Access the full text.

Sign up today, get DeepDyve free for 14 days.

Condensed Matter
, Volume 2018 (1810) – Sep 29, 2018

/lp/arxiv-cornell-university/polynomial-interpolation-as-detector-of-orbital-equation-equivalence-YqphAnMglI

- ISSN
- 0129-1831
- eISSN
- ARCH-3331
- DOI
- 10.1142/S0129183118500961
- Publisher site
- See Article on Publisher Site

October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence International Journal of Modern Physics C c World Scientiﬁc Publishing Company 1 2,3,4 Owen J. Brison and Jason A.C. Gallas Departamento de Matem´atica, Universidade de Lisboa, 1649-003 Lisboa, Portugal, Instituto de Altos Estudos da Para´ıba, Rua Silvino Lopes 419-2502, 58039-190 Jo˜ao Pessoa, Brazil, Complexity Sciences Center, 9225 Collins Ave. 1208, Surfside FL 33154, USA, Max-Planck-Institut fu¨r Physik komplexer Systeme, 01187 Dresden, Germany Received 8 August 2018 Accepted 24 August 2018 Published 19 September 2018 https://doi.org/10.1142/S0129183118500961 Equivalence between algebraic equations of motion may be detected by using a p-adic method, methods using factorization and linear algebra, or by systematic computer search of suitable Tschirnhausen transformations. Here, we show standard polynomial interpolation to be a competitive alternative method for detecting orbital equivalences and ﬁeld isomorphisms. Eﬃcient algorithms for ascertaining equivalences are relevant for signiﬁcantly minimizing computer searches in theoretical and practical applications. Keywords: Polynomial equivalence; Polynomial isomorphism; Algebraic computation. PACS Nos.: 02.70.Wz, 02.10.De, 03.65.Fd 1. Introduction Massive simulations in computer clusters have revealed that the control parameter space of dissipative dynamical systems is riddled with stability islands characterized individually by periodic motions of ever increasing periods which accumulate, ex- hibiting conspicuous and interesting regularities. Even in systems governed by sim- ple polynomial maps, the number of periodic orbits displays an explosive growth as a function of the nonlinearity . A few years ago it was realized that the nucleation of stability in classical systems occurs in a variety of ways which normally involve the presence of peak-doubling and peak-adding cascades extending over wide re- gions of the control space. For instance, transitions among distinct stable oscillatory phases may, or may not, be mediated by parameter intervals, windows, of chaotic 2,3,4 oscillations . Details of these and other novel regularities are summarized in re- cent surveys concerning complexities observed in the accumulations of doubling and 5 6 7,8 adding cascades in laser systems see Ref. , in chemistry , in biochemical models , and in the dynamics of cancer . arXiv:1810.02312v1 [nlin.CD] 29 Sep 2018 October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence 2 Owen J. Brison and Jason A.C. Gallas Eﬃcient algorithms for explicitly computing ﬁeld isomorphisms and their in- verses for dynamical systems governed by polynomial maps are central players for analytically assessing the aforementioned regularities present in both theoretical and practical applications. Polynomial maps are useful because metric properties of systems governed by sets of nonlinear diﬀerential equations cannot yet be analyti- cally determined due to the lack of methods to solve them exactly. Since stability windows quickly become very narrow as oscillations increase, accuracy is tanta- mount to exact analytical work. One added advantage of polynomial maps is that the equations of motion generated by them are always exact, not contaminated by the unavoidable round-oﬀ and discretization errors arising from numerical approx- imations of diﬀerential equations. Currently, the common methods used for determining ﬁeld isomorphism and equivalence among equations of motion are: i) a p-adic method reported by Zassen- 10,11 haus and Liang and used to study isomorphisms among quintic polynomials of all three signatures (n, ℓ), namely (1, 2), (3, 1) and (5, 0), where n refers to the number of real roots while ℓ refers to the number of pairs of complex roots; ii) methods using factorization and linear algebra ; iii) a systematic search of suitable Tschirnhausen transformations constrained by some quantity of interest, usually polynomial discriminants. This approach is eﬃcient for systems with low-degree equations of motion . The purpose of this paper is to introduce an alternative method to detect equiva- lence among orbital equations of motion, namely Lagrange polynomial interpolation. While standard methods look for isomorphisms focusing primarily on properties of the number ﬁelds involved, the method based on polynomial interpolation seeks isomorphisms directly among irreducible polynomials, because they are the objects that arise automatically as equations of motion governing periodic trajectories of 15,16,17 dynamical systems of algebraic origin . After obtaining sets of equations of motion, the most important task in dynamics is to determined whether or not such equations are interconnected (i.e. deﬁne or not the same number ﬁeld), and when they are, to obtain complete sets of explicit expressions for the transformations and corresponding inverses interconnecting them. 2. Polynomial interpolation as isomorphism detector Although already used in 1779 by Edward Waring and an easy consequence of a formula published in 1783 by Euler, “Lagrange” interpolation is a result rediscovered by Lagrange in 1795 which is nowadays traditionally used in numerical analysis for polynomial interpolation. The history of polynomial interpolation is however considerably older than the works above, as reviewed by Meijering . For a given set of k pairs (a , b ), i = 1,··· , k with all a distinct, the interpola- i i i tion polynomial is deﬁned as the lowest degree polynomial that assumes the value October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence Polynomial interpolation as detector of orbital equivalence 3 b at the point a for each i. The interpolation polynomial is a linear combination i i L(x) = b ℓ (x), i i i=0 where (x − a )··· (x − a )(x − a )··· (x − a ) x − a 1 i−1 i+1 k j ℓ (x) = = (a − a )··· (a − a )(a − a )··· (a − a ) a − a i 1 i i−1 i i+1 i k i j j6=i are the basis polynomials. The interpolator L(x) has degree k ≤ (k−1) and L(a ) = b ,··· , L(a ) = b . In all cases of interest here, not only are the a distinct but the 1 k k i b are too, so we may reverse the interpolations. Now, it is not diﬃcult to see that by taking b = a ,··· , b = a , b = a 1 2 k−1 k k 1 we obtain L(a ) = a ,··· , L(a ) = a or, in other words, for this choice of b the 1 2 k 1 i “action” of L(x) is to induce a (cyclic) permutation among the elements a ,··· , a . 1 k This is the basic observation that will be explored in the remainder of the paper as an eﬃcient detector of equivalence and isomorphism among polynomial equations of motion. In the applications considered here, the elements a and b are usually i i roots of algebraic orbital equations which have real coeﬃcients and, consequently, may be real or complex numbers. Depending on the numerical values of (a , b ), the i i transformations resulting from permutations of the b will have coeﬃcients deﬁned by real or complex numbers. Isomorphisms of interest to us here are the ones char- acterized by rational coeﬃcients. As seen in the examples in the next Section, in most cases, these interesting isomorphisms involve just integer coeﬃcients. 3. Applications 3.1. Equivalences of Vandermonde’s totally real cyclic quintic As a ﬁrst application, we apply polynomial interpolation to ﬁnd direct and in- verse transformations that establish the equivalence among a pair of cyclic quintics 4 19,14 of minimum discriminant Δ = 14641 = 11 originally considered by Cohn , namely 5 4 3 2 V (x) = x − x − 4x + 3x + 3x − 1, (Vandermonde’s quintic) (1) 5 4 3 2 G(x) = x + 2x − 5x − 2x + 4x − 1. (2) The polynomial V (x) represents a period-ﬁve orbital equation of motion for at least three paradigmatic physical models in the so-called generating partition limit , 2 2 namely the quadratic map x = 2 − x , the H´enon map (x, y) 7→ (2 − x , y), t+1 21,22 2 2 and the canonical quartic map , namely x = (x − 2) − 2. For details see, t+1 23,24 e.g. Refs. . Apart from these fruitful applications, V (x) is the celebrated quintic solved by Vandermonde (1735-1796) using “Lagrange” resolvents a number of years before Lagrange, and radical expressions at a time when it was still unknown that 25,26,27 a general solution of quintic equations was not possible . October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence 4 Owen J. Brison and Jason A.C. Gallas Table 1. The ten transformations interconnecting V (x) and G(x). b b b b b Direct transforms: V (x) → G(x) 1 2 3 4 5 3 2 g g g g g D = −x + x + 3x − 2 4 1 2 5 3 1 g g g g g D = −x + 2x 5 2 4 3 1 2 3 2 g g g g g D = x − x − 2x + 1 1 5 3 2 4 3 4 2 g g g g g D = x − 4x − x + 2 3 4 5 1 2 4 4 3 2 g g g g g D = −x + x + 4x − 2x − 3 2 3 1 4 5 5 b b b b b Inverse transforms: G(x) → V (x) 1 2 3 4 5 4 3 2 v v v v v I = 4x + 10x − 15x − 15x + 9 2 3 5 1 4 1 4 3 2 v v v v v I = x + 2x − 5x − 3x + 3 2 3 1 4 5 2 4 3 2 v v v v v I = −2x − 5x + 7x + 7x − 3 4 1 2 5 3 3 4 3 2 v v v v v I = −x − 2x + 5x + 2x − 3 3 4 5 1 2 4 4 3 2 v v v v v I = −2x − 5x + 8x + 9x − 5 3 1 2 4 5 5 To deﬁne basis polynomials and interpolations, we ﬁx the roots of V (x) and G(x) in the following orders: v ≃ −1.68, v ≃ −0.83, v ≃ 0.28, v ≃ 1.30, v ≃ 1.91, 1 2 3 4 5 g ≃ −3.22, g ≃ −1.08, g ≃ 0.37, g ≃ 0.54, g ≃ 1.39. 1 2 3 4 5 Fixing a = v , we compute the ﬁve basis elements ℓ (x). To check for the ex- i i i istence of transformations allowing the passage from V (x) to G(x) one needs to consider the 120 permutations of the roots g , using each permuted set of roots as the points b . This generates a set {L (x)} of transformations, for n = 1,··· , 120. i n Proceeding in this way, we ﬁnd that, although the transformations may have both real and complex coeﬃcients, some permutations produce transformations having just integer or rational coeﬃcients. For the passage from V (x) to G(x) we ﬁnd ﬁve transformations given in the upper part of Table 1, together with the root permuta- tions leading to them. To obtain the inverse transformations, allowing the passage from G(x) to V (x), we ﬁx a = g and investigate the nature of the coeﬃcients of i i the 120 transformations obtained by taking b = v for all possible permutations i i of the roots v . As before, we ﬁnd ﬁve inverse transformations, also listed in Ta- ble 1 with the root permutations leading to them. To obtain the transformations we wrote a program for Maple 2018 (X86 64 LINUX) running on a Dell XPS 13 notebook. Recently, using systematic search of coeﬃcients , it was possible to dis- cover the nine new transformations that had not been found with p-adic methods. Systematic coeﬃcient search of the ten transformations required about 2.1 seconds and used 133.1 MB of memory. In contrast, Lagrange interpolation has enabled these transformations to be obtained in just 0.25 seconds and using only 13.7 MB of memory, a considerable improvement. This speedup of the classiﬁcation of orbital points is, of course, a desirable feature, that opens the possibility of investigating orbital equations of considerably higher degrees. While a systematic search of suitable Tschirnhausen transformations was able to ﬁnd the transformations in Table 1, the interpolation polynomials introduced here have the advantage of revealing concomitantly, as a byproduct, the nature of the action of the individual transformations on the roots. We remark that some of October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence Polynomial interpolation as detector of orbital equivalence 5 the transformations in Table 1 are reducible, e.g., D (x) is clearly reducible as also are D , D , I , and I . 1 4 2 3 3.2. Hasse’s problem: Equivalence of quintics with complex roots As a second application, we use polynomial interpolation to uncover four new trans- formations, Eqs. (3)-(6), providing a complete and “symmetric solution”, i.e. a so- lution providing both direct and inverse connections for a classical problem posed by Hasse, who conjectured the possible existence of an isomorphism between three 2 28,29 quintics sharing a factor 47 in their discriminant . Such isomorphism was in- deed conﬁrmed by Zassenhaus and Liang , who used a p-adic method to uncover a pair of generating automorphisms of the Hilbert class ﬁeld over Q( −47). Their pair of transformations is certainly enough to establish isomorphism of the quintics but, as mentioned, does not provide an unbiasedly balanced and symmetric solu- tion to Hasse’s problem, i.e. a solution containing all possible direct and inverse transformations among all polynomials involved. Hasse’s problem is concerned with relations between the zeros of three quintic 30 31 28,29 equations obtained by Weber , by Fricke , and by Hasse , while inves- tigating class invariants for modular equations with discriminant −47. The three quintics found by these authors are, respectively, 5 3 2 f = x − x − 2 x − 2 x − 1, θ the real root; W W 5 4 3 2 f = x − x + x + x − 2 x + 1, θ the real root; F F 5 3 2 f = x + 10 x − 235 x + 2610 x − 9353, θ the real root. H H As reported by Zassenhaus and Liang , Hasse asked whether or not θ , θ , θ W F H generate the same ﬁeld. And if so, how to express these roots in terms of each other? Zassenhaus and Liang demonstrated that the polynomials indeed generate the same ﬁeld as manifest by the following transformations: θ = 5θ − 5θ − 2, H W θ = − θ − 2θ + 1. W F Proceeding as before, we ﬁnd the additional root interconnections which read, in the notation used by Zassenhaus and Liang: 4 3 θ = −θ + θ + θ + 1, (3) F W W W 4 3 2 θ = 10θ − 5θ + 5θ + 10θ − 12, (4) H F F F F 4 3 2 θ = 6θ + 23θ + 194θ − 1308θ + 9821 , (5) W H H H H 4 3 2 θ = − θ + 13θ + 179θ + 717θ − 444 . (6) F H H H H which may be easily veriﬁed. Note the conspicuous presence of non-integer coeﬃ- cients in Eqs. (5) and (6), not a common occurrence in the literature. Thus, poly- nomial interpolation is also able to deal with situations involving not only real but October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence 6 Owen J. Brison and Jason A.C. Gallas also complex roots. Of course, not only the real roots but also all complex roots are properly transformed by the same transformations above. When added to the known pair, the four new transformations reported here solve Hasse’s problem completely and symmetrically, with no bias. Table 2. Direct transformations among the s (x) of Eqs. (7)-(10). b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 1 2 5 4 3 2 1 4 2 6 5 3 2x − x − 14x − 4x + 10x + 2 5 4 3 2 2 1 3 4 6 5 −3x + x + 21x + 9x − 11x − 4 5 4 3 2 3 2 5 1 4 6 −6x + 2x + 43x + 17x − 28x − 7 5 4 3 2 4 6 1 5 3 2 4x − 2x − 28x − 7x + 18x + 2 5 3 2 5 3 6 2 1 4 3x − 22x − 15x + 12x + 6 6 5 4 3 2 1 −x b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 1 3 5 4 3 2 1 2 5 3 6 4 −7x + 2x + 50x + 22x − 30x − 8 5 4 3 2 2 3 1 6 4 5 6x − 2x − 43x − 17x + 29x + 7 4 3 2 3 6 2 4 5 1 −x + x + 6x − x − 2 5 4 3 2 4 5 6 1 2 3 −x + x + 7x − 2x − 7x + 2 5 4 3 2 5 1 4 2 3 6 −2x + x + 14x + 4x − 9x − 2 5 4 3 2 6 4 3 5 1 2 4x − x − 29x − 13x + 18x + 5 b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 1 4 5 4 3 2 1 3 4 2 6 5 −4x + x + 29x + 13x − 18x − 5 5 4 3 2 2 6 3 5 4 1 2x − x − 14x − 4x + 9x + 2 5 4 3 2 3 2 1 6 5 4 x − x − 7x + 2x + 7x − 2 4 3 2 4 1 5 3 2 6 x − x − 6x + x + 2 5 4 3 2 5 4 6 1 3 2 −6x + 2x + 43x + 17x − 29x − 7 5 4 3 2 6 5 2 4 1 3 7x − 2x − 50x − 22x + 30x + 8 b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 2 3 4 3 2 1 5 4 2 6 3 −x − x + 6x + x − 2 5 4 3 2 2 1 5 3 4 6 −4x − x + 29x − 13x − 18x + 5 5 4 3 2 3 2 1 6 5 4 x + x − 7x − 2x + 7x + 2 5 4 3 2 4 6 3 5 2 1 7x + 2x − 50x + 22x + 30x − 8 5 4 3 2 5 4 6 1 3 2 −6x − 2x + 43x − 17x − 29x + 7 5 4 3 2 6 3 2 4 1 5 2x + x − 14x + 4x + 9x − 2 b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 2 4 5 4 3 2 1 4 5 3 6 2 −2x − x + 14x − 4x − 9x + 2 5 4 3 2 2 3 1 6 4 5 6x + 2x − 43x + 17x + 29x − 7 5 4 3 2 3 1 4 2 5 6 −7x − 2x + 50x − 22x − 30x + 8 5 4 3 2 4 5 6 1 2 3 −x − x + 7x + 2x − 7x − 2 5 4 3 2 5 6 2 4 3 1 4x + x − 29x + 13x + 18x − 5 4 3 2 6 2 3 5 1 4 x + x − 6x − x + 2 b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 3 4 5 4 3 2 1 3 2 5 4 6 2x − 6x − 7x + 8x + 3x − 1 4 3 2 2 6 5 1 3 4 −2x + 6x + 7x − 8x − 2 4 3 2 3 2 6 4 1 5 x − 2x − 6x + 2 5 4 3 2 4 1 3 6 5 2 −2x + 7x + 4x − 12x + 3x + 2 3 2 5 4 1 2 6 3 −x + 3x + 3x − 3 6 5 4 3 2 1 −x October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence Polynomial interpolation as detector of orbital equivalence 7 Table 3. Inverse transformations among the s (x) of Eqs. (7)-(10). b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 2 1 5 4 3 2 1 3 6 2 5 4 −6x − 2x + 43x − 17x − 2x + 7 5 4 3 2 2 1 3 4 6 5 −3x − x + 21x − 9x − 11x + 4 5 3 2 3 6 5 1 4 2 3x − 22x + 15x + 12x − 6 5 4 3 2 4 2 1 5 3 6 2x + x − 14x + 4x + 10x − 2 5 4 3 2 5 4 2 6 1 3 4x + 2x − 28x + 7x + 18x − 2 6 5 4 3 2 1 −x b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 3 1 5 4 3 2 1 2 4 6 3 5 x − 2x − 6x − x + 4x + 2 5 4 3 2 2 4 5 3 1 6 x − 3x − 3x + 3x − x 5 4 3 2 3 1 2 5 6 4 −x + 3x + 3x − 3x + 2x 5 4 3 2 4 5 6 1 2 3 −x + 2x + 7x − 2x − 7x + 1 5 4 3 2 5 6 3 2 4 1 x − 4x − x + 9x − 2x − 2 5 4 2 6 3 1 4 5 2 −x + 4x − 6x + 4x b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 4 1 5 4 3 2 1 4 2 3 6 5 −x − 4x + x + 9x + 2x − 2 5 4 2 2 5 4 1 3 6 x + 4x − 6x − 4x 5 4 3 2 3 2 1 6 5 4 x + 2x − 7x − 2x + 7x + 1 5 4 3 2 4 6 5 2 1 3 x + 3x − 3x − 3x − 2x 5 4 3 2 5 3 6 4 2 1 −x − 2x + 6x − x − 4x + 2 5 4 3 2 6 1 3 5 4 2 −x − 3x + 3x + 3x + x b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 3 2 5 4 2 1 4 6 3 2 5 x − 4x + 6x − 4x 5 4 3 2 2 1 4 5 3 6 −x + 4x + x − 9x + 2x + 2 5 4 3 2 3 2 1 6 5 4 x − 2x − 7x + 2x + 7x − 1 5 4 3 2 4 6 5 2 1 3 x − 3x − 3x + 3x − 2x 5 4 3 2 5 3 2 4 6 1 −x + 3x + 3x − 3x + x 5 4 3 2 6 5 3 1 4 2 −x + 2x + 6x + x − 4x − 2 b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 4 2 5 4 3 2 1 6 4 2 3 5 x + 3x − 3x − 3x − x 5 4 3 2 2 4 1 3 5 6 x + 2x − 6x + x + 4x − 2 5 4 3 2 3 1 2 5 6 4 −x − 3x + 3x + 3x + 2x 5 4 3 2 4 5 6 1 2 3 −x − 2x + 7x + 2x − 7x − 1 5 4 2 5 2 3 6 4 1 −x − 4x + 6x + 4x 5 4 3 2 6 3 5 4 1 2 x + 4x − x − 9x − 2x + 2 b b b b b b Transforms from s (x) → s (x) 1 2 3 4 5 6 4 3 5 4 3 2 1 3 2 5 4 6 2x + 6x − 7x − 8x + 3x + 1 4 3 2 2 6 3 1 5 4 −x − 2x + 6x − 2 4 3 2 3 4 6 2 1 5 2x + 6x − 7x − 8x + 2 3 2 4 1 5 6 3 2 −x − 3x + 3x + 3 5 4 3 2 5 2 1 4 6 3 −2x − 7x + 4x + 12x + 3x − 2 6 5 4 3 2 1 −x As described by Pohst , Hasse’s problem played an important role in estab- lishing computational algebraic number theory at a time when computations of all kinds were taboo. In the early 1960s, Zassenhaus developed algorithmic means for number theoretical experiments in algebraic number theory. A major success was October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence 8 Owen J. Brison and Jason A.C. Gallas the proof of the isomorphism of the three quintic ﬁelds which occurred as candidates for the real subﬁeld of the Hilbert class ﬁeld of Q( −47). This problem, pointed out by Hasse, could not be solved by theoretical methods. It was the numerical solu- tion by Zassenhaus and Liang that gave major credit to methods in constructive algebraic number theory. 3.3. Equivalence among totally real sextics with small coeﬃcients A very interesting and much studied class of equations of motion involves sextic polynomials . Their splitting ﬁelds may contain quadratic and cubics subﬁelds or no subﬁelds at all. Thus, sextics may appear generically as orbital clusters al- gebraically entangling orbits into distinct groups of periodicity. Of particular inter- est is knowledge concerning cyclic sextics. The minimum discriminant of sextics, 3 4 300, 125 = 5 · 7 , was found by Liang and Zassenhaus for the polynomial s (x) a totally real cyclic sextic : 6 5 4 3 2 s (x) = x − x − 7x + 2x + 7x − 2x − 1, (7) 6 5 4 3 2 s (x) = x + x − 7x − 2x + 7x + 2x − 1, (8) 6 5 4 3 2 s (x) = x − 2x − 7x + 2x + 7x − x − 1, (9) 6 5 4 3 2 s (x) = x + 2x − 7x − 2x + 7x + x − 1. (10) The same minimum discriminant is also shared by s (x), s (x), s (x), simple rein- 2 3 4 carnations of s (x) after the substitutions x → ±x and x → ±1/x and suitable simpliﬁcations. Curiously, the roots of the above polynomials are interconnected in subtle ways by a multitude of transformations, given in Tables 2 and 3. The values of the b given in these tables indicate the order of the roots needed to ﬁnd the cor- responding transformations. As before for Vandermonde’s quintic, we assume the roots of the s (x) to be ordered from the smallest to the largest. Thus, denoting by (i) s , i = 1, . . . , 6, the ordered roots of s (x), the transformation shown in the ﬁrst (i) line of Table 2 is obtained when ﬁxing b = s , and so on. Similarly as for the s (x), polynomials characterized by totally real cyclic sextics and second lowest discriminant, namely 371, 293, are the following: 6 5 4 3 2 t (x) = x − x − 5 x + 4 x + 6 x − 3 x − 1, 6 5 4 3 2 t (x) = x + x − 5 x − 4 x + 6 x + 3 x − 1, 6 5 4 3 2 t (x) = x − 3 x − 6 x + 4 x + 5 x − x − 1, 6 5 4 3 2 t (x) = x + 3 x − 6 x − 4 x + 5 x + x − 1, 6 5 4 3 2 t (x) = x − 2 x − 7 x + 6 x + 5 x − 5 x + 1, 6 5 4 3 2 t (x) = x + 2 x − 7 x − 6 x + 5 x + 5 x + 1, 6 5 4 3 2 t (x) = x − 5 x + 5 x + 6 x − 7 x − 2 x + 1, 6 5 4 3 2 t (x) = x + 5 x + 5 x − 6 x − 7 x + 2 x + 1. As for the s (s), note the conspicuous presence of two groups of four elements arising from the substitutions x → ±x and x → ±1/x and suitable simpliﬁcations. October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence Polynomial interpolation as detector of orbital equivalence 9 Table 4. Bridges among sextics arising in orbital clusters of the H´enon Hamiltonian repeller. Note the non-integer coeﬃcients in the ﬁrst column, not a common occurrence in the literature. Direct Inverse 1 2 4 3 2 X → Y : − x + 3 Y → X : −x + 2x + 9x − 13x − 16 4 2 4 3 2 X → Z : − x + 3x − x − 3 Z → X : −x − 2x + 11x + 11x − 30 4 3 2 2 Y → Z : x − 2x − 10x + 13x + 22 Z → Y : −x − x + 6 = −(x + 3)(x − 2) 1 4 2 5 4 3 2 U → W : − x + 3x − x − 3 W → U : −4x − 6x + 63x + 74x − 234x − 219 1 2 5 4 3 2 − x + 3 2x + 3x − 32x − 38x + 121x + 116 To each pair of polynomials t (x) corresponds a set of six transformations anal- ogous to the ones in Tables 2 and 3, resulting from similar root permutations. The total number of such transformations is 64, too many to be recorded here explicitly. However, the polynomials t (x) allow them to be obtained easily if so desired, along with the proper root permutations leading to them. 3.4. Equivalent totally real cyclic sextics with larger coeﬃcients Table 4 reports isomorphisms having a direct bearing on the inner workings of the H´enon Hamiltonian repeller . As seen in the previous Section, sextics leading to minimal discriminants tend to have comparatively small coeﬃcients. However, in real-life applications the coeﬃcients are not always so small, for instance the (5) 2 2 cluster A (x) = X(x)Y (x)Z (x) deﬁning the orbital coordinates of six period- ﬁve trajectories of the H´enon Hamiltonian repeller involves totally real sextics with larger coeﬃcients: 6 5 4 3 2 18 X(x) = x − 2 x − 14 x + 24 x + 32 x − 16 x − 8, Δ = 2 · 31 · 241 · 389, 6 5 4 3 2 6 Y (x) = x − 2 x − 16 x + 26 x + 81 x − 84 x − 125, Δ = 2 · 31 · 241 · 389, 6 5 4 3 2 6 Z(x) = x + 2 x − 16 x − 22 x + 85 x + 60 x − 151, Δ = 2 · 31 · 241 · 389. (5) 2 Similarly, a trio of period-ﬁve orbits is amalgamated into B (x) = U(x)W (x), deﬁning the eighteen orbital coordinates as roots of a pair of totally real sextics , with Galois group 6T7: 6 4 3 2 18 4 2 U(x) = x − 22x + 8x + 124x − 88x − 32, Δ = 2 · 3 · 659 , 6 5 4 3 2 6 4 2 W (x) = x + 4x − 12x − 58x + 12x + 202x + 139, Δ = 2 · 3 · 659 . Table 4 shows that the orbits algebraically entangled together to form the above pair of orbital clusters have their coordinates related by simple transformations that, surprisingly, allow back and forth passage among seemingly distinct orbits. Unfor- tunately, the Galois group of X(x), Y (x), Z(x), U(x), and W (x) is the symmetric group, meaning that these sextics cannot be solved in terms of a ﬁnite number of radical extractions and elementary arithmetic operations. But the transformations interconnecting these sextics show clearly that knowledge of just two sets of six roots is enough to interconnect in phase-space all orbital points of the equations October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence 10 Owen J. Brison and Jason A.C. Gallas Table 5. Some representative bridges that allow direct and inverse passage among polynomials of the families f (x) and g (x). There are no connections between the f (x) and g (x) and i i i i 2 3 vice-versa, despite the fact that they all share the same discriminant 810, 448 = 2 · 37 . Direct Inverse f → f : −x f → f : −x 1 2 2 1 x − 1 x + 1 5 4 3 2 5 4 3 2 −x + 2x + 4x − 6x − 4x + 2 −x − 3x + 2x + 8x − x − 2 5 4 3 2 5 4 3 2 −x + 3x + 2x − 8x − x + 2 −x − 2x + 4x + 6x − 4x − 2 5 4 3 2 5 4 3 2 x − 3x − 2x + 8x + x − 3 x + 2x − 4x − 6x + 4x + 3 5 4 3 2 5 4 3 2 x − 2x − 4x + 6x + 4x − 3 x + 3x − 2x − 8x + x + 3 g → g : −x g → g : −x 1 2 2 1 5 4 3 2 5 4 3 2 x − 5x + 8x − 9x + 8x − 5 x + 5x + 8x + 9x + 8x + 5 5 4 3 2 5 4 3 2 g → g : −x + 4x − 4x + 5x − 3x + 2 g → g : −x + 3x − 6x + 7x − 2x 1 3 3 1 5 4 3 2 5 4 3 2 x − 4x + 4x − 5x + 3x − 1 x − 2x + 4x − 3x − x + 1 5 4 3 2 5 4 3 2 g → g : −x + 4x − 4x + 5x − 3x + 1 g → g : −x − 2x − 4x − 3x + x + 1 1 4 4 1 5 4 3 2 5 4 3 2 x − 4x + 4x − 5x + 3x − 2 x + 3x + 6x + 7x + 2x g → g : −x g → g : −x 1 5 5 1 5 4 3 2 5 3 2 x − 5x + 8x − 9x + 8x − 4 x − 2x + 5x − x + 2 g → g : x + 1 g → g : x − 1 1 6 6 1 5 4 3 2 5 3 2 −x + 5x − 8x + 9x − 8x + 4 −x + 2x + 5x + x + 2 (5) algebraically entangled in each cluster. The sextic trio of cluster A (x) is not iso- (5) morphic to the pair of sextics of cluster B (x). Table 4 contains transformations with non-integer coeﬃcients, something that we have not been able to ﬁnd in the literature. 3.5. Equivalence among distinct families of isodiscriminant sextics As a quite remarkable ﬁnal example, we consider a family of ten totally real sextics 2 3 sharing the same discriminant, 810, 448 = 2 ·37 , but formed by two non-isomorphic families f (x) and g (x), deﬁned in Eqs. (11)-(20). The four sextics f (x) are iso- i i i morphic among themselves, as also are the six g (x). However, none of the f (x) is i i isomorphic to any of the g (x) and vice-versa, despite the fact that they all share the same discriminant. In the notation of Butler and McKay adopted by Maple, the Galois group of the f (x) is 6T1, a cyclic semiabelian group, while the group of the g (x) is 6T8, a solvable, semiabelian group. 6 5 4 3 f (x) = x − 3x − 2x + 9x − 5x + 1, (11) 6 5 4 3 f (x) = x + 3x − 2x − 9x + 5x + 1, (12) 6 5 3 2 f (x) = x − 5x + 9x − 2x − 3x + 1, (13) 6 5 3 2 f (x) = x + 5x − 9x − 2x + 3x + 1, (14) 4 October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence Polynomial interpolation as detector of orbital equivalence 11 6 5 4 3 2 g (x) = x − 5x + 8x − 9x + 8x − 5x + 1, (15) 6 5 4 3 2 g (x) = x + 5x + 8x + 9x + 8x + 5x + 1, (16) 6 5 4 3 2 g (x) = x − 3 x + 6 x − 7 x + 2 x + x − 1, (17) 6 5 4 3 2 g (x) = x + 3x + 6x + 7x + 2x − x − 1, (18) 6 5 4 3 2 g (x) = x − x − 2 x + 7 x − 6 x + 3 x − 1, (19) 6 5 4 3 2 g (x) = x + x − 2 x − 7 x − 6 x − 3 x − 1. (20) Once again, note in Eqs. (11)-(20) two groups of four elements underlying the sub- stitutions x → ±x and x → ±1/x, and the presence of two outliers, namely the reciprocal polynomials g (x) and g (x). 1 2 Proceeding as before, from the roots of Eqs. (11)-(20) one may easily obtain the large set of transformations allowing back and forth passage, local to the global, among both groups of sextics. There is a total of six transformations connecting each pair of f (x) but just two transformations connecting pairs of g (x). The complete i i set of transformations is omitted here, with just a few representative ones being given in Table 5. 4. Conclusions and outlook This paper has shown that Lagrange interpolation works as an eﬃcient detector of equivalence and isomorphism among orbital equations of motion of algebraic dynam- ical systems governed by discrete-time mappings. This is a startling new application for a well-known interpolation technique of numerical analysis. Here, it is not used to approximate anything but, instead, as means of obtaining exact analytical ex- pressions for isomorphisms. We found polynomial interpolation to eﬃciently detect equivalences among equations of any signature, i.e. among polynomials having only real roots or not. The method is simple to implement and very fast. We antici- pate polynomial interpolation to be a helpful tool to locate equivalences among the huge number of orbital equations in systems of algebraic origin and polynomials in general. In particular, it should help to uncover equivalences among, e.g., the complicated amalgamation polynomial clusters arising in the Hamiltonian repeller limit of the H´enon map , and among orbits of the Pincherle map, a paradigmatic 23,36,37,38 map underlying the operating kernel of the so-called chaotic computer . Acknowledgments JACG thanks helpful email exchanges with J. Voight and J.C. Interlando. The latter brought Ref. to our attention. OJB acknowledges the partial support of the Funda¸ca˜o para a Ciˆencia e Tecnologia, project UID/MAT/04721/2013 (Cen- tro de An´alise Funcional, Estruturas Lineares e Aplica¸co˜es - Grupo de Estruturas Lineares, Alg´ebricas e Combinat´orias, Universidade de Lisboa, Portugal). JACG was supported by CNPq, Brazil. This work was also supported by the Max-Planck October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence 12 Owen J. Brison and Jason A.C. Gallas Institute for the Physics of Complex Systems, Dresden, in the framework of the Advanced Study Group on Optical Rare Events. References 1. O.J. Brison and J.A.C. Gallas, What is the eﬀective impact of the explosive orbital growth in discrete-time one-dimensional polynomial dynamical systems? Physica A 410, 413–418 (2014). 2. M.J.B. Hauser and J.A.C. Gallas Nonchaos-mediated mixed-mode oscillations in an enzyme reaction system, J. Phys. Chem. Lett. 5, 4187–4193 (2014). 3. L. Junges and J.A.C. Gallas, Stability charts for continuous wide-range control of two mutually delay-coupled semiconductor lasers, New J. Phys. 17, 053038 (2015). 4. J.G. Freire, M.R. Gallas, and J.A.C. Gallas, Chaos-free oscillations, Europhys. Lett. 118, 38003 (2017). 5. J.A.C. Gallas, Spiking systematics in some CO laser models, Adv. Atom. Molec. Opt. Phys. 65, 127–191 (2016). 6. R.J. Field, Chaos in the Belousov Zhabotinsky reaction, Mod. Phys. Lett. B 29, 1530015 (2015). 7. J.G. Freire, M.R. Gallas, and J.A.C. Gallas, Impact of predator dormancy on prey- predator dynamics, Chaos 28, 053118 (2018). 8. M.R. Gallas and J.A.C. Gallas, Nested arithmetic progressions of oscillatory phases in Olsen’s enzyme reaction model, Chaos 25, 064603 (2015). 9. M.R. Gallas, M.R. Gallas, and J.A.C. Gallas, Distribution of chaos and periodic spikes in a three-cell population model of cancer, European Phys. J. Special Topics 223, 2131–2144 (2014). 10. H. Zassenhaus and J. Liang, On a problem of Hasse, Math. Comp. 23, 515–519 (1969). 11. P. Cartier and Y. Roy, On the enumeration of quintic ﬁelds with small discriminants, J. reine angew. Math. 268/269, 213–215 (1974). 12. H. Hasse, Vorlesungen u¨ber Zahlentheorie, Zweite Auﬂage (Springer, Berlin, 1964). 13. M. van Hoeij and V. Pal, Isomorphisms of algebraic number ﬁelds, J. Theorie Nombres de Bordeaux 24, 293–305 (2012). 14. J.A.C. Gallas, Equivalence among orbital equations of polynomial maps, Int. J. Mod. Phys. C 29, ??? (2018), in print. DOI 10.1142/S0129183118500821 15. J.H. Silverman, The Arithmetic of Dynamical Systems (Springer, New York, 2007). 16. K. Gatermann, Computer Algebra Methods for Equivariant Dynamical Systems (Springer, Berlin, 2000). 17. K. Schmidt, Dynamical Systems of Algebraic Origin (Birkh¨auser, Basel, 1995). 18. E. Meijering, A chronology of interpolation: From ancient astronomy to modern signal and image processing, Proc. IEEE 90, 319–342 (2002). 19. H. Cohn, A numerical study of quintics of small discriminant, Commun. Pure Appl. Math. 8, 377–386 (1955). 20. R. Gilmore and M. Lefranc, The Topology of Chaos, 2nd Edition (Wiley-VCH, Wein- heim, 2011). 21. J.A.C. Gallas, Dissecting shrimps: results for some one-dimensional physical systems, Physica A 202, 196–223 (1994). 22. J.A.C. Gallas, Structure of the parameter space of a ring cavity, Appl. Phys. B Sup- plement 60, S203–S213 (1995). 23. J.A.C. Gallas, Method for extracting arbitrarily large orbital equations of the Pin- cherle map, Results in Physics 6, 561–567 (2016). 24. J. Argyris, G. Faust, M. Haase, R. Friedrich, An Exploration of Dynamical Systems October 5, 2018 0:27 WSPC/INSTRUCTION FILE orbital˙equivalence Polynomial interpolation as detector of orbital equivalence 13 and Chaos, 2nd ed. (Springer, New York, 2015). 25. C. Itzigsohn (Editor), Abhandlungen aus der reinen Mathematik von N. Vandermonde, (Springer, Berlin, 1888). 26. N. Nielsen, Observations sur des recherches alg`ebriques plus anciennes que le th´eor`eme d’Abel, Det Kgl. Danske Vidensk. Selskab., Mathematisk-fysiske Meddelelser. 8, 1–31 (1928). 27. H. Lebesgue, L’Œuvre math´ematique de Vandermonde, L’Enseignement Math´emati- que 1, 203–223 (1955). 28. H. Hasse, Uber den Klassenko¨rper zum quadratischen Zahlk¨orper mit der Diskrimi- nante −47, Acta Arith. 9, 419–434 (1964). 29. H. Hasse and J. Liang, Uber den Klassenko¨rper zum quadratischen Zahlk¨orper mit der Diskriminante −47 (Fortsetzung), Acta Arith. 16, 89–97 (1969). 30. H. Weber, Lehrbuch der Algebra, Dritter Band, 2. Auﬂ. §131, page 485 (Vieweg, Braunschweig, 1908) 31. R. Fricke, Lehrbuch der Algebra, Dritter Band, Kap. 5 §4, page 492 (Vieweg, Braun- schweig, 1928) 32. M. Pohst, In memoriam Hans Zassenhaus 1912-1991, J. Number Theory 47, 1–19 (1994). 33. A. Endler and J.A.C. Gallas, Reductions and simpliﬁcations of orbital sums in a Hamiltonian repeller, Phys. Lett. A 352, 124–128 (2006). 34. J. Liang and H. Zassenhaus, The minimum discriminant of sixth degree totally com- plex algebraic number ﬁelds, J. Number Theory 9, 16-35 (1977). 35. G. Butler and J. McKay, The transitive groups of degree up to eleven, Commun. Algebra 11, 863–911 (1983). 36. W.L. Ditto, A. Miliotis, K. Murali, S. Sinha, and M.L. Spano, Chaogates: morphing logic gates that exploit dynamical patterns, Chaos 20, 037107 (2010). 37. D.N. Guerra, A.R. Bulsara, W.L. Ditto, S. Sinha, K. Murali, and P. Mohanty, Noise- assisted reprogrammable nanomechanical logic gate, Nano Lett. 10, 1168–1171 (2010). 38. S. Sinha and W.L. Ditto, Dynamics based computation, Phys. Rev. Lett. 81, 2156– 2159 (1998).

Condensed Matter – arXiv (Cornell University)

**Published: ** Sep 29, 2018

Loading...

You can share this free article with as many people as you like with the url below! We hope you enjoy this feature!

Read and print from thousands of top scholarly journals.

System error. Please try again!

Already have an account? Log in

Bookmark this article. You can see your Bookmarks on your DeepDyve Library.

To save an article, **log in** first, or **sign up** for a DeepDyve account if you don’t already have one.

Copy and paste the desired citation format or use the link below to download a file formatted for EndNote

Access the full text.

Sign up today, get DeepDyve free for 14 days.

All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.