))αρz(y),\displaystyle=\lambda_{\rho_{\lambda_{y}(z)}(\alpha(x))}\alpha\rho_{z}(y),

which proves Lemma 3.4(3).

We now prove Lemma 3.4(4). By Lemma 3.4(2), we have

λρy(x)α\displaystyle\lambda_{\rho_{y}(x)}\alpha =λαλx(y)1αλα(x)λy.\displaystyle=\lambda_{\alpha\lambda_{x}(y)}^{-1}\alpha\lambda_{\alpha(x)}\lambda_{y}. (3.4)

And by (3.3), we have

λρλy(z)α(x)α\displaystyle\lambda_{\rho_{\lambda_{y}(z)}\alpha(x)}\alpha =λαλα(x)λy(z)1αλxλλy(z).\displaystyle=\lambda_{\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\alpha\lambda_{x}\lambda_{\lambda_{y}(z)}. (3.5)

Then using (2) and (3.4) we obtain that

αρα(z)ρy(x)\displaystyle\alpha\rho_{\alpha(z)}\rho_{y}(x) =αλλρy(x)α(z)1ρy(x)\displaystyle=\alpha\lambda_{\lambda_{\rho_{y}(x)}\alpha(z)}^{-1}\rho_{y}(x)
=αλλαλx(y)1αλα(x)λy(z)1λλx(y)1(x).\displaystyle=\alpha\lambda_{\lambda_{\alpha\lambda_{x}(y)}^{-1}\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\lambda_{\lambda_{x}(y)}^{-1}(x). (3.6)

By (2) and (3.5) we get

ραρz(y)ρλy(z)α(x)\displaystyle\rho_{\alpha\rho_{z}(y)}\rho_{\lambda_{y}(z)}\alpha(x) =λλρλy(z)α(x)αλλy(z)1(y)1ρλy(z)α(x)\displaystyle=\lambda_{\lambda_{\rho_{\lambda_{y}(z)}\alpha(x)}\alpha\lambda_{\lambda_{y}(z)}^{-1}(y)}^{-1}\rho_{\lambda_{y}(z)}\alpha(x)
=λλαλα(x)λy(z)1αλxλλy(z)λλy(z)1(y)1ρλy(z)α(x)\displaystyle=\lambda_{\lambda_{\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\alpha\lambda_{x}\lambda_{\lambda_{y}(z)}\lambda_{\lambda_{y}(z)}^{-1}(y)}^{-1}\rho_{\lambda_{y}(z)}\alpha(x)
=λλαλα(x)λy(z)1αλx(y)1λλα(x)λy(z)1α(x).\displaystyle=\lambda_{\lambda_{\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\alpha\lambda_{x}(y)}^{-1}\lambda_{\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\alpha(x). (3.7)

Substituting yy for α(y)\alpha(y) in (5), we have αλxλλα(x)1α(y)=λα2(y)λλα(y)1α(x)α\alpha\lambda_{x}\lambda_{\lambda_{\alpha(x)}^{-1}\alpha(y)}=\lambda_{\alpha^{2}(y)}\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}\alpha, which together with (6) yields

λα2(y)λλα(y)1α(x)α=λyλλα(y)1α(x)α.\lambda_{\alpha^{2}(y)}\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}\alpha=\lambda_{y}\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}\alpha. (3.8)

Replacing xx by αλx(y)\alpha\lambda_{x}(y) and yy by αλα(x)λy(z)\alpha\lambda_{\alpha(x)}\lambda_{y}(z) in (4), respectively, we have

αλα2λx(y)λλαλx(y)1αλα(x)λy(z)=λα2λα(x)λy(z)λλαλα(x)λy(z)1αλx(y)α.\alpha\lambda_{\alpha^{2}\lambda_{x}(y)}\lambda_{\lambda_{\alpha\lambda_{x}(y)}^{-1}\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}=\lambda_{\alpha^{2}\lambda_{\alpha(x)}\lambda_{y}(z)}\lambda_{\lambda_{\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\alpha\lambda_{x}(y)}\alpha.

By (3.1) and (3.8), we get that

αλλx(y)λλαλx(y)1αλα(x)λy(z)=λλα(x)λy(z)λλαλα(x)λy(z)1αλx(y)α,\alpha\lambda_{\lambda_{x}(y)}\lambda_{\lambda_{\alpha\lambda_{x}(y)}^{-1}\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}=\lambda_{\lambda_{\alpha(x)}\lambda_{y}(z)}\lambda_{\lambda_{\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\alpha\lambda_{x}(y)}\alpha,

whence

αλλαλx(y)1αλα(x)λy(z)1λλx(y)1=λλαλα(x)λy(z)1αλx(y)1λλα(x)λy(z)1α.\alpha\lambda_{\lambda_{\alpha\lambda_{x}(y)}^{-1}\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\lambda_{\lambda_{x}(y)}^{-1}=\lambda_{\lambda_{\alpha\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\alpha\lambda_{x}(y)}^{-1}\lambda_{\lambda_{\alpha(x)}\lambda_{y}(z)}^{-1}\alpha. (3.9)

Lemma 3.4(4) follows from (3.6), (3.7) and (3.9). ∎

Corollary 3.10.

Let (X,r,α)(X,r,\alpha) be a left non-degenerate involutive solution to HYBE. Then

  1. (1)

    λxα=λα2(x)α\lambda_{x}\alpha=\lambda_{\alpha^{2}(x)}\alpha for all xXx\in X;

  2. (2)

    λxα2=α2λx\lambda_{x}\alpha^{2}=\alpha^{2}\lambda_{x} for all xXx\in X.

Proof.

(1) Replacing yy by α(y)\alpha(y) in Theorem 3.9(5) we have

αλxλλα(x)1α(y)=λα2(y)λλα(y)1α(x)α,\alpha\lambda_{x}\lambda_{\lambda_{\alpha(x)}^{-1}\alpha(y)}=\lambda_{\alpha^{2}(y)}\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}\alpha,

which together with Theorem 3.9(6) yields

λα2(y)λλα(y)1α(x)α=λyλλα(y)1α(x)α.\lambda_{\alpha^{2}(y)}\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}\alpha=\lambda_{y}\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}\alpha.

By Theorem 3.9(3), λα2(y)αλλy1(x)=λyαλλy1(x)\lambda_{\alpha^{2}(y)}\alpha\lambda_{\lambda_{y}^{-1}(x)}=\lambda_{y}\alpha\lambda_{\lambda_{y}^{-1}(x)}. Thus λα2(y)α=λyα\lambda_{\alpha^{2}(y)}\alpha\ =\lambda_{y}\alpha.

(2) By Theorem 3.9(3) and Corollary 3.10(1) we have

αλα(x)=λα2(x)α=λxα.\alpha\lambda_{\alpha(x)}=\lambda_{\alpha^{2}(x)}\alpha=\lambda_{x}\alpha.

By Theorem 3.9(3), we have α2λx=αλα(x)α=λxα2\alpha^{2}\lambda_{x}=\alpha\lambda_{\alpha(x)}\alpha=\lambda_{x}\alpha^{2}. ∎

Theorem 3.11.

A triple (X,r,α)(X,r,\alpha) with r:X×XX×Xr:X\times X\rightarrow X\times X and α:XX\alpha:X\rightarrow X is a left non-degenerate involutive solution to HYBE if and only if the following statements are true for all x,yXx,y\in X,

  1. (1)

    λx\lambda_{x}\ is bijective;

  2. (2)

    ρy(x)=λλx(y)1(x)\rho_{y}(x)=\lambda_{\lambda_{x}(y)}^{-1}(x);

  3. (3)

    αλx=λα(x)α\alpha\lambda_{x}=\lambda_{\alpha(x)}\alpha;

  4. (4)

    λxα=αλα(x)\lambda_{x}\alpha=\alpha\lambda_{\alpha(x)};

  5. (5)

    αλxλλα(x)1(y)=λα(y)λλy1α(x)α\alpha\lambda_{x}\lambda_{\lambda_{\alpha(x)}^{-1}(y)}=\lambda_{\alpha(y)}\lambda_{\lambda_{y}^{-1}\alpha(x)}\alpha.

Proof.

(\Rightarrow) By Theorem 3.9 we need only prove (4). By (3) and Corollary 3.10, αλα(x)=λα2(x)α=λxα\alpha\lambda_{\alpha(x)}=\lambda_{\alpha^{2}(x)}\alpha=\lambda_{x}\alpha, as desired.

(\Leftarrow) It suffices to prove (4) and (6) of Theorem 3.9. Replacing yy by α(y)\alpha(y) in (5) and using (3) and (4) we get

αλxλλα(x)1α(y)=λα2(y)λλα(y)1α(x)α=λyλλα(y)1α(x)α,\alpha\lambda_{x}\lambda_{\lambda_{\alpha(x)}^{-1}\alpha(y)}=\lambda_{\alpha^{2}(y)}\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}\alpha=\lambda_{y}\lambda_{\lambda_{\alpha(y)}^{-1}\alpha(x)}\alpha,

which is Theorem 3.9(6).

Replacing xx by α(x)\alpha(x) in (5) we get αλα(x)λλα2(x)1(y)=λα(y)λλy1α2(x)α\alpha\lambda_{\alpha(x)}\lambda_{\lambda_{\alpha^{2}(x)}^{-1}(y)}=\lambda_{\alpha(y)}\lambda_{\lambda_{y}^{-1}\alpha^{2}(x)}\alpha. It follows from (3) and (4) that

λα(y)λλy1α2(x)α=αλα(x)λλα2(x)1(y)=λα2(x)λλα3(x)1α(y)α=λα2(x)λλα(x)1α(y)α=αλα(x)λλx1(y).\lambda_{\alpha(y)}\lambda_{\lambda_{y}^{-1}\alpha^{2}(x)}\alpha=\alpha\lambda_{\alpha(x)}\lambda_{\lambda_{\alpha^{2}(x)}^{-1}(y)}=\lambda_{\alpha^{2}(x)}\lambda_{\lambda_{\alpha^{3}(x)}^{-1}\alpha(y)}\alpha\\ =\lambda_{\alpha^{2}(x)}\lambda_{\lambda_{\alpha(x)}^{-1}\alpha(y)}\alpha=\alpha\lambda_{\alpha(x)}\lambda_{\lambda_{x}^{-1}(y)}.

This proves Theorem 3.9(4). ∎

When α=id\alpha=\operatorname{id}, Theorem 3.11 gives Theorem 2.3.

Example 3.12.

A triple (X,r,α)(X,r,\alpha) such that α(X)={θ}\alpha(X)=\{\theta\} is a left non-degenerate involutive solution to HYBE if and only if λx\lambda_{x} is bijective, λx(θ)=θ\lambda_{x}(\theta)=\theta and ρy(x)=λλx(y)1(x)\rho_{y}(x)=\lambda_{\lambda_{x}(y)}^{-1}(x) for all x,yXx,y\in X.

4. Hom-cycle sets

In this section, we introduce a Hom-version of cycle sets and relate them to left non-degenerate involutive solutions to HYBE.

Definition 4.1.

A left Hom-quasigroup (X,α)(X,\alpha) is called a left Hom-cycle set if the following conditions are satisfied for all x,y,zXx,y,z\in X,

α((xy)(α(x)z))=(yx)(α(y)α(z)),\displaystyle\alpha((xy)(\alpha(x)z))=(yx)(\alpha(y)\alpha(z)), (4.1)
α((α(x)y)(xz))=(yα(x))(α(y)α(z)),\displaystyle\alpha((\alpha(x)y)(xz))=(y\alpha(x))(\alpha(y)\alpha(z)), (4.2)
α((α(x)α(y))(xz))=(α(y)α(x))(yα(z)).\displaystyle\alpha((\alpha(x)\alpha(y))(xz))=(\alpha(y)\alpha(x))(y\alpha(z)). (4.3)

In what follows, left cycle sets and left Hom-cycle sets are referred to cycle sets and Hom-cycle sets.

Clearly, XX is a cycle set if and only if (X,idX)(X,\operatorname{id}_{X}) is a Hom-cycle set. Hence we will identify the cycle set XX with the Hom-cycle set (X,idX)(X,\operatorname{id}_{X}).

Denote by 𝖧𝖢𝖲\mathsf{HCS} the category of Hom-cycle sets, which is a full subcategory of 𝖧𝖰𝖦\mathsf{HQG}.

Denote by 𝖲𝗁𝗒𝖻𝖾\mathsf{S_{hybe}} the category of left non-degenerate involutive solution to HYBE, which is a full subcategory of 𝖧𝖰𝖲\mathsf{HQS}.

Theorem 4.2.

The functors G:𝖲𝗁𝗒𝖻𝖾𝖧𝖢𝖲G:\mathsf{S_{hybe}}\to\mathsf{HCS} and S:𝖧𝖢𝖲𝖲𝗁𝗒𝖻𝖾S:\mathsf{HCS}\to\mathsf{S_{hybe}} are mutually inverse, and so the categories 𝖲𝗁𝗒𝖻𝖾\mathsf{S_{hybe}} and 𝖧𝖢𝖲\mathsf{HCS} are isomorphic.

Proof.

Let (X,r,α)(X,r,\alpha) be a left non-degenerate involutive solution to HYBE. G(X,r,α)G(X,r,\alpha) is a left Hom-quasigroup. Furthermore, since xy=λx1(y)x\cdot y=\lambda_{x}^{-1}(y), Theorem 3.9(4)–(6) imply the conditions (4.1)–(4.3). Thus G(X,r,α)G(X,r,\alpha) is a Hom-cycle set.

Conversely, let (X,α)(X,\alpha) be a Hom-cycle set and S(X,α)=(X,r,α)S(X,\alpha)=(X,r,\alpha). Then (X,r,α)(X,r,\alpha) is a left non-degenerate involutive Hom-quadratic set. By Lemma 2.1(2) and Corollary 2.5, (X,r,α)(X,r,\alpha) satisfies Theorem 3.9(1)–(3). Furthermore, the conditions (4.1)–(4.3) imply Theorem 3.9(4)–(6). Thus S(X,α)S(X,\alpha) is a left non-degenerate involutive solution to HYBE. ∎

Theorem 4.2 generalizes [30, Proposition 1].

Theorem 4.3.

A left Hom-quasigroup (X,α)(X,\alpha) is a Hom-cycle set if and only if the following conditions hold for all x,yXx,y\in X,

α2(x)α(y)=xα(y),\displaystyle\alpha^{2}(x)\alpha(y)=x\alpha(y), (4.4)
(xα(y))(α(x)α(z))=(yα(x))(α(y)α(z)).\displaystyle(x\alpha(y))(\alpha(x)\alpha(z))=(y\alpha(x))(\alpha(y)\alpha(z)). (4.5)
Proof.

Suppose S(X,α)=(X,r,α)S(X,\alpha)=(X,r,\alpha).

(\Rightarrow) By Theorem 4.2, (X,r,α)(X,r,\alpha) is a left non-degenerate involutive solution to HYBE. Thus (4.4) follows from Theorem 3.11(4), and (4.5) follows from (4.2).

(\Leftarrow) By Theorem 4.2 it suffices to prove that (X,r,α)(X,r,\alpha) is a left non-degenerate involutive solution to HYBE. We need to verify that (X,r,α)(X,r,\alpha) satisfies (1)–(5) of Theorem 3.11. In fact, Lemma 2.1(2) and Corollary 2.5 imply that (X,r,α)(X,r,\alpha) satisfies (1)–(3) in Theorem 3.11, and the conditions (4.4) and (4.5) imply that (X,r,α)(X,r,\alpha) satisfies (4)–(5) in Theorem 3.11, as desired. ∎

Example 4.4.

Let XX be a nonempty set with a map α:XX\alpha:X\to X, and let f:XXf:X\to X be a bijection such that fα=αff\alpha=\alpha f. Define an operation on XX by xy=f(y)xy=f(y) for all x,yXx,y\in X. Then (X,α)(X,\alpha) is a Hom-cycle set, which corresponds to the left non-degenerate involutive solution to HYBE defined in Example 3.8.

Example 4.5.

Let (X,α)(X,\alpha) be a left Hom-quasigroup with α(X)={θ}\alpha(X)=\{\theta\}. Then (X,α)(X,\alpha) is a Hom-cycle set if and only if θ\theta is a right zero, i.e., xθ=θx\theta=\theta for all xXx\in X. In this case, (X,α)(X,\alpha) corresponds to the solution defined in Example 3.12.

Example 4.6.

A trivial solution to HYBE corresponds to a right zero groupoid with an endomorphism.

Example 4.7.

Let (X,+)(X,+) be an Abelian group with endomorphisms φ,ψ,α\varphi,\psi,\alpha and define xy=φ(x)+ψ(y)x\cdot y=\varphi(x)+\psi(y) for all x,yXx,y\in X. Then (X,,α)(X,\cdot,\alpha) is a Hom-cycle set if and only if the following hold:

  1. (1)

    ψ\psi is bijective;

  2. (2)

    φα=αφ\varphi\alpha=\alpha\varphi and ψα=αψ\psi\alpha=\alpha\psi;

  3. (3)

    φα2=φ\varphi\alpha^{2}=\varphi;

  4. (4)

    φ2=φψαψφα\varphi^{2}=\varphi\psi\alpha-\psi\varphi\alpha.

In fact, (1) is equivalent to saying that (X,)(X,\cdot) is a left quasigroup by [2, Section 4.1]; (2) is equivalent to the assertion that α\alpha is an endomorphism of (X,)(X,\cdot); (3) and (4) are equivalent to (4.4) and (4.5), respectively.

Lemma 4.8.

If (X,α)(X,\alpha) is a left Hom-quasigroup satisfying (4.4), then

(xα2(y))α(z)=(xy)α(z),(x\alpha^{2}(y))\alpha(z)=(xy)\alpha(z), (4.6)

for all x,y,zXx,y,z\in X.

Proof.

By using (4.4) we have

(xy)α(z)=α2(xy)α(z)=(α2(x)α2(y))α(z)=(xα2(y))α(z),(xy)\alpha(z)=\alpha^{2}(xy)\alpha(z)=(\alpha^{2}(x)\alpha^{2}(y))\alpha(z)=(x\alpha^{2}(y))\alpha(z),

for all x,y,zXx,y,z\in X. ∎

Lemma 4.9.

If (X,α)(X,\alpha) is a left Hom-quasigroup satisfying (4.4), then (4.1), (4.2) and (4.3) are equivalent.

Proof.

Replacing xx by α(x)\alpha(x) in (4.1) gives (4.2).

Replacing yy by α(y)\alpha(y) in (4.2) gives (4.3).

Suppose (4.3) holds. Replacing xx by α(x)\alpha(x) and yy by α(y)\alpha(y) in (4.3) gives α((α2(x)α2(y))(α(x)z))=(α2(y)α2(x))(α(y)α(z))\alpha((\alpha^{2}(x)\alpha^{2}(y))(\alpha(x)z))=(\alpha^{2}(y)\alpha^{2}(x))(\alpha(y)\alpha(z)). It follows that

α((xy)(α(x)z))=(yα2(x))(α(y)α(z)),\alpha((xy)(\alpha(x)z))=(y\alpha^{2}(x))(\alpha(y)\alpha(z)),

which together with (4.6) gives (4.1). ∎

Theorem 4.10.

Let (X,α)(X,\alpha) be a left Hom-quasigroup. Then (X,α)(X,\alpha) is a Hom-cycle set if and only if it satisfies (4.4) and one of (4.1), (4.2) and (4.3).

Proof.

It follows from Lemma 4.9 and Theorem 4.3. ∎

Theorem 4.11.

A Hom-cycle set is non-degenerate if and only if the corresponding solution to HYBE is non-degenerate.

Proof.

Let (X,α)(X,\alpha) be a Hom-cycle set and (X,r,α)(X,r,\alpha) the corresponding solution to HYBE.

(\Rightarrow) Suppose (X,α)(X,\alpha) is non-degenerate and \circ is the dual operation defined in Lemma 2.10. Then (2.5) can be rewritten as λxy1(yx)=x\lambda_{x\circ y}^{-1}(y\circ x)=x for all x,yXx,y\in X. Thus yx=λxy(x)y\circ x=\lambda_{x\circ y}(x) for all x,yXx,y\in X. It follows by interchanging xx and yy that xy=λyx(y)x\circ y=\lambda_{y\circ x}(y) for all x,yXx,y\in X. Substituting the last equation into (2.5), we get

λyx(y)(yx)=x.\lambda_{y\circ x}(y)\cdot\newline (y\circ x)=x.

Denote by τy\tau_{y} the left multiplication by yy with respect to the operation \circ. Then ρyτy(x)=ρy(yx)=λyx(y)(yx)=x\rho_{y}\tau_{y}(x)=\rho_{y}(y\circ x)=\lambda_{y\circ x}(y)\cdot(y\circ x)=x. Noting that xλx(y)=yx\lambda_{x}(y)=y, by (2.4) we have

τyρy(x)=(xλx(y))(λx(y)x)=x.\tau_{y}\rho_{y}(x)=(x\lambda_{x}(y))\circ(\lambda_{x}(y)x)=x.

Thus ρy\rho_{y} is bijective, and so (X,r,α)(X,r,\alpha) is non-degenerate.

(\Leftarrow) Suppose (X,r,α)(X,r,\alpha) is non-degenerate. Define xy=ρx1(y)x\circ y=\rho_{x}^{-1}(y) for all x,yXx,y\in X. Replacing yy by λx1(y)\lambda_{x}^{-1}(y) in (2.1) we obtain ρλx1(y)(x)=λy1(x)\rho_{\lambda_{x}^{-1}(y)}(x)=\lambda_{y}^{-1}(x), whence ρλx1(y)1λy1(x)=x\rho_{\lambda_{x}^{-1}(y)}^{-1}\lambda_{y}^{-1}(x)=x, which gives (2.4). Similarly, replacing yy by ρx1(y)\rho_{x}^{-1}(y) in (2.2) we can get (2.5). Thus (X,α)(X,\alpha) is non-degenerate by Lemma 2.10. ∎

Denote by 𝗇𝖽𝖧𝖢𝖲\mathsf{ndHCS} the category of non-degenerate left Hom-cycle sets. Then 𝗇𝖽𝖧𝖢𝖲\mathsf{ndHCS} is a full subcategory of 𝖧𝖢𝖲\mathsf{HCS}.

Denote by 𝗇𝖽𝗂𝖲𝗁𝗒𝖻𝖾\mathsf{ndiS_{hybe}} the category of non-degenerate involutive solutions to HYBE. Then 𝗇𝖽𝗂𝖲𝗁𝗒𝖻𝖾\mathsf{ndiS_{hybe}} is a full subcategory of 𝖲𝗁𝗒𝖻𝖾\mathsf{S_{hybe}}.

Theorem 4.2 and Theorem 4.11 have the following corollary, which generalize [30, Proposition 2].

Corollary 4.12.

The functors G:𝗇𝖽𝗂𝖲𝗁𝗒𝖻𝖾𝗇𝖽𝖧𝖢𝖲G:\mathsf{ndiS_{hybe}}\to\mathsf{ndHCS} and S:𝗇𝖽𝖧𝖢𝖲𝗇𝖽𝗂𝖲𝗁𝗒𝖻𝖾S:\mathsf{ndHCS}\to\mathsf{ndiS_{hybe}} are mutually inverse, and so the categories 𝗇𝖽𝗂𝖲𝗁𝗒𝖻𝖾\mathsf{ndiS_{hybe}} and 𝗇𝖽𝖧𝖢𝖲\mathsf{ndHCS} are isomorphic.

Theorem 4.13.

Let (X,α)(X,\alpha) be a non-degenerate Hom-cycle set with the dual operation \circ. Then (X,,α)(X,\circ,\alpha) is a non-degenerate Hom-cycle set.

Proof.

Suppose S(X,α)=(X,r,α)S(X,\alpha)=(X,r,\alpha). By Theorem 4.11, (X,r,α)(X,r,\alpha) is a non-degenerate involutive solution to HYBE. Replacing yy by λx(y)\lambda_{x}(y) in (2.4), we have (xλx(y))(λx(y)x)=x(x\lambda_{x}(y))\circ(\lambda_{x}(y)x)=x, whence yρy(x)=xy\circ\rho_{y}(x)=x. Replacing xx by ρy1(x)\rho_{y}^{-1}(x) in the last equation gives yx=ρy1(x)y\circ x=\rho_{y}^{-1}(x), and so ρy(yx)=x\rho_{y}(y\circ x)=x. Thus (X,)(X,\circ) is a left quasigroup, and we have

α(x)=αρy(yx)=ρα(y)α(yx),\alpha(x)=\alpha\rho_{y}(y\circ x)=\rho_{\alpha(y)}\alpha(y\circ x),

whence α(yx)=ρα(y)1α(x)=α(y)α(x)\alpha(y\circ x)=\rho_{\alpha(y)}^{-1}\alpha(x)=\alpha(y)\circ\alpha(x). Hence (X,,α)(X,\circ,\alpha) is a left Hom-quasigroup. Suppose S(X,,α)=(X,r,α)S(X,\circ,\alpha)=(X,r^{\circ},\alpha). By Corollary 4.12, it suffices to prove that (X,r,α)(X,r^{\circ},\alpha) is a non-degenerate involutive solution to HYBE. Since λy1(x)=yx=ρy1(x)\lambda_{y}^{\circ-1}(x)=y\circ x=\rho_{y}^{-1}(x), we have λy=ρy\lambda_{y}^{\circ}=\rho_{y}. Replacing yy by ρx(y)\rho_{x}(y) in (2.5), we have (xρx(y))(ρx(y)x)=x(x\circ\rho_{x}(y))(\rho_{x}(y)\circ x)=x. Thus

y(ρx(y)x)=x=yλy(x),y(\rho_{x}(y)\circ x)=x=y\lambda_{y}(x),

and so we have λy(x)=ρx(y)x=λx(y)x=ρy(x)\lambda_{y}(x)=\rho_{x}(y)\circ x=\lambda_{x}^{\circ}(y)\circ x=\rho_{y}^{\circ}(x), whence ρy=λy\rho_{y}^{\circ}=\lambda_{y}. Thus r(x,y)=(ρx(y),λy(x))r^{\circ}(x,y)=(\rho_{x}(y),\lambda_{y}(x)). Clearly r=τrτr^{\circ}=\tau r\tau, where τ(x,y)=(y,x)\tau(x,y)=(y,x). Since (X,r,α)(X,r,\alpha) is a non-degenerate involutive solution to HYBE, so is (X,r,α)(X,r^{\circ},\alpha), as desired. ∎

Finite cycle sets and cycle sets with bijective square maps, especially square-free cycle sets, are non-degenerate [30], but Hom-cycle sets are not the case.

Example 4.14.

Let X={1,2,3,4}X=\{1,2,3,4\} with a map α:XX\alpha:X\rightarrow X such that α(X)={1}\alpha(X)=\{1\}. Define an operation on XX by the following multiplication table:

123411234212343143241324\begin{array}[]{c|cccc}\cdot&1&2&3&4\\ \hline\cr 1&1&2&3&4\\ 2&1&2&3&4\\ 3&1&4&3&2\\ 4&1&3&2&4\end{array}

Then (X,,α)(X,\cdot,\alpha) is a square-free Hom-cycle set, but not non-degenerate since Δ(3,4)=(2,2)=Δ(2,2)\Delta(3,4)=(2,2)=\Delta(2,2).

5. Twists

Let (X,α)(X,\alpha) be a left Hom-quasigroup. Define an operation on XX by xy=α(x)yx\cdot^{\prime}y=\alpha(x)y for all x,yXx,y\in X. Then (X,)(X,\cdot^{\prime}) is a left quasigroup, and

α(xy)=α(α(x)y)=α2(x)α(y)=α(x)α(y).\alpha(x\cdot^{\prime}y)=\alpha\left(\alpha(x)y\right)=\alpha^{2}(x)\alpha(y)=\alpha(x)\cdot^{\prime}\alpha(y).

Thus α\alpha is an endomorphism of (X,)(X,\cdot^{\prime}). Hence (X,,α)(X,\cdot^{\prime},\alpha) is a left Hom-quasigroup. We call (X,,α)(X,\cdot^{\prime},\alpha) the twist of (X,α)(X,\alpha), denoted by T(X,α)T(X,\alpha).

Using twist we can define a functor T:𝖧𝖰𝖦𝖧𝖰𝖦T:\mathsf{HQG}\to\mathsf{HQG} in a natural way.

A left Hom-quasigroup (X,α)(X,\alpha) is called an im-cycle set if the following are satisfied:

α3(x)α(y)=α(x)α(y),\displaystyle\alpha^{3}(x)\alpha(y)=\alpha(x)\alpha(y), (5.1)
(α(x)α(y))(α(x)α(z))=(α(y)α(x))(α(y)α(z)),\displaystyle(\alpha(x)\alpha(y))(\alpha(x)\alpha(z))=(\alpha(y)\alpha(x))(\alpha(y)\alpha(z)), (5.2)

for all x,y,zXx,y,z\in X.

It is clear that (X,α)(X,\alpha) is an im-cycle set if and only if α(X)\alpha(X) is a cycle set and (5.1) holds.

Theorem 5.1.

A left Hom-quasigroup is an im-cycle set if and only if its twist is a Hom-cycle set.

Proof.

Let (X,,α)(X,\cdot,\alpha) be a left Hom-quasigroup. Then its twist (X,,α)(X,\cdot^{\prime},\alpha) is also a left Hom-quasigroup.

(\Rightarrow) By (5.1) we have α2(x)α(y)=xα(y)\alpha^{2}(x)\cdot^{\prime}\alpha(y)=x\cdot^{\prime}\alpha(y) for all x,yXx,y\in X. Replacing xx by α(x)\alpha(x) and yy by α(y)\alpha(y) in (5.2), we obtain

(α2(x)α2(y))(α2(x)α(z))=(α2(y)α2(x))(α2(y)α(z)),(\alpha^{2}(x)\alpha^{2}(y))(\alpha^{2}(x)\alpha(z))=(\alpha^{2}(y)\alpha^{2}(x))(\alpha^{2}(y)\alpha(z)),

which implies that (xα(y))(α(x)α(z))=(yα(x))(α(y)α(z))(x\cdot^{\prime}\alpha(y))\cdot^{\prime}(\alpha(x)\cdot^{\prime}\alpha(z))=(y\cdot^{\prime}\alpha(x))\cdot^{\prime}(\alpha(y)\cdot^{\prime}\alpha(z)). By Theorem 4.3, (X,,α)(X,\cdot^{\prime},\alpha) is a Hom-cycle set.

(\Leftarrow) Since (X,,α)(X,\cdot^{\prime},\alpha) is a Hom-cycle set, applying Theorem 4.3 to (X,,α)(X,\cdot^{\prime},\alpha) yields (5.1) and

α(α(x)α(y))(α2(x)α(z))=α(α(y)α(x))(α2(y)α(z)).\alpha(\alpha(x)\alpha(y))(\alpha^{2}(x)\alpha(z))=\alpha(\alpha(y)\alpha(x))(\alpha^{2}(y)\alpha(z)).

Replacing xx by α(x)\alpha(x) and yy by α(y)\alpha(y) in the last equation, we have

(α3(x)α3(y))(α3(x)α(z))=(α3(y)α3(x))(α3(y)α(z)).(\alpha^{3}(x)\alpha^{3}(y))(\alpha^{3}(x)\alpha(z))=(\alpha^{3}(y)\alpha^{3}(x))(\alpha^{3}(y)\alpha(z)).

By (5.1), we see that (α(X),)(\alpha(X),\cdot) is a cycle set. Thus (X,α)(X,\alpha) is an im-cycle set. ∎

Corollary 5.2.

The twist of an im-cycle set is a Hom-cycle set.

Corollary 5.3.

Let XX be a cycle set with an endomorphism α\alpha such that (5.1) holds. Then the twist of (X,α)(X,\alpha) is a Hom-cycle set.

Theorem 5.4.

The twist of a Hom-cycle set is an im-cycle set.

Proof.

Let (X,,α)(X,\cdot,\alpha) be a Hom-cycle set with the twist (X,,α)(X,\cdot^{\prime},\alpha). By (4.4), we can rewrite (4.5) as

α2(xα(y))(α(x)α(z))=α2(yα(x))(α(y)α(z)),\alpha^{2}(x\alpha(y))(\alpha(x)\alpha(z))=\alpha^{2}(y\alpha(x))(\alpha(y)\alpha(z)),

that is,

(α2(x)α3(y))(α(x)α(z))=(α2(y)α3(x))(α(y)α(z)),(\alpha^{2}(x)\alpha^{3}(y))(\alpha(x)\alpha(z))=(\alpha^{2}(y)\alpha^{3}(x))(\alpha(y)\alpha(z)),

for all x,y,zXx,y,z\in X. Thus by (4.6), we have

(α2(x)α(y))(α(x)α(z))=(α2(y)α(x))(α(y)α(z)),(\alpha^{2}(x)\alpha(y))(\alpha(x)\alpha(z))=(\alpha^{2}(y)\alpha(x))(\alpha(y)\alpha(z)),

which implies (xy)(xα(z))=(yx)(yα(z))(x\cdot^{\prime}{y})\cdot^{\prime}(x\cdot^{\prime}\alpha(z))=(y\cdot^{\prime}x)\cdot^{\prime}(y\cdot^{\prime}\alpha(z)) for all x,y,zXx,y,z\in X. Hence (α(X),)(\alpha(X),\cdot^{\prime}) is a cycle set. By (4.4), we have

α4(x)α(y)=α2(x)α(y)\alpha^{4}(x)\cdot\alpha(y)=\alpha^{2}(x)\cdot\alpha(y)

for all x,yXx,y\in X, whence α3(x)α(y)=α(x)α(y)\alpha^{3}(x)\cdot^{\prime}\alpha(y)=\alpha(x)\cdot^{\prime}\alpha(y) for all x,yXx,y\in X. Thus (X,,α)(X,\cdot^{\prime},\alpha) is an im-cycle set. ∎

Theorem 5.5.

If (X,α)(X,\alpha) be a non-degenerate Hom-cycle set, then its twist is non-degenerate.

Proof.

Let T(X,α)=(X,,α)T(X,\alpha)=(X,\cdot^{\prime},\alpha). By Theorem 4.13, (X,,α)(X,\circ,\alpha) is a Hom-cycle set. Applying Theorem 4.3 to (X,,α)(X,\circ,\alpha) we obtain

α2(x)α(y)=xα(y),\alpha^{2}(x)\circ\alpha(y)=x\circ\alpha(y), (5.3)

for all x,yXx,y\in X. Define xy=α(x)yx\circ^{\prime}y=\alpha(x)\circ y for all x,yXx,y\in X. Then by (4.4), (5.3) and Lemma 2.10 we have

(xy)(yx)=α(α(x)y)(α(y)x)=(xα(y))(α(y)x)=x,\displaystyle(x\cdot^{\prime}y)\circ^{\prime}(y\cdot^{\prime}x)=\alpha(\alpha(x)y)\circ(\alpha(y)x)=(x\alpha(y))\circ(\alpha(y)x)=x,
(xy)(yx)=α(α(x)y)(α(y)x)=(xα(y))(α(y)x)=x.\displaystyle(x\circ^{\prime}y)\cdot^{\prime}(y\circ^{\prime}x)=\alpha(\alpha(x)\circ y)(\alpha(y)\circ x)=(x\circ\alpha(y))(\alpha(y)\circ x)=x.

Thus (X,,α)(X,\cdot^{\prime},\alpha) is non-degenerate by Lemma 2.10. ∎

Lemma 5.6.

Let (X,α)(X,\alpha) be a left Hom-quasigroup with α(X)\alpha(X) singleton. Then its twist is a non-degenerate left Hom-quasigroup.

Proof.

Let α(X)={θ}\alpha(X)=\{\theta\}. Then θ\theta is a right zero of the left quasigroup XX. Let T(X,α)=(X,,α)T(X,\alpha)=(X,\cdot^{\prime},\alpha). Thus

Δ(x,y)=(xy,yx)=(α(x)y,α(y)x)=(θy,θx)\Delta(x,y)=(x\cdot^{\prime}y,y\cdot^{\prime}x)=(\alpha(x)y,\alpha(y)x)=(\theta y,\theta x)

for all x,yXx,y\in X. It follows Δ=σθ×σθ\Delta=\sigma_{\theta}\times\sigma_{\theta}. Since σθ\sigma_{\theta} is bijective, so is Δ\Delta. Hence (X,,α)(X,\cdot^{\prime},\alpha) is non-degenerate. ∎

Remark 5.7.

A Hom-cycle set is not necessarily the twist of an im-cycle set, and vice versa. For example, let (X,α)(X,\alpha) be as in Example 4.14. It is degenerate and both a Hom-cycle set and an im-cycle set, but it is isomorphic to neither the twist of a Hom-cycle set nor the twist of an im-cycle set by Lemma 5.6.

The following example shows that twist of a Hom-cycle set is not necessarily a Hom-cycle set.

Example 5.8.

Let FF be a field of characteristic 2\neq 2, XX the vector space F3{F}^{3} and φ,ψ,α\varphi,\psi,\alpha the linear endomorphisms of XX defined under the natural basis by the following matrices, respectively,

A=(010001000),B=(100011001),C=(101010001).A=\begin{pmatrix}0&1&0\\ 0&0&1\\ 0&0&0\end{pmatrix},~{}B=\begin{pmatrix}1&0&0\\ 0&1&1\\ 0&0&1\end{pmatrix},~{}C=\begin{pmatrix}-1&0&1\\ 0&-1&0\\ 0&0&-1\end{pmatrix}.

Then it is easy to check that ψ\psi is bijective and

φα=αφ,ψα=αψ,φα2=φ,φ2=φψψφ,φ2αφ2.\ \varphi\alpha=\alpha\varphi,~{}\psi\alpha=\alpha\psi,~{}\varphi\alpha^{2}=\varphi,~{}\varphi^{2}=\varphi\psi-\psi\varphi,~{}\varphi^{2}\alpha\neq\varphi^{2}.

Define xy=φ(x)+ψ(y)x\cdot y=\varphi(x)+\psi(y) for all x,yXx,y\in X. Then (X,)(X,\cdot) is a cycle set by [2, Section 4.1] and α\alpha is an endomorphism of (X,)(X,\cdot) satisfying α2(x)y=xy\alpha^{2}(x)y=xy for all x,yXx,y\in X.

The twist T(X,,α)T(X,\cdot,\alpha) of (X,,α)(X,\cdot,\alpha) is a Hom-cycle set By Corollary 5.3. Clearly, the twist of T(X,,α)T(X,\cdot,\alpha) equals (X,,α)(X,\cdot,\alpha), which is not a Hom-cycle set since (4) in Example 4.7 does not hold.

Corollary 5.9.

If (X,α)(X,\alpha) is a non-degenerate Hom-cycle set, then the map q:XX,xα(x)xq^{\prime}:X\to X,~{}~{}x\mapsto\alpha(x)x is bijective.

Proof.

By Theorem 5.5 the twist (X,,α)(X,\cdot^{\prime},\alpha) is non-degenerate. Since qq^{\prime} is the square map of (X,)(X,\cdot^{\prime}), qq^{\prime} is bijective by Lemma 2.11. ∎

Remark 5.10.

The converses of Theorem 5.5 and Corollary 5.9 do not hold. Example 4.14 provides a counterexample to the both cases.

We now apply Theorem 5.1, Corollary 5.3 and Theorem 5.4 to solutions to HYBE.

Let (X,r,α)(X,r,\alpha) be a left non-degenerate involutive Hom-quadratic set. We call STG(X,r,α)STG(X,r,\alpha) the twist of (X,r,α)(X,r,\alpha). The twist of (X,r,α)(X,r,\alpha) is also a left non-degenerate involutive Hom-quadratic set.

Theorem 5.11.

Let (X,r,α)(X,r,\alpha) be a left non-degenerate involutive Hom-quadratic set with the twist (X,r,α)(X,r^{\prime},\alpha). Then

r(x,y)=(λα(x)(y),λλα2(x)α(y)1(x))r^{\prime}(x,y)=(\lambda_{\alpha(x)}(y),\lambda_{\lambda_{\alpha^{2}(x)}\alpha(y)}^{-1}(x)) (5.4)

for all x,yXx,y\in X. If (X,r,α)(X,r,\alpha) is additionally a solution to HYBE, then

r(x,y)=(λα(x)(y),ρα(y)(x))r^{\prime}(x,y)=(\lambda_{\alpha(x)}(y),\rho_{\alpha(y)}(x)) (5.5)

for all x,yXx,y\in X.

Proof.

Since STG(X,r,α)=(X,r,α)STG(X,r,\alpha)=(X,r^{\prime},\alpha), by Theorem 2.7 we have

TG(X,r,α)=G(X,r,α).TG(X,r,\alpha)=G(X,r^{\prime},\alpha).

Thus G(X,r,α)G(X,r^{\prime},\alpha) is the twist of G(X,r,α)G(X,r,\alpha). Let G(X,r,α)=(X,,α)G(X,r,\alpha)=(X,\cdot,\alpha). Then G(X,r,α)=(X,,α)G(X,r^{\prime},\alpha)=(X,\cdot^{\prime},\alpha), where xy=α(x)yx\cdot^{\prime}y=\alpha(x)y for all x,yXx,y\in X. Thus λx=λα(x)\lambda^{\prime}_{x}=\lambda_{\alpha(x)}, and so

ρy(x)=λλx(y)1(x)=λαλα(x)(y)1(x)=λλα2(x)α(y)1(x).\rho^{\prime}_{y}(x)=\lambda^{\prime-1}_{\lambda^{\prime}_{x}(y)}(x)=\lambda^{-1}_{\alpha\lambda_{\alpha(x)}(y)}(x)=\lambda^{-1}_{\lambda_{\alpha^{2}(x)}\alpha(y)}(x).

Therefore (5.4) holds.

Furthermore, if (X,r,α)(X,r,\alpha) is a solution to HYBE, then (5.5) follows from (5.4), Corollary 3.10(1) and (2.3). ∎

Remark 5.12.

By Example 5.8 and Theorem 4.2, the twist of a left non-degenerate involutive solution to HYBE is not necessarily a solution to HYBE.

Lemma 5.13.

Let (X,r,α)(X,r,\alpha) be a left non-degenerate involutive Hom-quadratic set and i,ji,j be nonnegative integers such that 0ji20\leq j-i\leq 2. Then the following are equivalent.

  1. (1)

    r(αi+2×αj)=(1×α2)r(αi×αj)r(\alpha^{i+2}\times\alpha^{j})=(1\times\alpha^{2})r(\alpha^{i}\times\alpha^{j});

  2. (2)

    λαi+2(x)αj=λαi(x)αj\lambda_{\alpha^{i+2}(x)}\alpha^{j}=\lambda_{\alpha^{i}(x)}\alpha^{j} for all xXx\in X.

Proof.

We first note that for all x,yXx,y\in X,

r(αi+2×αj)(x,y)=(λαi+2(x)αj(y),ραj(y)αi+2(x)),\displaystyle r(\alpha^{i+2}\times\alpha^{j})(x,y)=(\lambda_{\alpha^{i+2}(x)}\alpha^{j}(y),\rho_{\alpha^{j}(y)}\alpha^{i+2}(x)),
(1×α2)r(αi×αj)(x,y)=(λαi(x)αj(y),α2ραj(y)αi(x)).\displaystyle(1\times\alpha^{2})r(\alpha^{i}\times\alpha^{j})(x,y)=(\lambda_{\alpha^{i}(x)}\alpha^{j}(y),\alpha^{2}\rho_{\alpha^{j}(y)}\alpha^{i}(x)).

It suffices to prove that (2) implies ραj(y)αi+2(x)=α2ραj(y)αi(x)\rho_{\alpha^{j}(y)}\alpha^{i+2}(x)=\alpha^{2}\rho_{\alpha^{j}(y)}\alpha^{i}(x) for all x,yXx,y\in X. Since 0ji20\leq j-i\leq 2, we have λαi(x)αj=αjλαi+2j(x)\lambda_{\alpha^{i}(x)}\alpha^{j}=\alpha^{j}\lambda_{\alpha^{i+2-j}(x)}, and so αjλαi+2j(x)1=λαi(x)1αj\alpha^{j}\lambda_{\alpha^{i+2-j}(x)}^{-1}=\lambda_{\alpha^{i}(x)}^{-1}\alpha^{j}. Thus λαi+2(x)1αj=λαi(x)1αj\lambda_{\alpha^{i+2}(x)}^{-1}\alpha^{j}=\lambda_{\alpha^{i}(x)}^{-1}\alpha^{j}. Consequently,

α2ραj(y)αi(x)=λλαi+2(x)αj+2(y)1αi+2(x)=λαi+2(λxαji(y))1αi+2(x)=λαi(λxαji(y))1αi+2(x)=λλαi(x)αj(y)1αi+2(x)=λλαi+2(x)αj(y)1αi+2(x)=ραj(y)αi+2(x),\alpha^{2}\rho_{\alpha^{j}(y)}\alpha^{i}(x)=\lambda_{\lambda_{\alpha^{i+2}(x)}\alpha^{j+2}(y)}^{-1}\alpha^{i+2}(x)=\lambda_{\alpha^{i+2}(\lambda_{x}\alpha^{j-i}(y))}^{-1}\alpha^{i+2}(x)\\ =\lambda_{\alpha^{i}(\lambda_{x}\alpha^{j-i}(y))}^{-1}\alpha^{i+2}(x)=\lambda_{\lambda_{\alpha^{i}(x)}\alpha^{j}(y)}^{-1}\alpha^{i+2}(x)\\ =\lambda_{\lambda_{\alpha^{i+2}(x)}\alpha^{j}(y)}^{-1}\alpha^{i+2}(x)=\rho_{\alpha^{j}(y)}\alpha^{i+2}(x),

as desired. ∎

Corollary 5.14.

Let (X,r,α)(X,r,\alpha) be a left non-degenerate involutive solution to HYBE. Then r(α2×α)=(id×α2)r(id×α)r(\alpha^{2}\times\alpha)=(\operatorname{id}\times\alpha^{2})r(\operatorname{id}\times\alpha).

Proof.

It follows from Corollary 3.10 and Lemma 5.13. ∎

Theorem 5.15.

Let (X,r,α)(X,r,\alpha) be left non-degenerate involutive Hom-quadratic set with the twist (X,r,α)(X,r^{\prime},\alpha).

  1. (1)

    If (X,r,α)(X,r,\alpha) is a solution to HYBE, then (α(X),r)(\alpha(X),r^{\prime}) is a solution to YBE and r(α2×id)=(id×α2)rr^{\prime}(\alpha^{2}\times\operatorname{id})=(\operatorname{id}\times\alpha^{2})r^{\prime} on α(X)\alpha(X).

  2. (2)

    The twist (X,r,α)(X,r^{\prime},\alpha) is a solution to HYBE if and only if (α(X),r)(\alpha(X),r) is a solution to YBE and r(α2×id)=(id×α2)rr(\alpha^{2}\times\operatorname{id})=(\operatorname{id}\times\alpha^{2})r on α(X)\alpha(X).

  3. (3)

    If (X,r)(X,r) is a solution to YBE and r(α3×α)=(α×α3)rr(\alpha^{3}\times\alpha)=(\alpha\times\alpha^{3})r, then (X,r,α)(X,r^{\prime},\alpha) is a solution to HYBE.

Proof.

Let S(X,r,α)=(X,,α)S(X,r,\alpha)=(X,\cdot,\alpha). Then S(X,r,α)=(X,,α)S(X,r^{\prime},\alpha)=(X,\cdot^{\prime},\alpha), the twist of (X,,α)(X,\cdot,\alpha).

(1) Theorem 4.2 implies that (X,,α)(X,\cdot,\alpha) is a Hom-cycle set. By Theorem 5.4, (α(X),)(\alpha(X),\cdot^{\prime}) is a cycle set and (5.1) holds with respect to the operation \cdot^{\prime}. It follows that α2(x)y=xy\alpha^{2}(x)\cdot^{\prime}y=x\cdot^{\prime}y for all x,yα(X)x,y\in\alpha(X), which implies r(α2×id)=(id×α2)rr^{\prime}(\alpha^{2}\times\operatorname{id})=(\operatorname{id}\times\alpha^{2})r^{\prime} on α(X)\alpha(X) by Lemma 5.13. By Theorem 2.9, (α(X),r)(\alpha(X),r^{\prime}) is a solution to YBE.

(2) By Theorem 4.2, (X,r,α)(X,r^{\prime},\alpha) is a left non-degenerate involutive solution to HYBE if and only if (X,,α)(X,\cdot^{\prime},\alpha) is a Hom-cycle set. Equivalently, by Theorem 5.1 (α(X),)(\alpha(X),\cdot) is a left cycle set satisfying (5.1). This is equivalent to that (α(X),r)(\alpha(X),r) is a solution to YBE satisfying r(α2×id)=(id×α2)rr(\alpha^{2}\times\operatorname{id})=(\operatorname{id}\times\alpha^{2})r on α(X)\alpha(X) by Theorem 2.9 and Lemma 5.13.

(3) follows from the sufficiency in (2). ∎

Acknowledgements

This work is supported by NSF of China (No.12171194, No.11971289).

References

  • Baxter [1972] R. J. Baxter. Partition function of the eight-vertex lattice model. Ann. Physics, 70(1):193–228, 1972. URL https://doi.org/10.1016/0003-4916(72)90335-1.
  • Bonatto et al. [2021] M. Bonatto, M. Kinyon, D. Stanovskỳ, and P. Vojtěchovskỳ. Involutive latin solutions of the Yang-Baxter equation. J. Algebra, 565:128–159, 2021. URL https://doi.org/10.1016/j.jalgebra.2020.09.001.
  • Castelli et al. [2018] M. Castelli, F. Catino, and G. Pinto. A new family of set-theoretic solutions of the Yang-Baxter equation. Comm. Algebra, 46(4):1622–1629, 2018. URL https://doi.org/10.1080/00927872.2017.1350700.
  • Castelli et al. [2019] M. Castelli, F. Catino, and G. Pinto. Indecomposable involutive set-theoretic solutions of the Yang-Baxter equation. J. Pure Appl. Algebra, 223(10):4477–4493, 2019. URL https://doi.org/10.1016/j.jpaa.2019.01.017.
  • Castelli et al. [2020a] M. Castelli, F. Catino, and G. Pinto. About a question of Gateva-Ivanova and Cameron on square-free set-theoretic solutions of the Yang-Baxter equation. Comm. Algebra, 48(6):2369–2381, 2020a. URL https://doi.org/10.1080/00927872.2020.1713328.
  • Castelli et al. [2020b] M. Castelli, G. Pinto, and W. Rump. On the indecomposable involutive set-theoretic solutions of the Yang-Baxter equation of prime-power size. Comm. Algebra, 48(5):1941–1955, 2020b. URL https://doi.org/10.1080/00927872.2019.1710163.
  • Castelli et al. [2021] M. Castelli, F. Catino, and P. Stefanelli. Indecomposable involutive set-theoretic solutions of the Yang-Baxter equation and orthogonal dynamical extensions of cycle sets. Mediterr. J. Math., 18(6):Paper No. 246, 27pp, 2021. URL https://doi.org/10.1007/s00009-021-01912-4.
  • Cedó et al. [2014] F. Cedó, E. Jespers, and J. Okniński. Braces and the Yang-Baxter equation. Comm. Math. Phys., 327(1):101–116, 2014. URL https://doi.org/10.1007/s00220-014-1935-y.
  • Chen and Zhang [2014] Y. Chen and L. Zhang. The category of Yetter-Drinfel’d Hom-modules and the quantum Hom-Yang-Baxter equation. J. Math. Phys., 55(3):031702, 18pp, 2014. URL https://doi.org/10.1063/1.4868964.
  • Dehornoy [2015] P. Dehornoy. Set-theoretic solutions of the Yang-Baxter equation, RC-calculus, and Garside germs. Adv. Math., 282:93–127, 2015. URL https://doi.org/10.1016/j.aim.2015.05.008.
  • Dehornoy et al. [2015] P. Dehornoy, F. Digne, E. Godelle, D. Krammer, and J. Michel. Foundations of Garside theory, volume 22 of EMS Tracts in Mathematics. European Mathematical Society (EMS), Zürich, 2015. URL https://doi.org/10.4171/139.
  • Drinfeld [1992] V. G. Drinfeld. On some unsolved problems in quantum group theory. In Quantum groups (Leningrad, 1990), volume 1510 of Lecture Notes in Math., pages 1–8. Springer, Berlin, 1992. URL https://doi.org/10.1007/BFb0101175.
  • Etingof and Latour [2005] P. Etingof and F. Latour. The dynamical Yang-Baxter equation, representation theory, and quantum integrable systems, volume 29 of Oxford Lecture Series in Mathematics and its Applications. Oxford University Press, Oxford, 2005.
  • Etingof and Schiffmann [2001] P. Etingof and O. Schiffmann. Lectures on the dynamical Yang-Baxter equations. In Quantum groups and Lie theory (Durham), 1999), volume 290 of London Math. Soc. Lecture Note Ser., pages 89–129. Cambridge Univ. Press, Cambridge, 2001.
  • Etingof and Varchenko [1998] P. Etingof and A. Varchenko. Solutions of the quantum dynamical Yang-Baxter equation and dynamical quantum groups. Comm. Math. Phys., 196(3):591–640, 1998. URL https://doi.org/10.1007/s002200050437.
  • Etingof et al. [1999] P. Etingof, T. Schedler, and A. Soloviev. Set-theoretical solutions to the quantum Yang-Baxter equation. Duke Math. J., 100(2):169–209, 1999. URL https://doi.org/10.1215/S0012-7094-99-10007-X.
  • Gateva-Ivanova [2004] T. Gateva-Ivanova. A combinatorial approach to the set-theoretic solutions of the Yang-Baxter equation. J. Math. Phys., 45(10):3828–3858, 2004. URL https://doi.org/10.1063/1.1788848.
  • Gateva-Ivanova [2018] T. Gateva-Ivanova. Set-theoretic solutions of the Yang-Baxter equation, braces and symmetric groups. Adv. Math., 338:649–701, 2018. URL https://doi.org/10.1016/j.aim.2018.09.005.
  • Gateva-Ivanova and Majid [2008] T. Gateva-Ivanova and S. Majid. Matched pairs approach to set theoretic solutions of the Yang-Baxter equation. J. Algebra, 319(4):1462–1529, 2008. URL https://doi.org/10.1016/j.jalgebra.2007.10.035.
  • Gateva-Ivanova and Van den Bergh [1998] T. Gateva-Ivanova and M. Van den Bergh. Semigroups of II-type. J. Algebra, 206(1):97–112, 1998. URL https://doi.org/10.1006/jabr.1997.7399.
  • Guarnieri and Vendramin [2017] L. Guarnieri and L. Vendramin. Skew braces and the Yang-Baxter equation. Math. Comp., 86(307):2519–2534, 2017. URL https://doi.org/10.1090/mcom/3161.
  • Jespers and Okniński [2007] E. Jespers and J. Okniński. Noetherian semigroup algebras, volume 7 of Algebra and Applications. Springer, Dordrecht, 2007. URL https://doi.org/10.1007/1-4020-5810-1.
  • Jiao and Huang [2018] Z. Jiao and G. Huang. Solutions of Hom-Yang-Baxter equation from monoidal Hom-(co)algebra structures. Math Notes, 104:121–134, 2018. URL https://doi.org/10.1134/S0001434618070131.
  • Kamiya and Shibukawa [2011] N. Kamiya and Y. Shibukawa. Dynamical Yang-Baxter maps associated with homogeneous pre-systems. J. Gen. Lie Theory Appl., 5:Art. ID G110106, 9pp, 2011. URL https://doi.org/10.4303/jglta/G110106.
  • Lebed and Vendramin [2017] V. Lebed and L. Vendramin. Homology of left non-degenerate set-theoretic solutions to the Yang-Baxter equation. Adv. Math., 304:1219–1261, 2017. URL https://doi.org/10.1016/j.aim.2016.09.024.
  • Lu et al. [2000] J.-H. Lu, M. Yan, and Y.-C. Zhu. On the set-theoretical Yang-Baxter equation. Duke Math. J., 104(1):1–18, 2000. URL https://doi.org/10.1215/S0012-7094-00-10411-5.
  • Matsumoto [2013] D. K. Matsumoto. Dynamical braces and dynamical Yang-Baxter maps. J. Pure Appl. Algebra, 217(2):195–206, 2013. URL https://doi.org/10.1016/j.jpaa.2012.06.012.
  • Matsumoto and Shimizu [2018] D. K. Matsumoto and K. Shimizu. Quiver-theoretical approach to dynamical Yang-Baxter maps. J. Algebra, 507:47–80, 2018. URL https://doi.org/10.1016/j.jalgebra.2018.04.003.
  • Panaite et al. [2019] F. Panaite, P. T. Schrader, and M. D. Staic. Hom-tensor categories and the Hom-Yang-Baxter equation. Appl. Categ. Structures, 27(4):323–363, 2019. URL https://doi.org/10.1007/s10485-019-09556-y.
  • Rump [2005] W. Rump. A decomposition theorem for square-free unitary solutions of the quantum Yang-Baxter equation. Adv. Math., 193(1):40–55, 2005. URL https://doi.org/10.1016/j.aim.2004.03.019.
  • Rump [2007] W. Rump. Braces, radical rings, and the quantum Yang-Baxter equation. J. Algebra, 307(1):153–170, 2007. URL https://doi.org/10.1016/j.jalgebra.2006.03.040.
  • Rump [2016] W. Rump. Dynamical groups and braces. J. Algebra Appl., 15(07):1650135, 31pp, 2016. URL https://doi.org/10.1142/S0219498816501358.
  • Shcherbacov [2017] V. Shcherbacov. Elements of quasigroup theory and applications. Monographs and Research Notes in Mathematics. CRC Press, Boca Raton, FL, 2017. URL https://doi.org/10.1201/9781315120058.
  • Shibukawa [2005] Y. Shibukawa. Dynamical Yang-Baxter maps. Int. Math. Res. Not., 2005(36):2199–2221, 2005. URL https://doi.org/10.1155/IMRN.2005.2199.
  • Shibukawa [2016] Y. Shibukawa. Hopf algebroids and rigid tensor categories associated with dynamical Yang-Baxter maps. J. Algebra, 449:408–445, 2016. URL https://doi.org/10.1016/j.jalgebra.2015.11.007.
  • Smoktunowicz [2018] A. Smoktunowicz. On Engel groups, nilpotent groups, rings, braces and the Yang-Baxter equation. Trans. Amer. Math. Soc., 370(9):6535–6564, 2018. URL https://doi.org/10.1090/tran/7179.
  • Veselov [2007] A. Veselov. Yang-Baxter maps: dynamical point of view. In Combinatorial aspect of integrable systems, volume 17 of MSJ Mem., pages 145–167. Math. Soc. Japan, Tokyo, 2007. URL https://doi.org/10.2969/msjmemoirs/01701C060.
  • Wang et al. [2022] S. Wang, X. Zhang, and S. Guo. Hom-Yang-Baxter equations and Hom-Yang-Baxter systems. Comm. Algebra, 2022. URL https://doi.org/10.1080/00927872.2022.2137518.
  • Yang [1967] C. N. Yang. Some exact results for the many-body problem in one dimension with repulsive delta-function interaction. Phys. Rev. Lett., 19(23):1312–1315, 1967. URL https://doi.org/10.1103/PhysRevLett.19.1312.
  • Yau [2009] D. Yau. The Hom-Yang-Baxter equation, Hom-Lie algebras, and quasi-triangular bialgebras. J. Phys. A, 42(16):165202, 12pp, 2009. URL https://doi.org/10.1088/1751-8113/42/16/165202.
  • Yau [2011] D. Yau. The Hom-Yang-Baxter equation and Hom-Lie algebras. J. Math. Phys., 52(5):053502, 19pp, 2011. URL https://doi.org/10.1063/1.3571970.
  • Yau [2012] D. Yau. Hom-quantum groups: I. Quasi-triangular Hom-bialgebras. J. Phys. A, 45(6):065203, 23pp, 2012. URL https://doi.org/10.1088/1751-8113/45/6/065203.
  • Yau [2015] D. Yau. The classical Hom-Yang-Baxter equation and Hom-Lie bialgebras. Int. Electron. J. Algebra, 17(17):11–45, 2015. URL https://doi.org/10.24330/ieja.266210.