DOCSLIB.ORG

Alternating Sign Matrices and Latin Squares

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Alternating Sign Matrices and Latin Squares Alternating Sign Matrices and Latin Squares Cian O'Brien Rachel Quinlan and Kevin Jennings National University of Ireland, Galway [email protected] Postgraduate Modelling Group, NUI Galway October 4th, 2019 Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 1 / 9 20 1 0 03 61 −1 1 07 6 7 40 1 −1 15 0 0 1 0 They are a generalisation of the permutation matrices. 2+ 0 0 0 3 6 0 0 0 +7 6 7 4 0 + 0 0 5 0 0 + 0 Alternating Sign Matrices An Alternating Sign Matrix (ASM) is (0; 1; −1)-matrix for which all row and column sums are 1, and the non-zero elements in each row and column alternate in sign. Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 2 / 9 They are a generalisation of the permutation matrices. 2+ 0 0 0 3 6 0 0 0 +7 6 7 4 0 + 0 0 5 0 0 + 0 Alternating Sign Matrices An Alternating Sign Matrix (ASM) is (0; 1; −1)-matrix for which all row and column sums are 1, and the non-zero elements in each row and column alternate in sign. 20 1 0 03 61 −1 1 07 6 7 40 1 −1 15 0 0 1 0 Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 2 / 9 They are a generalisation of the permutation matrices. 2+ 0 0 0 3 6 0 0 0 +7 6 7 4 0 + 0 0 5 0 0 + 0 Alternating Sign Matrices An Alternating Sign Matrix (ASM) is (0; 1; −1)-matrix for which all row and column sums are 1, and the non-zero elements in each row and column alternate in sign. 2 0 + 0 0 3 6+ − + 0 7 6 7 4 0 + − +5 0 0 + 0 Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 2 / 9 2+ 0 0 0 3 6 0 0 0 +7 6 7 4 0 + 0 0 5 0 0 + 0 Alternating Sign Matrices An Alternating Sign Matrix (ASM) is (0; 1; −1)-matrix for which all row and column sums are 1, and the non-zero elements in each row and column alternate in sign. 2 0 + 0 0 3 6+ − + 0 7 6 7 4 0 + − +5 0 0 + 0 They are a generalisation of the permutation matrices. Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 2 / 9 Alternating Sign Matrices An Alternating Sign Matrix (ASM) is (0; 1; −1)-matrix for which all row and column sums are 1, and the non-zero elements in each row and column alternate in sign. 2 0 + 0 0 3 6+ − + 0 7 6 7 4 0 + − +5 0 0 + 0 They are a generalisation of the permutation matrices. 2+ 0 0 0 3 6 0 0 0 +7 6 7 4 0 + 0 0 5 0 0 + 0 Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 2 / 9 2+ 0 0 3 2 0 + 0 3 2 0 0 +3 4 0 + 0 5 % 4+ − +5 % 4 0 + 0 5 0 0 + 0 + 0 + 0 0 Alternating Sign Hypermatrices An Alternating Sign Hypermatrix (ASHM) is a hypermatrix (n × n × n array) for which every plane is an ASM. Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 3 / 9 2+ 0 0 3 2 0 + 0 3 2 0 0 +3 4 0 + 0 5 % 4+ − +5 % 4 0 + 0 5 0 0 + 0 + 0 + 0 0 Alternating Sign Hypermatrices An Alternating Sign Hypermatrix (ASHM) is a hypermatrix (n × n × n array) for which every plane is an ASM. Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 3 / 9 Alternating Sign Hypermatrices An Alternating Sign Hypermatrix (ASHM) is a hypermatrix (n × n × n array) for which every plane is an ASM. 2+ 0 0 3 2 0 + 0 3 2 0 0 +3 4 0 + 0 5 % 4+ − +5 % 4 0 + 0 5 0 0 + 0 + 0 + 0 0 Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 3 / 9 2 0 0 + 0 0 3 6 0 + − + 0 7 6 7 6+ − + − +7 6 7 4 0 + − + 0 5 0 0 + 0 0 Non-Zero Entry Bounds for ASMs The number of non-zero entries in the rows/columns of an ASM is bounded above by (1; 3; 5;:::; 5; 3; 1). Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 4 / 9 Non-Zero Entry Bounds for ASMs The number of non-zero entries in the rows/columns of an ASM is bounded above by (1; 3; 5;:::; 5; 3; 1). 2 0 0 + 0 0 3 6 0 + − + 0 7 6 7 6+ − + − +7 6 7 4 0 + − + 0 5 0 0 + 0 0 Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 4 / 9 + 0 0 0 0 ! 0 + 0 0 0 ! 0 0 + 0 0 ! 0 0 0 + 0 ! 0 0 0 0 +! 0 + 0 0 0 + − + 0 0 0 + − + 0 0 0 + − + 0 0 0 + 0 0 0 + 0 0 % 0 + − + 0 % + − + − + % 0 + − + 0 % 0 0 + 0 0 0 0 0 + 0 0 0 + − + 0 + − + 0 + − + 0 0 0 + 0 0 0 0 0 0 0 + 0 0 0 + 0 0 0 + 0 0 0 + 0 0 0 + 0 0 0 0 Non-Zero Entry Bounds for ASHMs The number of non-zero entries in the planes of an ASHM is bounded above by 21 1 ··· 1 13 61 3 ··· 3 17 6 7 6. .. .7 6. .7 6 7 41 3 ··· 3 15 1 1 ··· 1 1 Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 5 / 9 Non-Zero Entry Bounds for ASHMs The number of non-zero entries in the planes of an ASHM is bounded above by 21 1 ··· 1 13 61 3 ··· 3 17 6 7 6. .. .7 6. .7 6 7 41 3 ··· 3 15 1 1 ··· 1 1 + 0 0 0 0 ! 0 + 0 0 0 ! 0 0 + 0 0 ! 0 0 0 + 0 ! 0 0 0 0 +! 0 + 0 0 0 + − + 0 0 0 + − + 0 0 0 + − + 0 0 0 + 0 0 0 + 0 0 % 0 + − + 0 % + − + − + % 0 + − + 0 % 0 0 + 0 0 0 0 0 + 0 0 0 + − + 0 + − + 0 + − + 0 0 0 + 0 0 0 0 0 0 0 + 0 0 0 + 0 0 0 + 0 0 0 + 0 0 0 + 0 0 0 0 Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 5 / 9 21 2 33 42 3 15 3 1 2 Each n × n latin square can be decomposed uniquely into a sum of scalar multiples of mutually orthogonal permutation matrices. Therefore each latin square corresponds uniquely to a permutation hypermatrix. For a permutation hypermatrix M, define L(M) to be Pn L(M)i;j;k = k=1 k × Mi;j;k . 2+ 0 0 3 2 0 + 0 3 2 0 0 +3 4 0 0 +5 % 4+ 0 0 5 % 4 0 + 0 5 0 + 0 0 0 + + 0 0 Latin Squares An n × n Latin Square is an n × n array of n symbols such that each symbol occurs exactly once in each row and column. Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 6 / 9 Each n × n latin square can be decomposed uniquely into a sum of scalar multiples of mutually orthogonal permutation matrices. Therefore each latin square corresponds uniquely to a permutation hypermatrix. For a permutation hypermatrix M, define L(M) to be Pn L(M)i;j;k = k=1 k × Mi;j;k . 2+ 0 0 3 2 0 + 0 3 2 0 0 +3 4 0 0 +5 % 4+ 0 0 5 % 4 0 + 0 5 0 + 0 0 0 + + 0 0 Latin Squares An n × n Latin Square is an n × n array of n symbols such that each symbol occurs exactly once in each row and column. 21 2 33 42 3 15 3 1 2 Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 6 / 9 Therefore each latin square corresponds uniquely to a permutation hypermatrix. For a permutation hypermatrix M, define L(M) to be Pn L(M)i;j;k = k=1 k × Mi;j;k . 2+ 0 0 3 2 0 + 0 3 2 0 0 +3 4 0 0 +5 % 4+ 0 0 5 % 4 0 + 0 5 0 + 0 0 0 + + 0 0 Latin Squares An n × n Latin Square is an n × n array of n symbols such that each symbol occurs exactly once in each row and column. 21 2 33 42 3 15 3 1 2 Each n × n latin square can be decomposed uniquely into a sum of scalar multiples of mutually orthogonal permutation matrices. Cian O'Brien (NUIG) ASBG-Colourings October 4th, 2019 6 / 9 Therefore each latin square corresponds uniquely to a permutation hypermatrix. For a permutation hypermatrix M, define L(M) to be Pn L(M)i;j;k = k=1 k × Mi;j;k . 2+ 0 0 3 2 0 + 0 3 2 0 0 +3 4 0 0 +5 % 4+ 0 0 5 % 4 0 + 0 5 0 + 0 0 0 + + 0 0 Latin Squares An n × n Latin Square is an n × n array of n symbols such that each symbol occurs exactly once in each row and column. 21 2 33 21 0 03 20 1 03 20 0 13 42 3 15 = 40 0 15 + 2 41 0 05 + 3 40 1 05 3 1 2 0 1 0 0 0 1 1 0 0 Each n × n latin square can be decomposed uniquely into a sum of scalar multiples of mutually orthogonal permutation matrices.
Load more

Recommended publications
  • Math 511 Advanced Linear Algebra Spring 2006
    MATH 511 ADVANCED LINEAR ALGEBRA SPRING 2006 Sherod Eubanks HOMEWORK 2 x2:1 : 2; 5; 9; 12 x2:3 : 3; 6 x2:4 : 2; 4; 5; 9; 11 Section 2:1: Unitary Matrices Problem 2 If ¸ 2 σ(U) and U 2 Mn is unitary, show that j¸j = 1. Solution. If ¸ 2 σ(U), U 2 Mn is unitary, and Ux = ¸x for x 6= 0, then by Theorem 2:1:4(g), we have kxkCn = kUxkCn = k¸xkCn = j¸jkxkCn , hence j¸j = 1, as desired. Problem 5 Show that the permutation matrices in Mn are orthogonal and that the permutation matrices form a sub- group of the group of real orthogonal matrices. How many different permutation matrices are there in Mn? Solution. By definition, a matrix P 2 Mn is called a permutation matrix if exactly one entry in each row n and column is equal to 1, and all other entries are 0. That is, letting ei 2 C denote the standard basis n th element of C that has a 1 in the i row and zeros elsewhere, and Sn be the set of all permutations on n th elements, then P = [eσ(1) j ¢ ¢ ¢ j eσ(n)] = Pσ for some permutation σ 2 Sn such that σ(k) denotes the k member of σ. Observe that for any σ 2 Sn, and as ½ 1 if i = j eT e = σ(i) σ(j) 0 otherwise for 1 · i · j · n by the definition of ei, we have that 2 3 T T eσ(1)eσ(1) ¢ ¢ ¢ eσ(1)eσ(n) T 6 .
    [Show full text]
  • Chapter Four Determinants
    Chapter Four Determinants In the first chapter of this book we considered linear systems and we picked out the special case of systems with the same number of equations as unknowns, those of the form T~x = ~b where T is a square matrix. We noted a distinction between two classes of T ’s. While such systems may have a unique solution or no solutions or infinitely many solutions, if a particular T is associated with a unique solution in any system, such as the homogeneous system ~b = ~0, then T is associated with a unique solution for every ~b. We call such a matrix of coefficients ‘nonsingular’. The other kind of T , where every linear system for which it is the matrix of coefficients has either no solution or infinitely many solutions, we call ‘singular’. Through the second and third chapters the value of this distinction has been a theme. For instance, we now know that nonsingularity of an n£n matrix T is equivalent to each of these: ² a system T~x = ~b has a solution, and that solution is unique; ² Gauss-Jordan reduction of T yields an identity matrix; ² the rows of T form a linearly independent set; ² the columns of T form a basis for Rn; ² any map that T represents is an isomorphism; ² an inverse matrix T ¡1 exists. So when we look at a particular square matrix, the question of whether it is nonsingular is one of the first things that we ask. This chapter develops a formula to determine this. (Since we will restrict the discussion to square matrices, in this chapter we will usually simply say ‘matrix’ in place of ‘square matrix’.) More precisely, we will develop infinitely many formulas, one for 1£1 ma- trices, one for 2£2 matrices, etc.
    [Show full text]
  • Rotation Matrix - Wikipedia, the Free Encyclopedia Page 1 of 22
    Rotation matrix - Wikipedia, the free encyclopedia Page 1 of 22 Rotation matrix From Wikipedia, the free encyclopedia In linear algebra, a rotation matrix is a matrix that is used to perform a rotation in Euclidean space. For example the matrix rotates points in the xy -Cartesian plane counterclockwise through an angle θ about the origin of the Cartesian coordinate system. To perform the rotation, the position of each point must be represented by a column vector v, containing the coordinates of the point. A rotated vector is obtained by using the matrix multiplication Rv (see below for details). In two and three dimensions, rotation matrices are among the simplest algebraic descriptions of rotations, and are used extensively for computations in geometry, physics, and computer graphics. Though most applications involve rotations in two or three dimensions, rotation matrices can be defined for n-dimensional space. Rotation matrices are always square, with real entries. Algebraically, a rotation matrix in n-dimensions is a n × n special orthogonal matrix, i.e. an orthogonal matrix whose determinant is 1: . The set of all rotation matrices forms a group, known as the rotation group or the special orthogonal group. It is a subset of the orthogonal group, which includes reflections and consists of all orthogonal matrices with determinant 1 or -1, and of the special linear group, which includes all volume-preserving transformations and consists of matrices with determinant 1. Contents 1 Rotations in two dimensions 1.1 Non-standard orientation
    [Show full text]
  • Alternating Sign Matrices, Extensions and Related Cones
    See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/311671190 Alternating sign matrices, extensions and related cones Article in Advances in Applied Mathematics · May 2017 DOI: 10.1016/j.aam.2016.12.001 CITATIONS READS 0 29 2 authors: Richard A. Brualdi Geir Dahl University of Wisconsin–Madison University of Oslo 252 PUBLICATIONS 3,815 CITATIONS 102 PUBLICATIONS 1,032 CITATIONS SEE PROFILE SEE PROFILE Some of the authors of this publication are also working on these related projects: Combinatorial matrix theory; alternating sign matrices View project All content following this page was uploaded by Geir Dahl on 16 December 2016. The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document and are linked to publications on ResearchGate, letting you access and read them immediately. Alternating sign matrices, extensions and related cones Richard A. Brualdi∗ Geir Dahly December 1, 2016 Abstract An alternating sign matrix, or ASM, is a (0; ±1)-matrix where the nonzero entries in each row and column alternate in sign, and where each row and column sum is 1. We study the convex cone generated by ASMs of order n, called the ASM cone, as well as several related cones and polytopes. Some decomposition results are shown, and we find a minimal Hilbert basis of the ASM cone. The notion of (±1)-doubly stochastic matrices and a generalization of ASMs are introduced and various properties are shown. For instance, we give a new short proof of the linear characterization of the ASM polytope, in fact for a more general polytope.
    [Show full text]
  • More Gaussian Elimination and Matrix Inversion
    Week 7 More Gaussian Elimination and Matrix Inversion 7.1 Opening Remarks 7.1.1 Introduction * View at edX 289 Week 7. More Gaussian Elimination and Matrix Inversion 290 7.1.2 Outline 7.1. Opening Remarks..................................... 289 7.1.1. Introduction..................................... 289 7.1.2. Outline....................................... 290 7.1.3. What You Will Learn................................ 291 7.2. When Gaussian Elimination Breaks Down....................... 292 7.2.1. When Gaussian Elimination Works........................ 292 7.2.2. The Problem.................................... 297 7.2.3. Permutations.................................... 299 7.2.4. Gaussian Elimination with Row Swapping (LU Factorization with Partial Pivoting) 304 7.2.5. When Gaussian Elimination Fails Altogether................... 311 7.3. The Inverse Matrix.................................... 312 7.3.1. Inverse Functions in 1D.............................. 312 7.3.2. Back to Linear Transformations.......................... 312 7.3.3. Simple Examples.................................. 314 7.3.4. More Advanced (but Still Simple) Examples................... 320 7.3.5. Properties...................................... 324 7.4. Enrichment......................................... 325 7.4.1. Library Routines for LU with Partial Pivoting................... 325 7.5. Wrap Up.......................................... 327 7.5.1. Homework..................................... 327 7.5.2. Summary...................................... 327 7.1. Opening
    [Show full text]
  • Pattern Avoidance in Alternating Sign Matrices
    PATTERN AVOIDANCE IN ALTERNATING SIGN MATRICES ROBERT JOHANSSON AND SVANTE LINUSSON Abstract. We generalize the definition of a pattern from permu- tations to alternating sign matrices. The number of alternating sign matrices avoiding 132 is proved to be counted by the large Schr¨oder numbers, 1, 2, 6, 22, 90, 394 . .. We give a bijection be- tween 132-avoiding alternating sign matrices and Schr¨oder-paths, which gives a refined enumeration. We also show that the 132, 123- avoiding alternating sign matrices are counted by every second Fi- bonacci number. 1. Introduction For some time, we have emphasized the use of permutation matrices when studying pattern avoidance. This and the fact that alternating sign matrices are generalizations of permutation matrices, led to the idea of studying pattern avoidance in alternating sign matrices. The main results, concerning 132-avoidance, are presented in Section 3. In Section 5, we enumerate sets of alternating sign matrices avoiding two patterns of length three at once. In Section 6 we state some open problems. Let us start with the basic definitions. Given a permutation π of length n we will represent it with the n×n permutation matrix with 1 in positions (i, π(i)), 1 ≤ i ≤ n and 0 in all other positions. We will also suppress the 0s. 1 1 3 1 2 1 Figure 1. The permutation matrix of 132. In this paper, we will think of permutations in terms of permutation matrices. Definition 1.1. (Patterns in permutations) A permutation π is said to contain a permutation τ as a pattern if the permutation matrix of τ is a submatrix of π.
    [Show full text]
  • Matrices That Commute with a Permutation Matrix
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Matrices That Commute With a Permutation Matrix Jeffrey L. Stuart Department of Mathematics University of Southern Mississippi Hattiesburg, Mississippi 39406-5045 and James R. Weaver Department of Mathematics and Statistics University of West Florida Pensacola. Florida 32514 Submitted by Donald W. Robinson ABSTRACT Let P be an n X n permutation matrix, and let p be the corresponding permuta- tion. Let A be a matrix such that AP = PA. It is well known that when p is an n-cycle, A is permutation similar to a circulant matrix. We present results for the band patterns in A and for the eigenstructure of A when p consists of several disjoint cycles. These results depend on the greatest common divisors of pairs of cycle lengths. 1. INTRODUCTION A central theme in matrix theory concerns what can be said about a matrix if it commutes with a given matrix of some special type. In this paper, we investigate the combinatorial structure and the eigenstructure of a matrix that commutes with a permutation matrix. In doing so, we follow a long tradition of work on classes of matrices that commute with a permutation matrix of a specified type. In particular, our work is related both to work on the circulant matrices, the results for which can be found in Davis’s compre- hensive book [5], and to work on the centrosymmetric matrices, which can be found in the book by Aitken [l] and in the papers by Andrew [2], Bamett, LZNEAR ALGEBRA AND ITS APPLICATIONS 150:255-265 (1991) 255 Q Elsevier Science Publishing Co., Inc., 1991 655 Avenue of the Americas, New York, NY 10010 0024-3795/91/$3.50 256 JEFFREY L.
    [Show full text]
  • The Waldspurger Transform of Permutations and Alternating Sign Matrices
    Séminaire Lotharingien de Combinatoire 78B (2017) Proceedings of the 29th Conference on Formal Power Article #41, 12 pp. Series and Algebraic Combinatorics (London) The Waldspurger Transform of Permutations and Alternating Sign Matrices Drew Armstrong∗ and James McKeown Department of Mathematics, University of Miami, Coral Gables, FL Abstract. In 2005 J. L. Waldspurger proved the following theorem: Given a finite real reflection group G, the closed positive root cone is tiled by the images of the open weight cone under the action of the linear transformations 1 g. Shortly after this − E. Meinrenken extended the result to affine Weyl groups and then P. V. Bibikov and V. S. Zhgoon gave a uniform proof for a discrete reflection group acting on a simply- connected space of constant curvature. In this paper we show that the Waldspurger and Meinrenken theorems of type A give an interesting new perspective on the combinatorics of the symmetric group. In particular, for each permutation matrix g S we define a non-negative integer 2 n matrix WT(g), called the Waldspurger transform of g. The definition of the matrix WT(g) is purely combinatorial but it turns out that its columns are the images of the fundamental weights under 1 g, expressed in simple root coordinates. The possible − columns of WT(g) (which we call UM vectors) biject to many interesting structures including: unimodal Motzkin paths, abelian ideals in the Lie algebra sln(C), Young diagrams with maximum hook length n, and integer points inside a certain polytope. We show that the sum of the entries of WT(g) is half the entropy of the corresponding permutation g, which is known to equal the rank of g in the MacNeille completion of the Bruhat order.
    [Show full text]
  • Applied Linear Algebra and Differential Equations
    Applied Linear Algebra and Differential Equations Lecture notes for MATH 2350 Jeffrey R. Chasnov The Hong Kong University of Science and Technology Department of Mathematics Clear Water Bay, Kowloon Hong Kong Copyright ○c 2017-2019 by Jeffrey Robert Chasnov This work is licensed under the Creative Commons Attribution 3.0 Hong Kong License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/hk/ or send a letter to Creative Commons, 171 Second Street, Suite 300, San Francisco, California, 94105, USA. Preface What follows are my lecture notes for a mathematics course offered to second-year engineering students at the the Hong Kong University of Science and Technology. Material from our usual courses on linear algebra and differential equations have been combined into a single course (essentially, two half-semester courses) at the request of our Engineering School. I have tried my best to select the most essential and interesting topics from both courses, and to show how knowledge of linear algebra can improve students’ understanding of differential equations. All web surfers are welcome to download these notes and to use the notes and videos freely for teaching and learning. I also have some online courses on Coursera. You can click on the links below to explore these courses. If you want to learn differential equations, have a look at Differential Equations for Engineers If your interests are matrices and elementary linear algebra, try Matrix Algebra for Engineers If you want to learn vector calculus (also known as multivariable calculus, or calcu- lus three), you can sign up for Vector Calculus for Engineers And if your interest is numerical methods, have a go at Numerical Methods for Engineers Jeffrey R.
    [Show full text]
  • Numerically Stable Coded Matrix Computations Via Circulant and Rotation Matrix Embeddings
    1 Numerically stable coded matrix computations via circulant and rotation matrix embeddings Aditya Ramamoorthy and Li Tang Department of Electrical and Computer Engineering Iowa State University, Ames, IA 50011 fadityar,[email protected] Abstract Polynomial based methods have recently been used in several works for mitigating the effect of stragglers (slow or failed nodes) in distributed matrix computations. For a system with n worker nodes where s can be stragglers, these approaches allow for an optimal recovery threshold, whereby the intended result can be decoded as long as any (n − s) worker nodes complete their tasks. However, they suffer from serious numerical issues owing to the condition number of the corresponding real Vandermonde-structured recovery matrices; this condition number grows exponentially in n. We present a novel approach that leverages the properties of circulant permutation matrices and rotation matrices for coded matrix computation. In addition to having an optimal recovery threshold, we demonstrate an upper bound on the worst-case condition number of our recovery matrices which grows as ≈ O(ns+5:5); in the practical scenario where s is a constant, this grows polynomially in n. Our schemes leverage the well-behaved conditioning of complex Vandermonde matrices with parameters on the complex unit circle, while still working with computation over the reals. Exhaustive experimental results demonstrate that our proposed method has condition numbers that are orders of magnitude lower than prior work. I. INTRODUCTION Present day computing needs necessitate the usage of large computation clusters that regularly process huge amounts of data on a regular basis. In several of the relevant application domains such as machine learning, arXiv:1910.06515v4 [cs.IT] 9 Jun 2021 datasets are often so large that they cannot even be stored in the disk of a single server.
    [Show full text]
  • Contents September 1
    Contents September 1 : Examples 2 September 3 : The Gessel-Lindstr¨om-ViennotLemma 3 September 15 : Parametrizing totally positive unipotent matrices { Part 1 6 September 17 : Parametrizing totally positive unipotent matrices { part 2 7 September 22 : Introduction to the 0-Hecke monoid 9 September 24 : The 0-Hecke monoid continued 11 September 29 : The product has the correct ranks 12 October 1 : Bruhat decomposition 14 October 6 : A unique Bruhat decomposition 17 October 8: B−wB− \ N − + as a manifold 18 October 13: Grassmannians 21 October 15: The flag manifold, and starting to invert the unipotent product 23 October 20: Continuing to invert the unipotent product 28 October 22 { Continuing to invert the unipotent product 30 October 27 { Finishing the inversion of the unipotent product 32 October 29 { Chevalley generators in GLn(R) 34 November 5 { LDU decomposition, double Bruhat cells, and total positivity for GLn(R) 36 November 10 { Kasteleyn labelings 39 November 12 { Kasteleyn labellings continued 42 November 17 { Using graphs in a disc to parametrize the totally nonnegative Grassmannian 45 November 19 { cyclic rank matrices 45 December 1 { More on cyclic rank matrices 48 December 3 { Bridges and Chevalley generators 51 December 8 { finishing the proof 55 1 2 September 1 : Examples We write [m] = f1; 2; : : : ; mg. Given an m × n matrix M, and subsets I =⊆ [m] and I J ⊆ [n] with #(I) = #(J), we define ∆J (M) = det(Mij)i2I; j2J , where we keep the elements of I and the elements of J in the same order. For example, 13 M12 M15 ∆25(M) = det : M32 M35 I The ∆J (M) are called the minors of M.
    [Show full text]
  • Review of Matrices and Block Structures
    APPENDIX A Review of Matrices and Block Structures Numerical linear algebra lies at the heart of modern scientific computing and computational science. Today it is not uncommon to perform numerical computations with matrices having millions of components. The key to understanding how to implement such algorithms is to exploit underlying structure within the matrices. In these notes we touch on a few ideas and tools for dissecting matrix structure. Specifically we are concerned with the block structure matrices. 1. Rows and Columns m n Let A R ⇥ so that A has m rows and n columns. Denote the element of A in the ith row and jth column 2 as Aij. Denote the m rows of A by A1 ,A2 ,A3 ,...,Am and the n columns of A by A 1,A2,A3,...,An. For example, if · · · · · · · · 32 15 73 − A = 2 27 32 100 0 0 , 2 −89 0 47− 22 21 333 − − then A = 100, 4 5 2,4 − A1 = 32 1573,A2 = 2 27 32 100 0 0 ,A3 = 89 0 47 22 21 33 · − · − − · − − and ⇥ ⇤ ⇥ ⇤ ⇥ ⇤ 3 2 1 5 7 3 − A 1 = 2 ,A2 = 27 ,A3 = 32 ,A4 = 100 ,A5 = 0 ,A6 = 0 . · 2 −893 · 2 0 3 · 2 47 3 · 2−22 3 · 2 213 · 2333 − − 4 5 4 5 4 5 4 5 4 5 4 5 Exercise 1.1. If 3 41100 220010− 2 3 C = 100001, − 6 0002147 6 7 6 0001037 6 7 4 5 4 −2 2 3 what are C4,4, C 4 and C4 ? For example, C2 = 220010and C 2 = 0 .
    [Show full text]