Structure in Loss of Orthogonality $

Total Page:16

File Type:pdf, Size:1020Kb

Structure in Loss of Orthogonality $ Structure in loss of orthogonality I Xiao-Wen Changa, Christopher C. Paigea,∗, David Titley-Peloquinb aSchool of Computer Science, McGill University, Montr´eal,Qu´ebec, Canada bDepartment of Bioresource Engineering, McGill University, Ste-Anne-de-Bellevue, Qu´ebec, Canada Abstract In [SIAM J. Matrix Anal. Appl., 31 (2009), pp. 565{583] it was shown that for any sequence of k unit 2-norm n-vectors, the columns of Vk, there is a special (n+k)-square unitary matrix Q(k) that can be used in the analysis of numerical algorithms based on (k) orthogonality. A k × k submatrix Sk of Q provides valuable theoretical information on the loss of orthogonality among the columns of Vk. Here it is shown that the singular value decomposition (SVD) and Jordan canonical form (JCF) of Sk both reveal the null space of Vk as well as orthonormal vectors available from a right-side orthogonal transformation of Vk. The JCF of Sk is shown to reveal more than its SVD does. The Lanczos orthogonal tridiagonalization process for a Hermitian matrix is then used to indicate the occurrence of some of these properties in practical computations. Keywords: Loss of orthogonality, singular value decomposition, Jordan canonical form, rounding error analysis, Lanczos process, eigenproblem. 2000 MSC: 65F15, 65F25, 65G50, 15A18 1. Introduction n×k If Vk 2 C has unit 2-norm columns, one can define the strictly upper triangular 4 −1 H matrix Sk = (I + Uk) Uk, where Uk is the strictly upper triangular part of Vk Vk, as well as the unitary matrix S (I −S )V H (k) 4 k k k k Q = H ; (1) Vk(Ik −Sk) In −Vk(Ik −Sk)Vk see Theorem 1 below. This Q(k) was described in [13] and can be the basis of the rounding error analysis of several numerical algorithms based on orthogonality, see, e.g., [14, 15]. But more generally the matrix Sk provides valuable theoretical information on the loss of orthogonality among the columns of any such Vk. Here properties of Sk are developed in general, and used to show the various properties of Vk that can occur. In particular, IWith best wishes to Paul Van Dooren, one of the brightest and most likeable of people. ∗Corresponding author Email addresses: [email protected] (Xiao-Wen Chang), [email protected] (Christopher C. Paige), [email protected] (David Titley-Peloquin) Preprint submitted to Linear Algebra and Its Applications July 7, 2020 we show that the Jordan canonical form (JCF) of Sk can reveal important properties that are not available from the singular value decomposition (SVD) of Sk. The paper is organized as follows. In the next two sections we give a very brief history followed by the notation used here. Section 4 summarizes the theorem on unitary Q(k) in (1), while section 5 derives some properties of Q(k) that we need. Section 6 deals with the SVD of Sk, and shows how it defines important subspaces related to Vk. Section 7 introduces the JCF of Sk, then section 8 shows how this reveals more properties of Vk. These are new results for general Vk with unit length columns, so proofs are given in these sections 7 & 8. Section 9 summarizes the Lanczos process and the result of its rounding error analysis in [14], which shows that the finite precision Lanczos process behaves as a higher dimensional exact Lanczos process for a slightly perturbed (k + n) × (k + n) matrix Ak. Section 10 states a theorem on how the Lanczos process converges, and then uses the JCF of Sk to reveal some surprising numerical behaviors of the Lanczos process and therefore of some other numerical iterative orthogonalization algorithms. 2. A very brief history of the Lanczos process and orthogonalization Although the orthogonal tridiagonalization of a Hermitian matrix A devised by Cor- nelius Lanczos [8] is simple and elegant mathematically, its numerical behavior has fasci- nated many for 70 years. The Lanczos process was originally discarded because of its loss of orthogonality, then brought back in importance and very gradually understood. There have been many useful works on this resuscitation, such as [11, 12, 20, 9, 10, 14, 15]. The ideas behind the Lanczos process led to other valuable algorithms such as in [4, 16, 2], and there has also been work on the sensitivity of the tridiagonal matrix and vectors resulting from the Lanczos process to perturbations in A, see for example [18]. But an understanding of the loss of orthogonality of the Lanczos process turned out to be crucial. A breakthrough in our understanding of loss of orthogonality in general was initiated by a comment by Charles Sheffield [21] to Gene Golub, which Gene related to Ake˚ Bj¨orck and Chris Paige around 1990, see [1]. This concerned the loss of orthogonality in modified Gram-Schmidt (MGS), but it was shown in [13] that it could be extended to apply to any sequence of unit-length vectors vj. A more complete background of this is given in [13, Section 2.2]. This approach was applied in [14] to give an augmented backward stability result for the Hermitian matrix Lanczos process [8], and this was used in [15] to prove the iterative convergence of the Lanczos process for the eigenproblem and solution of equations, along with more history in [15, Section 2]. Here we look more deeply into the properties of Sk in (1) and what it tells us about loss of orthogonality in general. 3. Notation 4 We use \=" for \is defined to be", and \≡" for \is equivalent to". Let In denote the n×k n × n unit matrix, with j-th column ej. We say Q1 2 C has orthonormal columns H n×k if Q1 Q1p= Ik and write Q1 2 U . For a vector v, we denote its Euclidean norm by 4 H n×m kvk2 = v v. For a matrix B = [b1; b2; : : : ; bm] 2 C we denote its Frobenius norm 4 by kBkF , its spectral norm by kBk2 = σmax(B) the maximum singular value of B, and its range by Range(B). For indices, i:j means i; i+1; : : : ; j, while Bi:j ≡ [bi; bi+1; : : : ; bj]. 2 We will be dealing with sequences of matrices of increasing dimensions, and will use the index k to denote the k-th matrix in a sequence, usually as for example Q(k), in (k) (k) (k) which case subscripts denote partitioning, as in Q ≡ [Q1 j Q2 ]. We often omit the particular superscript ·(k) when the meaning is clear. However there are five special matrices where we denote the k-th matrix by a subscript: Vk, Uk, Sk, Tk, and Ak. For these the (k + 1)-st matrix can be obtained from the k-th by adding a column, e.g., Vk+1 = [Vk; vk+1], or a column and a row, and there is no need for further subscripts. This makes their presentation and manipulation easier to understand in formulae. 4. Obtaining a unitary matrix from unit-length n-vectors The next theorem was given in full with proofs in [13]. It allows us to develop a (k) (k +n)×(k +n) unitary matrix Q from any n×k matrix Vk with unit-length columns. n Theorem 1 ([13, Theorem 2.1]). For integers n ≥ 1 and k ≥ 1 suppose that vj 2 C satisfies kvjk2 = 1; j =1:k+1, and Vk = [v1; : : : ; vk]. If Uk is the strictly upper triangular H H matrix satisfying Vk Vk = I + Uk + Uk , define the strictly upper triangular matrix Sk via 4 −1 −1 k×k Sk = (Ik + Uk) Uk = Uk(Ik + Uk) 2 C : (2) Then H H kSkk2 ≤ 1; Vk Vk = I , kSkk2 = 0; Vk Vk singular , kSkk2 = 1: (3) Here Sk is the unique strictly upper triangular k × k matrix such that " # h i (k) (k) S (I −S )V H (k) (k) (k) Q11 Q12 4 k k k k (k+n)×(k+n) Q ≡ Q1 Q2 ≡ (k) (k) = H 2U : Q Q Vk(Ik −Sk) In −Vk(Ik −Sk)V k n 21 22 k (4) Finally Sk and Sk+1 have the following relations S s S ≡ k k+1 2 (k+1)×(k+1); s =(I −S )V H v : (5) k+1 0 0 C k+1 k k k k+1 Here is an indication of a proof. From (2) it can be shown that −1 −1 −1 UkSk = SkUk;Uk =(Ik −Sk) Sk ≡Sk(Ik −Sk) ; (Ik −Sk) = Ik +Uk: (6) (k)H (k) (k) To prove Q1 Q1 = Ik in (4), use (2) and (6) to give (dropping · and ·k): S U(I + U)−1 U Q ≡ = = (I + U)−1; 1 V (I − S) V (I + U)−1 V H −H H H −1 −H H H −1 Q1 Q1 = (I +U) [V V +U U](I +U) = (I +U) [I +U +U +U U](I +U) = (I + U)−H [(I + U)H (I + U)](I + U)−1 = I: This was given in [15, §4]. Next, for example, kSkk2 ≤ 1 in (3) follows immediately. Finally, the first equation in (5) follows from the definition of Sk, and to prove the second equation in (5) we see from (6) that Sk+1=(Ik+1 −Sk+1)Uk+1, so that s V H v (I −S )V H v k+1 =S e =(I −S )U e =(I −S ) k k+1 = k k k k+1 . 0 k+1 k+1 k+1 k+1 k+1 k+1 k+1 k+1 0 0 3 5.
Recommended publications
  • A New Description of Space and Time Using Clifford Multivectors
    A new description of space and time using Clifford multivectors James M. Chappell† , Nicolangelo Iannella† , Azhar Iqbal† , Mark Chappell‡ , Derek Abbott† †School of Electrical and Electronic Engineering, University of Adelaide, South Australia 5005, Australia ‡Griffith Institute, Griffith University, Queensland 4122, Australia Abstract Following the development of the special theory of relativity in 1905, Minkowski pro- posed a unified space and time structure consisting of three space dimensions and one time dimension, with relativistic effects then being natural consequences of this space- time geometry. In this paper, we illustrate how Clifford’s geometric algebra that utilizes multivectors to represent spacetime, provides an elegant mathematical framework for the study of relativistic phenomena. We show, with several examples, how the application of geometric algebra leads to the correct relativistic description of the physical phenomena being considered. This approach not only provides a compact mathematical representa- tion to tackle such phenomena, but also suggests some novel insights into the nature of time. Keywords: Geometric algebra, Clifford space, Spacetime, Multivectors, Algebraic framework 1. Introduction The physical world, based on early investigations, was deemed to possess three inde- pendent freedoms of translation, referred to as the three dimensions of space. This naive conclusion is also supported by more sophisticated analysis such as the existence of only five regular polyhedra and the inverse square force laws. If we lived in a world with four spatial dimensions, for example, we would be able to construct six regular solids, and in arXiv:1205.5195v2 [math-ph] 11 Oct 2012 five dimensions and above we would find only three [1].
    [Show full text]
  • Lecture 4: April 8, 2021 1 Orthogonality and Orthonormality
    Mathematical Toolkit Spring 2021 Lecture 4: April 8, 2021 Lecturer: Avrim Blum (notes based on notes from Madhur Tulsiani) 1 Orthogonality and orthonormality Definition 1.1 Two vectors u, v in an inner product space are said to be orthogonal if hu, vi = 0. A set of vectors S ⊆ V is said to consist of mutually orthogonal vectors if hu, vi = 0 for all u 6= v, u, v 2 S. A set of S ⊆ V is said to be orthonormal if hu, vi = 0 for all u 6= v, u, v 2 S and kuk = 1 for all u 2 S. Proposition 1.2 A set S ⊆ V n f0V g consisting of mutually orthogonal vectors is linearly inde- pendent. Proposition 1.3 (Gram-Schmidt orthogonalization) Given a finite set fv1,..., vng of linearly independent vectors, there exists a set of orthonormal vectors fw1,..., wng such that Span (fw1,..., wng) = Span (fv1,..., vng) . Proof: By induction. The case with one vector is trivial. Given the statement for k vectors and orthonormal fw1,..., wkg such that Span (fw1,..., wkg) = Span (fv1,..., vkg) , define k u + u = v − hw , v i · w and w = k 1 . k+1 k+1 ∑ i k+1 i k+1 k k i=1 uk+1 We can now check that the set fw1,..., wk+1g satisfies the required conditions. Unit length is clear, so let’s check orthogonality: k uk+1, wj = vk+1, wj − ∑ hwi, vk+1i · wi, wj = vk+1, wj − wj, vk+1 = 0. i=1 Corollary 1.4 Every finite dimensional inner product space has an orthonormal basis.
    [Show full text]
  • Orthogonality Handout
    3.8 (SUPPLEMENT) | ORTHOGONALITY OF EIGENFUNCTIONS We now develop some properties of eigenfunctions, to be used in Chapter 9 for Fourier Series and Partial Differential Equations. 1. Definition of Orthogonality R b We say functions f(x) and g(x) are orthogonal on a < x < b if a f(x)g(x) dx = 0 . [Motivation: Let's approximate the integral with a Riemann sum, as follows. Take a large integer N, put h = (b − a)=N and partition the interval a < x < b by defining x1 = a + h; x2 = a + 2h; : : : ; xN = a + Nh = b. Then Z b f(x)g(x) dx ≈ f(x1)g(x1)h + ··· + f(xN )g(xN )h a = (uN · vN )h where uN = (f(x1); : : : ; f(xN )) and vN = (g(x1); : : : ; g(xN )) are vectors containing the values of f and g. The vectors uN and vN are said to be orthogonal (or perpendicular) if their dot product equals zero (uN ·vN = 0), and so when we let N ! 1 in the above formula it makes R b sense to say the functions f and g are orthogonal when the integral a f(x)g(x) dx equals zero.] R π 1 2 π Example. sin x and cos x are orthogonal on −π < x < π, since −π sin x cos x dx = 2 sin x −π = 0. 2. Integration Lemma Suppose functions Xn(x) and Xm(x) satisfy the differential equations 00 Xn + λnXn = 0; a < x < b; 00 Xm + λmXm = 0; a < x < b; for some numbers λn; λm. Then Z b 0 0 b (λn − λm) Xn(x)Xm(x) dx = [Xn(x)Xm(x) − Xn(x)Xm(x)]a: a Proof.
    [Show full text]
  • Inner Products and Orthogonality
    Advanced Linear Algebra – Week 5 Inner Products and Orthogonality This week we will learn about: • Inner products (and the dot product again), • The norm induced by the inner product, • The Cauchy–Schwarz and triangle inequalities, and • Orthogonality. Extra reading and watching: • Sections 1.3.4 and 1.4.1 in the textbook • Lecture videos 17, 18, 19, 20, 21, and 22 on YouTube • Inner product space at Wikipedia • Cauchy–Schwarz inequality at Wikipedia • Gram–Schmidt process at Wikipedia Extra textbook problems: ? 1.3.3, 1.3.4, 1.4.1 ?? 1.3.9, 1.3.10, 1.3.12, 1.3.13, 1.4.2, 1.4.5(a,d) ??? 1.3.11, 1.3.14, 1.3.15, 1.3.25, 1.4.16 A 1.3.18 1 Advanced Linear Algebra – Week 5 2 There are many times when we would like to be able to talk about the angle between vectors in a vector space V, and in particular orthogonality of vectors, just like we did in Rn in the previous course. This requires us to have a generalization of the dot product to arbitrary vector spaces. Definition 5.1 — Inner Product Suppose that F = R or F = C, and V is a vector space over F. Then an inner product on V is a function h·, ·i : V × V → F such that the following three properties hold for all c ∈ F and all v, w, x ∈ V: a) hv, wi = hw, vi (conjugate symmetry) b) hv, w + cxi = hv, wi + chv, xi (linearity in 2nd entry) c) hv, vi ≥ 0, with equality if and only if v = 0.
    [Show full text]
  • Eigenvalues, Eigenvectors and the Similarity Transformation
    Eigenvalues, Eigenvectors and the Similarity Transformation Eigenvalues and the associated eigenvectors are ‘special’ properties of square matrices. While the eigenvalues parameterize the dynamical properties of the system (timescales, resonance properties, amplification factors, etc) the eigenvectors define the vector coordinates of the normal modes of the system. Each eigenvector is associated with a particular eigenvalue. The general state of the system can be expressed as a linear combination of eigenvectors. The beauty of eigenvectors is that (for square symmetric matrices) they can be made orthogonal (decoupled from one another). The normal modes can be handled independently and an orthogonal expansion of the system is possible. The decoupling is also apparent in the ability of the eigenvectors to diagonalize the original matrix, A, with the eigenvalues lying on the diagonal of the new matrix, . In analogy to the inertia tensor in mechanics, the eigenvectors form the principle axes of the solid object and a similarity transformation rotates the coordinate system into alignment with the principle axes. Motion along the principle axes is decoupled. The matrix mechanics is closely related to the more general singular value decomposition. We will use the basis sets of orthogonal eigenvectors generated by SVD for orbit control problems. Here we develop eigenvector theory since it is more familiar to most readers. Square matrices have an eigenvalue/eigenvector equation with solutions that are the eigenvectors xand the associated eigenvalues : Ax = x The special property of an eigenvector is that it transforms into a scaled version of itself under the operation of A. Note that the eigenvector equation is non-linear in both the eigenvalue () and the eigenvector (x).
    [Show full text]
  • A Guided Tour to the Plane-Based Geometric Algebra PGA
    A Guided Tour to the Plane-Based Geometric Algebra PGA Leo Dorst University of Amsterdam Version 1.15{ July 6, 2020 Planes are the primitive elements for the constructions of objects and oper- ators in Euclidean geometry. Triangulated meshes are built from them, and reflections in multiple planes are a mathematically pure way to construct Euclidean motions. A geometric algebra based on planes is therefore a natural choice to unify objects and operators for Euclidean geometry. The usual claims of `com- pleteness' of the GA approach leads us to hope that it might contain, in a single framework, all representations ever designed for Euclidean geometry - including normal vectors, directions as points at infinity, Pl¨ucker coordinates for lines, quaternions as 3D rotations around the origin, and dual quaternions for rigid body motions; and even spinors. This text provides a guided tour to this algebra of planes PGA. It indeed shows how all such computationally efficient methods are incorporated and related. We will see how the PGA elements naturally group into blocks of four coordinates in an implementation, and how this more complete under- standing of the embedding suggests some handy choices to avoid extraneous computations. In the unified PGA framework, one never switches between efficient representations for subtasks, and this obviously saves any time spent on data conversions. Relative to other treatments of PGA, this text is rather light on the mathematics. Where you see careful derivations, they involve the aspects of orientation and magnitude. These features have been neglected by authors focussing on the mathematical beauty of the projective nature of the algebra.
    [Show full text]
  • Inner Product Spaces
    CHAPTER 6 Woman teaching geometry, from a fourteenth-century edition of Euclid’s geometry book. Inner Product Spaces In making the definition of a vector space, we generalized the linear structure (addition and scalar multiplication) of R2 and R3. We ignored other important features, such as the notions of length and angle. These ideas are embedded in the concept we now investigate, inner products. Our standing assumptions are as follows: 6.1 Notation F, V F denotes R or C. V denotes a vector space over F. LEARNING OBJECTIVES FOR THIS CHAPTER Cauchy–Schwarz Inequality Gram–Schmidt Procedure linear functionals on inner product spaces calculating minimum distance to a subspace Linear Algebra Done Right, third edition, by Sheldon Axler 164 CHAPTER 6 Inner Product Spaces 6.A Inner Products and Norms Inner Products To motivate the concept of inner prod- 2 3 x1 , x 2 uct, think of vectors in R and R as x arrows with initial point at the origin. x R2 R3 H L The length of a vector in or is called the norm of x, denoted x . 2 k k Thus for x .x1; x2/ R , we have The length of this vector x is p D2 2 2 x x1 x2 . p 2 2 x1 x2 . k k D C 3 C Similarly, if x .x1; x2; x3/ R , p 2D 2 2 2 then x x1 x2 x3 . k k D C C Even though we cannot draw pictures in higher dimensions, the gener- n n alization to R is obvious: we define the norm of x .x1; : : : ; xn/ R D 2 by p 2 2 x x1 xn : k k D C C The norm is not linear on Rn.
    [Show full text]
  • Math 480 Notes on Orthogonality the Word Orthogonal Is a Synonym for Perpendicular. Question 1: When Are Two Vectors V 1 and V2
    Math 480 Notes on Orthogonality The word orthogonal is a synonym for perpendicular. n Question 1: When are two vectors ~v1 and ~v2 in R orthogonal to one another? The most basic answer is \if the angle between them is 90◦" but this is not very practical. How could you tell whether the vectors 0 1 1 0 1 1 @ 1 A and @ 3 A 1 1 are at 90◦ from one another? One way to think about this is as follows: ~v1 and ~v2 are orthogonal if and only if the triangle formed by ~v1, ~v2, and ~v1 − ~v2 (drawn with its tail at ~v2 and its head at ~v1) is a right triangle. The Pythagorean Theorem then tells us that this triangle is a right triangle if and only if 2 2 2 (1) jj~v1jj + jj~v2jj = jj~v1 − ~v2jj ; where jj − jj denotes the length of a vector. 0 x1 1 . The length of a vector ~x = @ . A is easy to measure: the Pythagorean Theorem (once again) xn tells us that q 2 2 jj~xjj = x1 + ··· + xn: This expression under the square root is simply the matrix product 0 x1 1 T . ~x ~x = (x1 ··· xn) @ . A : xn Definition. The inner product (also called the dot product) of two vectors ~x;~y 2 Rn, written h~x;~yi or ~x · ~y, is defined by n T X hx; yi = ~x ~y = xiyi: i=1 Since matrix multiplication is linear, inner products satisfy h~x;~y1 + ~y2i = h~x;~y1i + h~x;~y2i h~x1; a~yi = ah~x1; ~yi: (Similar formulas hold in the first coordinate, since h~x;~yi = h~y; ~xi.) Now we can write 2 2 jj~v1 − ~v2jj = h~v1 − ~v2;~v1 − ~v2i = h~v1;~v1i − 2h~v1;~v2i + h~v2;~v2i = jj~v1jj − 2h~v1;~v2i + jj~v2jj; so Equation (1) holds if and only if h~v1;~v2i = 0: n Answer to Question 1: Vectors ~v1 and ~v2 in R are orthogonal if and only if h~v1;~v2i = 0.
    [Show full text]
  • Inner Product Spaces Isaiah Lankham, Bruno Nachtergaele, Anne Schilling (March 2, 2007)
    MAT067 University of California, Davis Winter 2007 Inner Product Spaces Isaiah Lankham, Bruno Nachtergaele, Anne Schilling (March 2, 2007) The abstract definition of vector spaces only takes into account algebraic properties for the addition and scalar multiplication of vectors. For vectors in Rn, for example, we also have geometric intuition which involves the length of vectors or angles between vectors. In this section we discuss inner product spaces, which are vector spaces with an inner product defined on them, which allow us to introduce the notion of length (or norm) of vectors and concepts such as orthogonality. 1 Inner product In this section V is a finite-dimensional, nonzero vector space over F. Definition 1. An inner product on V is a map ·, · : V × V → F (u, v) →u, v with the following properties: 1. Linearity in first slot: u + v, w = u, w + v, w for all u, v, w ∈ V and au, v = au, v; 2. Positivity: v, v≥0 for all v ∈ V ; 3. Positive definiteness: v, v =0ifandonlyifv =0; 4. Conjugate symmetry: u, v = v, u for all u, v ∈ V . Remark 1. Recall that every real number x ∈ R equals its complex conjugate. Hence for real vector spaces the condition about conjugate symmetry becomes symmetry. Definition 2. An inner product space is a vector space over F together with an inner product ·, ·. Copyright c 2007 by the authors. These lecture notes may be reproduced in their entirety for non- commercial purposes. 2NORMS 2 Example 1. V = Fn n u =(u1,...,un),v =(v1,...,vn) ∈ F Then n u, v = uivi.
    [Show full text]
  • A Feasible Method for Optimization with Orthogonality Constraints
    A FEASIBLE METHOD FOR OPTIMIZATION WITH ORTHOGONALITY CONSTRAINTS ZAIWEN WEN y AND WOTAO YIN z November 7, 2010 > Abstract. Minimization with orthogonality constraints (e.g., X X = I) and/or spherical constraints (e.g., kxk2 = 1) has wide applications in polynomial optimization, combinatorial optimization, eigenvalue problems, sparse PCA, p-harmonic flows, 1-bit compressive sensing, matrix rank minimization, etc. These problems are difficult because the constraints are not only non-convex but numerically expensive to preserve during iterations. To deal with these difficulties, we propose to use a Crank-Nicolson-like update scheme to preserve the constraints and based on it, develop curvilinear search algorithms with lower per-iteration cost compared to those based on projections and geodesics. The efficiency of the proposed algorithms is demonstrated on a variety of test problems. In particular, for the maxcut problem, it exactly solves a decomposition formulation for the SDP relaxation. For polynomial optimization, nearest correlation matrix estimation and extreme eigenvalue problems, the proposed algorithms run very fast and return solutions no worse than those from their state-of-the-art algorithms. For the quadratic assignment problem, a gap 0.842% to the best known solution on the largest problem \tai256c" in QAPLIB can be reached in 5 minutes on a typical laptop. Key words. Orthogonality constraint, spherical constraint, Stiefel manifold, Cayley transformation, curvilinear search, polynomial optimization, maxcut SDP, nearest correlation matrix, eigenvalue and eigenvector, invariant subspace, quadratic assignment problem AMS subject classifications. 49Q99, 65K05, 90C22, 90C26, 90C27, 90C30 1. Introduction. Matrix orthogonality constraints play an important role in many applications of science and engineering.
    [Show full text]
  • Clifford Algebra with Mathematica
    Clifford Algebra with Mathematica J.L. ARAGON´ G. ARAGON-CAMARASA Universidad Nacional Aut´onoma de M´exico University of Glasgow Centro de F´ısica Aplicada School of Computing Science y Tecnolog´ıa Avanzada Sir Alwyn William Building, Apartado Postal 1-1010, 76000 Quer´etaro Glasgow, G12 8QQ Scotland MEXICO UNITED KINGDOM [email protected] [email protected] G. ARAGON-GONZ´ ALEZ´ M.A. RODRIGUEZ-ANDRADE´ Universidad Aut´onoma Metropolitana Instituto Polit´ecnico Nacional Unidad Azcapotzalco Departamento de Matem´aticas, ESFM San Pablo 180, Colonia Reynosa-Tamaulipas, UP Adolfo L´opez Mateos, 02200 D.F. M´exico Edificio 9. 07300 D.F. M´exico MEXICO MEXICO [email protected] [email protected] Abstract: The Clifford algebra of a n-dimensional Euclidean vector space provides a general language comprising vectors, complex numbers, quaternions, Grassman algebra, Pauli and Dirac matrices. In this work, we present an introduction to the main ideas of Clifford algebra, with the main goal to develop a package for Clifford algebra calculations for the computer algebra program Mathematica.∗ The Clifford algebra package is thus a powerful tool since it allows the manipulation of all Clifford mathematical objects. The package also provides a visualization tool for elements of Clifford Algebra in the 3-dimensional space. clifford.m is available from https://github.com/jlaragonvera/Geometric-Algebra Key–Words: Clifford Algebras, Geometric Algebra, Mathematica Software. 1 Introduction Mathematica, resulting in a package for doing Clif- ford algebra computations. There exists some other The importance of Clifford algebra was recognized packages and specialized programs for doing Clif- for the first time in quantum field theory.
    [Show full text]
  • Linear Algebra Chapter 5: Norms, Inner Products and Orthogonality
    MATH 532: Linear Algebra Chapter 5: Norms, Inner Products and Orthogonality Greg Fasshauer Department of Applied Mathematics Illinois Institute of Technology Spring 2015 [email protected] MATH 532 1 Outline 1 Vector Norms 2 Matrix Norms 3 Inner Product Spaces 4 Orthogonal Vectors 5 Gram–Schmidt Orthogonalization & QR Factorization 6 Unitary and Orthogonal Matrices 7 Orthogonal Reduction 8 Complementary Subspaces 9 Orthogonal Decomposition 10 Singular Value Decomposition 11 Orthogonal Projections [email protected] MATH 532 2 Vector Norms Vector[0] Norms 1 Vector Norms 2 Matrix Norms Definition 3 Inner Product Spaces Let x; y 2 Rn (Cn). Then 4 Orthogonal Vectors n T X 5 Gram–Schmidt Orthogonalization & QRx Factorizationy = xi yi 2 R i=1 6 Unitary and Orthogonal Matrices n X ∗ = ¯ 2 7 Orthogonal Reduction x y xi yi C i=1 8 Complementary Subspaces is called the standard inner product for Rn (Cn). 9 Orthogonal Decomposition 10 Singular Value Decomposition 11 Orthogonal Projections [email protected] MATH 532 4 Vector Norms Definition Let V be a vector space. A function k · k : V! R≥0 is called a norm provided for any x; y 2 V and α 2 R 1 kxk ≥ 0 and kxk = 0 if and only if x = 0, 2 kαxk = jαj kxk, 3 kx + yk ≤ kxk + kyk. Remark The inequality in (3) is known as the triangle inequality. [email protected] MATH 532 5 Vector Norms Remark Any inner product h·; ·i induces a norm via (more later) p kxk = hx; xi: We will show that the standard inner product induces the Euclidean norm (cf.
    [Show full text]