ebook img

Simplifying the Reinsch algorithm for the Baker-Campbell-Hausdorff series PDF

0.26 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Simplifying the Reinsch algorithm for the Baker-Campbell-Hausdorff series

Simplifying the Reinsch algorithm for the Baker–Campbell–Hausdorff series Alexander Van–Brunt∗ and Matt Visser† School of Mathematics, Statistics, and Operations Research, Victoria University of Wellington; PO Box 600, Wellington 6140, New Zealand. (Dated: 21 January 2015; 27 January 2015; LATEX-ed January 28, 2015) The Baker–Campbell–Hausdorff series computes thequantity ∞ Z(X,Y)=ln eXeY = z (X,Y), 5 (cid:16) (cid:17) X n 1 n=1 0 where X and Y are not necessarily commuting, in terms of homogeneous multinomials z (X,Y) n 2 of degree n. (This is essentially equivalent to computing the so-called Goldberg coefficients.) The n Baker–Campbell–Hausdorff series is a general purpose tool of wide applicability in mathematical a physics, quantum physics, and many other fields. The Reinsch algorithm for the truncated series J permits one to calculate up to some fixed order N by using (N +1)×(N +1) matrices. We 7 show how to further simplify the Reinsch algorithm, making implementation (in principle) utterly 2 straightforward. This helps provide a deeper understanding of the Goldberg coefficients and their properties. Forinstanceweestablishstrictbounds(andsomeequalities)onthenumberofnon-zero ] Goldberg coefficients. Unfortunately, we shall see that the number of terms in the multinomial h z (X,Y) often grows very rapidly (in fact exponentially) with thedegree n. p n We also present some closely related results for thesymmetric product - h ∞ at S(X,Y)=ln(cid:16)eX/2eYeX/2(cid:17)=Xsn(X,Y). m n=1 [ Variations on these themes are straightforward. For instance, one can just as easily consider the series 2 ∞ v L(X,Y)=ln eXeYe−Xe−Y = ℓ (X,Y). (cid:16) (cid:17) X n 4 n=1 3 This type of series is of interest, for instance, when considering parallel transport around a closed 0 curve. Several other related series are investigated. 5 0 Keywords: Commutators, matrix exponentials, matrix logarithms, Baker–Campbell–Hausdorff . 1 formula. 0 5 1 : v i X r a ∗ [email protected][email protected] 2 CONTENTS I. Introduction 3 II. The simplified algorithm: Low-order terms — z to z 3 1 4 III. The simplified algorithm: Some medium-low order terms — z to z 5 5 8 A. The z term 5 5 B. The z term 6 6 C. The z term 7 7 D. The z term 9 8 IV. Some properties of the Goldberg coefficients 11 V. The number of non-zero Goldberg coefficients 12 A. Bounds on # 13 n B. Bounds on # 14 2n C. Comments 16 VI. Symmetric Baker–Campbell–Hausdorffformula 16 VII. Some other variants of the Baker–Campbell–Hausdorffformula 18 A. Loop Baker–Campbell–Hausdorffformula 18 B. Triangular Baker–Campbell–Hausdorffformula 19 C. Sum and difference Baker–Campbell–Hausdorffformula 19 D. Highly-symmetrized Baker–Campbell–Hausdorffformula 20 E. Symmetric sum and difference Baker–Campbell–Hausdorffformula 20 F. Highly-symmetrized sum and difference Baker–Campbell–Hausdorffformula 21 VIII. Discussion 21 Acknowledgments 21 A. Some Maple code 22 References 23 3 I. INTRODUCTION WhatisnowcalledtheBaker–Campbell–Hausdorffformula (orseries)hasbeenstudiedforwelloveracentury[1–5]. In the currentarticle we will discuss the relatively recently developed Reinsch algorithm[6–9], and relations between this algorithm and Goldberg’s results from 1956 [10–12]. One version of the Baker–Campbell–Hausdorff formula is the expansion [10–12] Z(X,Y)=ln(eXeY)= g(w)w, (1.1) Xw in terms of the rational-number Goldberg coefficients g(w), and “words” w constructed from the “alphabet” {X,Y}. The homogeneous multinomials z (X,Y) of degree n are then constructed in terms of words of length n: n z (X,Y)= g(w)w(X,Y). (1.2) n X |w|=n In fact what is now called the Baker–Campbell–Hausdorff theorem is the result that the multinomials z (X,Y) are n in fact representable in terms of nested commutators (Lie brackets). Specifically, Dynkin’s expansion [4, 5] amounts (with hindsight) to the observation that [11, 12] 1 z (X,Y)= g(w) [w(X,Y)], (1.3) n n X |w|=n where [w(X,Y)] denotes the nested commutator built from the word w(X,Y). For definiteness, we consider right- nested commutators of the form [ABCDEF ...] = [A,[B,[C,[D,[E,[F,...]]]]]]. This is sometimes called the “long commutator”. Note that in view of anti-symmetry andthe Jacobiidentity many different words w(X,Y) can mapto the same nested commutator [w(X,Y)]. Then ∞ Z(X,Y)=ln eXeY = z (X,Y), (1.4) n (cid:0) (cid:1) nX=1 andthegoalistofindefficientwaysofcalculatingthe g(w)and/orthez (X,Y). Since eXeY −1 =e−Ye−X wehave n Z(X,Y)=−Z(−Y,−X). Thisimpliesthesymmetriesz2n(X,Y)=−z2n(Y,X)andz2n(cid:0)+1(X,(cid:1)Y)=z2n+1(Y,X). The first few terms can be represented as 1 1 1 Z(X,Y)=ln(eXeY)=X +Y + [X,Y]+ [X−Y,[X,Y]]− [X,[Y,[X,Y]]]+... (1.5) 2 12 24 Unfortunately the expansion rapidly becomes extremely unwieldy. Specifically, we shall show that the limit superior of the number of terms in the Goldberg expansion grows exponentially in n. Thus even though explicit computer- aided computations can on a modern laptop easily be carried out to n=13, (or sometimes higher if one focusses on specific questions), beyond n = 8 or thereabouts the resulting formulae are simply too cumbersome to be usefully writtendownonpaper. (This style ofapproachcomplements whatwhatcanbe extractedby consideringspecial-case commutators, as in [13].) II. THE SIMPLIFIED ALGORITHM: LOW-ORDER TERMS — z1 TO z4 Our simplified versionofReinsch’s algorithmis this: Suppose one wishes to calculate up to some fixed wordlength N. Construct two (N +1)×(N +1) matrices that are zero except for the first super-diagonalwhere they contain N distinct elements. That is: 0 x 0 0 ··· 0 0 0 y 0 0 ··· 0 0 1 1 0 0 x 0 ··· 0 0 0 0 y 0 ··· 0 0  2   2  0 0 0 x ··· 0 0 0 0 0 y ··· 0 0 3 3 XN = ... ... ... ... ... ... ... , and YN = ... ... ... ... ... ... ... . (2.1) 0 0 0 0 ··· x 0  0 0 0 0 ··· y 0   N−1   N−1  0 0 0 0 ··· 0 x  0 0 0 0 ··· 0 y   N   N  0 0 0 0 ··· 0 0  0 0 0 0 ··· 0 0      4 Now compute (eg, using Maple or some equivalent package), the (N +1)×(N +1) matrix Z =ln(exp(X )exp(Y )). (2.2) N N N The first row of the matrix Z is (essentially) the information we want. Specifically, noting that the matrix Z is n N strictly upper triangular, let us denote 0 z z z ··· z z 1 2 3 N−1 N 0 0 ∗ ∗ ··· ∗ ∗   0 0 0 ∗ ··· ∗ ∗ ZN = ... ... ... ... ... ... ... . (2.3) 0 0 0 0 ··· ∗ ∗    0 0 0 0 ··· 0 ∗    0 0 0 0 ··· 0 0    Here the ∗’s denote nonzero quantities that are not of specific interest for current purposes. Brute force computation, (see Appendix A for appropriate Maple code), yields as the first four terms: z =x +y ; (2.4) 1 1 1 1 z = (x y −y x ); (2.5) 2 1 2 1 2 2 1 z = (x x y −2x y x +x y y +y x x −2y x y +y y x ); (2.6) 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 12 1 z = (x x y y −2x y x y +2y x y x −y y x x ). (2.7) 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 24 Already at the stage n=5 the terms are relatively turgid to explicitly write down. (Full explicit formulae for z , z , 5 6 z , and z are presented below in Section III.) The corresponding z (X,Y) multinomials are constructed by simply 7 8 n replacing x →X and y →Y while preserving the order of the letters. That is, for the first four terms, i i z (X,Y)=X +Y; (2.8) 1 1 z (X,Y)= (XY −YX); (2.9) 2 2 1 z (X,Y)= (X2Y −2XYX +XY2+YX2−2YXY +Y2X); (2.10) 3 12 1 z (X,Y)= (X2Y2−2XYXY +2YXYX −Y2X2). (2.11) 4 24 TheoriginalReinschalgorithm[6]involvedanextrasetofN signsymbolsσ ∈{+1,−1}N,andanadditional“symbol i conversionstage”, which the current algorithm avoids. The fundamental reason the algorithm works is that X and Y have been carefully constructed to not commute N N with each other, and to be linearly independent of each other. Furthermore, for m ∈ {1,...,N} all of the (X )m N and (Y )m are non-zero only on the m’th super-diagonal, and so are all linearly independent of each other. Indeed N (X )N+1 = 0 = (Y )N+1 so all matrix functions (in particular the matrix exponential and matrix logarithm) have N N finite Taylor series expansions. Specifically, let us consider the closely related ∞ × ∞ matrices δ x and δ y, where x and y need not i+1,j i+1,j commute. These matrices possess the property that if we have a homogeneous polynomial, P, of degree m, then P(δ x,δ y)=δ P(x,y). Consequently when we compute the Baker–Campbell–Hausdorffseries for these i+1,j i+1,j i+m,j specific matrices, then ∀i the (i,i+N) entry will be exactly all terms of degree N of the Baker–Campbell–Hausdorff expansion, for arbitrary variables x, y. In particular, now truncating to (N +1)×(N +1) matrices, this truncation naturallyeliminatesalltermsofordergreaterthanN;theothernon-zeroentriesintheresultingmatrixaretheterms of degree less than N. The introduction of the subscripts x and y to decorate the elements of these matrices is merely a way of getting i i around the software’s implicit assumption that the variables x and y commute. The elements of the first row of Z , N the z = [Z ] , are successively built up from a sum of exactly n products of the X and Y . Furthermore z n N 1,n+1 N N n will contain a x (respectively, a y ) if and only if the corresponding string of n matrices has a X (respectively, a i i N Y ) in its ith position. N This completes the description of the simplified Reinsch algorithm. 5 The next step, that of replacing words w by commutators [w], involves several elementary but tricky steps. For instance, from the definition of the long commutator [ABCDEF ...] = [A,[B,[C,[D,[E,[F,...]]]]]] we immediately have [wXa>1]=0=[wYb>1]. More subtly [w w ]=[w [w ]], which leads to 1 2 1 2 [w XYw ]=[w XY[w ]]=[w X[Y,[w ]]]=[w [X,[Y,[w ]]]]. (2.12) 1 2 1 2 1 2 1 2 Applying the Jacobi identity to this last term yields [w XYw ]=−[w [Y,[[w ],X]]]−[w [[w ],[X,Y]]]=[w [Y,[X,[w ]]]]+[w [[X,Y],[w ]]]. (2.13) 1 2 1 2 1 2 1 2 1 2 That is [w XYw ]=[w YXw ]+[w [X,Y]w ]. (2.14) 1 2 1 2 1 2 This quite general result immediately implies, for instance, [YXXY]=[XYXY]. When computing the commutator version of z , we encounter 3 [X2Y −2XYX +XY2+YX2−2YXY +Y2X]=[X2Y −2XYX −2YXY +Y2X] =[X2Y +2X2Y +2Y2Y +Y2X] =3[X2Y +Y2X] =3[(X −Y)XY]. (2.15) Similarly, when computing z one encounters 4 [X2Y2−2XYXY +2YXYX −Y2X2]=[−2XYXY +2YXYX]=−4[XYXY]. (2.16) This now yields z (X,Y)=X +Y; (2.17) 1 1 z (X,Y)= [XY]; (2.18) 2 2 1 z (X,Y)= [(X −Y)XY]; (2.19) 3 12 1 z (X,Y)= [XYXY]. (2.20) 4 24 (2.21) The word versions of z , z , z , and z , and the commutator versions of z and z , are presented below. 5 6 7 8 5 6 III. THE SIMPLIFIED ALGORITHM: SOME MEDIUM-LOW ORDER TERMS — z5 TO z8 In this Section we reproduce explicit results for z , z , z , and z . Beyond this stage the formulae are simply too 5 6 7 8 cumbersome to be usefully written down on paper. The only scientific justification for explicitly presenting even this level of detail is that it explicitly demonstrates the patterns and symmetries of the Goldberg coefficients g(w) in a somewhat non-trivial context. A. The z5 term For z we have: 5 y y x y x x y x y x y x y x y y x y x x y x y y x y x x y x 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 z =− + + − − − 5 120 30 30 120 120 120 y y y y x x y y y x x x y y x x x x y x y x y y y y x x y y 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 − + − + + − 720 180 120 180 180 120 y x x x y y x x x x y y x y y y y x x y y y x x x x y x y y 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 + − − − + − 180 720 120 120 180 120 x y x x y x y x x x y y y x y y y y x x x y y x y x y y x x 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 − + + + − − 120 180 180 180 120 120 x x y x y x x y x x x y y y y x x y y y x x x y y x x x x y 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 − − − + + − . (3.1) 120 120 720 180 180 720 6 Converting x →X, and y →Y so as to get a representation in terms of words in the {X,Y} alphabet, we obtain: i i Y2XYX XYXYX YXYXY YXYX2 YXY2X YX2YX z (X,Y)=− + + − − − 5 120 30 30 120 120 120 Y4X XY3X X2Y2X X3YX YXY3 YX2Y2 − + − + + − 720 180 120 180 180 120 YX3Y YX4 Y2XY2 Y2X2Y Y2X3 XYXY2 + − − − + − 180 720 120 120 180 120 XYX2Y XYX3 Y3XY Y3X2 XY2XY XY2X2 − + + + − − 120 180 180 180 120 120 X2YXY X2YX2 XY4 X2Y3 X3Y2 X4Y − − − + + − . (3.2) 120 120 720 180 180 720 Converting to right-nested commutators to get a Dynkin representation is tedious (due to antisymmetries and the Jacobi identity). Blanes and Casas [9] report the equivalent of 1 z (X,Y)= −[X4Y]+6[XYXYX]+2[XY3X]+2[YX3Y]+6[YXYXY]−[Y4X] . (3.3) 5 6! (cid:0) (cid:1) B. The z6 term For z we have: 6 y x y x y x y y y x y x y y x y x x y y x y y x y y x x y x x y x y x y 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 z =− − + + + + 6 60 360 240 240 240 60 y x y x x x y x y y x x y x x y x x y x y y y x y x x y y x y x x x y x 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 − + + − + − 360 240 240 360 240 360 x y y x y y x y y x x y x x y x y y x x y x x y y y y y x x x y y y x y 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 − − − − + + 240 240 240 240 1440 360 x x y y x y x x x y x y y y x x x x x y x y y y x y x x y y x y x x x y 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 − + + + − + 240 360 1440 360 240 360 y y y x x x x x x y y y x x x x y y x x y y y y 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 − + − − . (3.4) 360 360 1440 1440 Converting x →X and y →Y, so as to get a representation in terms of words in the {X,Y} alphabet, we obtain: i i YXYXYX Y3XYX Y2XYX2 Y2XY2X Y2X2YX XYXYXY z (X,Y)=− − + + + + 6 60 360 240 240 240 60 YXYX3 YXY2X2 YX2YZ2 YXY3X YX2Y2X YX3YX − + + − + − 360 240 240 360 240 360 XY2XY2 XY2X2Y X2YXY2 X2YX2Y Y4X2 XY3XY − − − − + + 240 240 240 240 1440 360 X2Y2XY X3YXY Y2X4 XYXY3 XYX2Y2 XYX3Y − + + + − + 240 360 1440 360 240 360 Y3X3 X3Y3 X4Y2 X2Y4 − + − − . (3.5) 360 360 1440 1440 Converting to right-nested commutators to get a Dynkin representation is tedious (due to antisymmetries and the Jacobi identity). Blanes and Casas [9] report the equivalent of 1 z (X,Y)= −2[X2Y2XY]+6[XYXYXY]−[XY4X]+[YX4Y] . (3.6) 6 2·6! (cid:0) (cid:1) 7 C. The z7 term For z we have: 7 x y x y x y x y y x y x y x y x y y x y x y x x y x y x y x y x y x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 z =− + + + − 7 140 840 840 840 140 y x y x y x x y x y x y y x y x y x x y x x x x y x y x y y y y x y x 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + + + − + 840 840 840 630 2016 x y y y x y x x x y y x y x x y y x y x y x y y x y x x x y y x y y x 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 − + + + + 630 840 840 840 840 x y y x x y x x x y x y x y x x y x y x x x x y x y y x x x y x x y x 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + + + + + 840 840 840 840 840 y y y x y x y y y y x y x x y y y x y y x y y y x x y x x y x y x y y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 − − − − + 630 5040 5040 5040 840 x y x y x x y x y x y x x x x y x y y x y x y x y y x x x y x x y x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + − + + + 840 630 840 840 840 x y x x y x x x y x y y y x x y x x y y x x y x x x y x y y x y x y y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + − + − + 840 630 840 630 840 y y x y x x y y y x y x x x y y x y y x y y y x y y x x y y x x y x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + − + − + 840 5040 840 1120 840 y y x x y x x y y x y y y x y y x x y y x y y x x x y x y x y x x x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 − − − − − 1120 5040 1120 5040 630 y x y x x x x y x y y x y y y x y y x x y y x y y x x x y x x y x y y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + + + − + 2016 840 840 5040 840 y x x y x x y y x x y x x x y x y y y x y y x y y y x x y x x y y x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + − − − + 840 5040 630 5040 840 y x x y y x x y x x x y x y y x x x y x x y x y y y y x y x x y y y x 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 − − − + − 1120 630 5040 2016 5040 y x x x y y x y x x x x y x y x y x y y y y x y x x y y x y y y y y x 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 − + − + − 5040 2016 630 840 5040 x x y y y y x x x x y y y x x x x x y y x x x x x x y x x y y y x x x 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + − + − − 2016 1512 2016 5040 1512 x x y y x y y x x y y x x y x x y y x x x x x x y x y y x x x y x x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 − − − − − 1120 1120 5040 5040 5040 x x x y x x x y y y y y x y y y y y y x x x y y y y x y x y y y y x x 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 − − − + + 1512 5040 5040 2016 2016 x x y y y x y x x y y y x x x x x y y x y x x x y y x x x x x x y x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 − − − − + 5040 5040 5040 5040 2016 x x x x y x x y y y y y y x x y x x x x x y y y x y y y y y y x x y y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + + − − − 2016 30240 5040 1512 5040 y y y x x x y y y y x x x x x y y x y y y x y y x x y y x y y x x x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 − + − − − 1512 3780 5040 1120 5040 x y y x x x x x x y x y y y x x y x x y y x x y x x x y x x y x x x x 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + − − − + 2016 5040 1120 5040 2016 y y y y x y y y y y y x x y y y y y x x x x y y y x y y x y y y x x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + + + − − 2016 2016 3780 5040 5040 y x y y y y y y x x y y y y y x x x y y y y x x x x y y y x x x x x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 − + − + − 5040 2016 1512 2016 5040 y x x x x x x y y x y y y y y y x x y y y y y x x x y y y y x x x x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + + − − + 30240 2016 5040 5040 2016 y y x x x x x x y x y y y y x y x x y y y x y x x x y y x y x x x x y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 − + − − + 5040 2016 5040 5040 2016 x y y y y y y x x y y y y y x x x y y y y x x x x y y y x x x x x y y 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 + − + + − 30240 5040 3780 3780 5040 x x x x x x y 1 2 3 4 5 6 7 + . (3.7) 30240 Thereare63occurrencesofeachx above,and63occurrencesofeachy . (Notethat63+63=126=number of terms.) i i 8 Converting x →X and y →Y so as to get a representation in terms of words in the {X,Y} alphabet we obtain: i i XYXYXYX Y2XYXYX YXY2XYX YX2YXYX YXYXYXY z7(X,Y)=− + + + − 140 840 840 840 140 YXYXYX2 YXYXY2X YXYX2YX X3YXYX Y4XYX + + + − + 840 840 840 630 2016 XY3XYX X2Y2XYX XY2XYXY XY2XYX2 XY2XY2X − + + + + 630 840 840 840 840 XY2X2YX X2YXYXY X2YXYX2 X2YXY2X X2YX2YX + + + + + 840 840 840 840 840 Y3XYXY Y3XYX2 Y3XY2X Y3X2YX XYXYXY2 − − − − + 630 5040 5040 5040 840 XYXYX2Y XYXYX3 XYXY2XY XYXY2X2 XYX2YXY + − + + + 840 630 840 840 840 XYX2YX2 XYXY3X XYX2Y2X XYX3YX Y2XYXY2 + − + − + 840 630 840 630 840 Y2XYX2Y Y2XYX3 Y2XY2XY Y2XY2X2 Y2X2YXY + − + − + 840 5040 840 1120 840 Y2X2YX2 Y2XY3X Y2X2Y2X Y2X3YX YXYX3Y − − − − − 1120 5040 1120 5040 630 YXYX4 YXY2XY2 YXY2X2Y YXY2X3 YX2YXY2 + + + − + 2016 840 840 5040 840 YX2YX2Y YX2YX3 YXY3XY YXY3X2 YX2Y2XY + − − − + 840 5040 630 5040 840 YX2Y2X2 YX3YXY YX3YX2 YXY4X YX2Y3X − − − + − 1120 630 5040 2016 5040 YX3Y2X YX4YX YXYXY3 YXYX2Y2 XY5X − + − + − 5040 2016 630 840 5040 X2Y4X X3Y3X X4Y2X X5YX XY3X3 + − + − − 2016 1512 2016 5040 1512 X2Y2XY2 X2Y2X2Y X2Y2X3 X3YXY2 X3YX2Y − − − − − 1120 1120 5040 5040 5040 X3YX3 Y5XY Y5X2 XY4XY XY4X2 − − − + + 1512 5040 5040 2016 2016 X2Y3XY X2Y3X2 X3Y2XY X3Y2X2 X4YXY − − − − + 5040 5040 5040 5040 2016 X4YX2 Y6X XYX5 Y3XY3 Y3X2Y2 + + − − − 2016 30240 5040 1512 5040 Y3X3Y Y3X4 XY2XY3 XY2X2Y2 XY2X3Y − + − − − 1512 3780 5040 1120 5040 XY2X4 X2YXY3 X2YX2Y2 X2YX3Y X2YX4 + − − − + 2016 5040 1120 5040 2016 Y4XY2 Y4X2Y Y4X3 XY3XY2 XY3X2Y + + + − − 2016 2016 3780 5040 5040 YXY5 YX2Y4 YX3Y3 YX4Y2 YX5Y − + − + − 5040 2016 1512 2016 5040 YX6 Y2XY4 Y2X2Y3 Y2X3Y2 Y2X4Y + + − − + 30240 2016 5040 5040 2016 Y2X5 XYXY4 XYX2Y3 XYX3Y2 XYX4Y − + − − + 5040 2016 5040 5040 2016 XY6 X2Y5 X3Y4 X4Y3 X5Y2 + − + + − 30240 5040 3780 3780 5040 X6Y + . (3.8) 30240 Thereare2occurrencesofX6above,5occurrencesofX5,12occurrencesofX4,28occurrencesofX3,64occurrences of X2, and144 occurrencesof X1. Similarly for Y. (Note 63×7=2×6+5×5+12×4+28×3+68×2+144×1.) Converting this to commutators in the Dynkin form is somewhat impractical without significant computer-aided computation. See for instance reference [7], and related online tables [8]. 9 D. The z8 term For z we have: 8 y y y x y x y x x y x y x y x y y y x y y x y x y y x x y x y x y y x y x y x x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 z = − − − − 8 1260 280 1680 1680 1680 y y x y x y y x y y x y x x y x y x y y x y x x y x y y x y y x y x y y x x y x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − − − − − 1680 1680 1680 1680 1680 y x y x x y x x y x y x y y y x y x y x x y y x y x y x x x y x y x y x y x x x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − + − + + 1680 1260 1680 1260 1260 y x y x y y x x y x y y y x y x y x x y y x y x y x x x y x y x y x x y x y x x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − + − + − 1680 1260 1680 1260 1680 y x x y x y y x y x x y x x y x y y y y y x y x x x y y x y x y x x x y x y x y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − − + + − 1680 1680 10080 1680 1260 y y y y x y x x y y y y x y y x y y y y x x y x x y y y x y x y x x y x x y x y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − − − − + 4032 4032 4032 1260 1680 x y y x y y x y x y y x x y x y x x y x y x y y x x y x y x x y x x y x y y x y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 + + + + + 1680 1680 1680 1680 1680 y y y x y x x x y y y x y y x x y y y x x y x x x y y x y x y y x y y x y x x y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 + + + + + 3024 10080 10080 1680 1680 y y y x y y y x y y y x x y y x y y y x x x y x x y x y x y y y x y x y x x y y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 + + + − + 3024 10080 3024 1260 1680 x y x y x x x y x y x y y x y y x y x y y x x y x y x x y x y y x y x x y x x y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − + + + + 1260 1680 1680 1680 1680 x y x y y y x y x y x x y y x y x y x x x y x y y y x y y x x x y y x x y x x x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − + − + + 1260 1680 1260 10080 10080 y y x y y y x x y y x x y y x x y y x x x y x x y y x y y y y x y y x x y y y x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 + + + − + 10080 2240 10080 4032 10080 y y x x x y y x y y x x x x y x y y x y x x x x y x y x x x x x y x y y x x x x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 + − − + − 10080 4032 4032 10080 4032 y x x y x x x x y x y y y x x x y x x y y x x x y x x x x y y x y x x x x x y x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − + + − + 4032 3024 10080 4032 10080 y x x x y x x x y x y y y y x x y x x y y y x x y x x x y y x x y x x x x y x x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 + − + + − 3024 4032 10080 10080 4032 y x y y y y y x y x x y y y y x y x x x y y y x x x y x x y y y x x y x x x y y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 + − + − − 10080 4032 3024 10080 10080 x x y y x y y y x x x y x x y y x x y y y x y y x x y y y x x y x x x y y x y y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − − − − − 10080 10080 10080 10080 10080 x x x y y x x y x x y y x x x y x x x x y y x y x x x x x y x y x x x x y x y y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − − + − + 10080 10080 4032 10080 4032 x x x x y x x y y y y y y y x x x y y y y y x y x x y y y y x y x x x y y y x y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 + − − + − 4032 60480 10080 4032 3024 23y y y y x x x x x y y y x y y y x y y y x x x y x x y y x x y y x x x y x y y y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − − − − − 120960 3024 3024 2240 3024 x x x y x x x y y y y y y x x x x y y y y x y y x y y y y x x y y y y x x x x x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − + + + + 3024 10080 4032 4032 10080 x y y x y y y y x y y x x x x y x x y x y y y y x x y x x x x y y y x x x x x x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 + + + + − 4032 4032 4032 4032 60480 x y x y y y y y x y x x y y y y x y x x x y y y x y x x x x y y x y x x x x x y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − + − + − 10080 4032 3024 4032 10080 x y y y x x y y x y y x x y y y x y y x x x y y x x y y y y y y x x x y y y y y 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 − − − + − 10080 10080 10080 60480 10080 23x x x x y y y y x x x x x y y y x x x x x x y y y x y x y x y x 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 + − + + . (3.9) 120960 10080 60480 280 Thereare62occurrencesofeachx above,and62occurrencesofeachy . (Notethat62+62=124=number of terms.) i i 10 Converting x →X and y →Y so as to get a representation in terms of words in the {X,Y} alphabet we obtain: i i Y3XYXYX XYXYXYXY Y2XY2XYX Y2X2YXYX Y2XYXYX2 z8(X,Y)= − − − − 1260 280 1680 1680 1680 Y2XYXY2X Y2XYX2YX YXY2XYX2 YXY2XY2X YXY2X2YX − − − − − 1680 1680 1680 1680 1680 YXYX2YX2 YXYXY3X YXYX2Y2X YXYX3YX YXYXYX3 − + − + + 1680 1260 1680 1260 1260 YXYXY2X2 YXY3XYX YX2Y2XYX YX3YXYX YX2YXYX2 − + − + − 1680 1260 1680 1260 1680 YX2YXY2X YX2YX2YX Y5XYX X2Y2XYXY X3YXYXY − − + + − 1680 1680 10080 1680 1260 Y4XYX2 Y4XY2X Y4X2YX XY3XYXY X2YX2YXY − − − − + 4032 4032 4032 1260 1680 XY2XY2XY XY2X2YXY X2YXYXY2 X2YXYX2Y X2YXY2XY + + + + + 1680 1680 1680 1680 1680 Y3XYX3 Y3XY2X2 Y3X2YX2 XY2XYXY2 XY2XYX2Y + + + + + 3024 10080 10080 1680 1680 Y3XY3X Y3X2Y2X Y3X3YX XYXYXY3 XYXYX2Y2 + + + − + 3024 10080 3024 1260 1680 XYXYX3Y XYXY2XY2 XYXY2X2Y XYX2YXY2 XYX2YX2Y − + + + + 1260 1680 1680 1680 1680 XYXY3XY XYX2Y2XY XYX3YXY Y2XY2X3 Y2X2YX3 − + − + + 1260 1680 1260 10080 10080 Y2XY3X2 Y2X2Y2X2 Y2X3YX2 Y2XY4X Y2X2Y3X + + + − + 10080 2240 10080 4032 10080 Y2X3Y2X Y2X4YX Y2XYX4 YXYX5 YXY2X4 + − − + − 10080 4032 4032 10080 4032 YX2YX4 YXY3X3 YX2Y2X3 YX4Y2X YX5YX − + + − + 4032 3024 10080 4032 10080 YX3YX3 YXY4X2 YX2Y3X2 YX3Y2X2 YX4YX2 + − + + − 3024 4032 10080 10080 4032 YXY5X YX2Y4X YX3Y3X X2YX2Y3 X2YX3Y2 + − + − − 10080 4032 3024 10080 10080 X2Y2XY3 X3YX2Y2 X2Y3XY2 X2Y3X2Y X3Y2XY2 − − − − − 10080 10080 10080 10080 10080 X3Y2X2Y X2Y2X3Y X4Y2XY X5YXY X4YXY2 − − + − + 10080 10080 4032 10080 4032 X4YX2Y Y6X2 XY5XY X2Y4XY X3Y3XY + − − + − 4032 60480 10080 4032 3024 23Y4X4 XY3XY3 XY3X3Y X2Y2X2Y2 X3YXY3 − − − − − 120960 3024 3024 2240 3024 X3YX3Y Y5X3 XY4XY2 XY4X2Y Y3X5 − + + + + 3024 10080 4032 4032 10080 XY2XY4 XY2X4Y X2YXY4 X2YX4Y Y2X6 + + + + − 4032 4032 4032 4032 60480 XYXY5 XYX2Y4 XYX3Y3 XYX4Y2 XYX5Y − + − + − 10080 4032 3024 4032 10080 XY3X2Y2 XY2X2Y3 XY2X3Y2 X2Y6 X3Y5 − − − + − 10080 10080 10080 60480 10080 23X4Y4 X5Y3 X6Y2 YXYXYXYX + − + + (3.10) 120960 10080 60480 280 Thereare2occurrencesofX6above,6occurrencesofX5,14occurrencesofX4,32occurrencesofX3,72occurrences of X2, and158 occurrencesof X1. Similarly for Y. (Note 62×8=2×6+6×5+14×4+32×3+72×2+158×1.) Converting this to commutators in the Dynkin form is somewhat impractical without significant computer-aided computation. See for instance reference [7], and related online tables [8].

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.