DOCSLIB.ORG

Parallel Systems in Symbolic and Algebraic Computation

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Parallel Systems in Symbolic and Algebraic Computation UCAM-CL-TR-537 Technical Report ISSN 1476-2986 Number 537 Computer Laboratory Parallel systems in symbolic and algebraic computation Mantsika Matooane June 2002 15 JJ Thomson Avenue Cambridge CB3 0FD United Kingdom phone +44 1223 763500 http://www.cl.cam.ac.uk/ c 2002 Mantsika Matooane ! This technical report is based on a dissertation submitted August 2001 by the author for the degree of Doctor of Philosophy to the University of Cambridge, Trinity College. Technical reports published by the University of Cambridge Computer Laboratory are freely available via the Internet: http://www.cl.cam.ac.uk/TechReports/ Series editor: Markus Kuhn ISSN 1476-2986 Abstract This thesis describes techniques that exploit the distributed memory in massively parallel processors to satisfy the peak memory requirements of some very large com- puter algebra problems. Our aim is to achieve balanced memory use, which differen- tiates this work from other parallel systems whose focus is on gaining speedup. It is widely observed that failures in computer algebra systems are mostly due to mem- ory overload: for several problems in computer algebra, some of the best available algorithms suffer from intermediate expression swell where the result is of reason- able size, but the intermediate calculation encounters severe memory limitations. This observation motivates our memory-centric approach to parallelizing computer algebra algorithms. The memory balancing is based on a randomized hashing algorithm for dynamic distribution of data. Dynamic distribution means that the intermediate data is allocated storage space at the time that it is created and therefore the system can avoid overloading some processing elements. Large scale computer algebra problems with peak memory demands of more than 10 gigabytes are considered. Distributed memory can scale to satisfy these requirements. For example, the Hitachi SR2201 which is the target architecture in this research provides up to 56 gigabytes of memory. The system has fine granularity: tasks sizes are small and data is partitioned in small blocks. The fine granularity provides flexibility in controlling memory balance but incurs higher communication costs. The communication overhead is reduced by an intelligent scheduler which performs asynchronous overlap of communication and computation. The implementation provides a polynomial algebra system with operations on multivariate polynomials and matrices with polynomial entries. Within this frame- work it is possible to find computations with large memory demands, for example, solving large sparse systems of linear equations and Gro¨bner base computations. The parallel algorithms that have been implemented are based on the standard algorithms for polynomial algebra. This demonstrates that careful attention to memory management aids solution of very large problems even without the benefit of advanced algorithms. The parallel implementation can be used to solve larger problems than have previously been possible. 3 4 Contents 1 Introduction 9 1.1 Motivation . 10 1.2 Randomized dynamic distribution . 13 1.3 Algebraic structures . 15 1.4 Outline of the dissertation . 16 2 Literature review 19 2.1 Target architecture . 19 2.2 Performance metrics . 25 2.3 Algorithms for data partitioning . 35 2.4 Computer algebra systems . 36 2.5 Parallel arithmetic . 41 2.6 Sparse systems of linear equations . 43 2.7 Gr¨obner bases . 48 2.8 Summary . 51 3 Data structures and memory allocation 53 3.1 Data locality . 54 3.2 Granularity . 55 3.3 Randomized memory balancing . 56 3.4 Local memory allocation . 58 3.5 Load factor . 60 3.6 Relocation . 60 3.7 Analysis of the randomized global hashing . 62 3.8 Weighted memory balancing . 63 3.9 The scheduler . 66 3.10 Summary . 69 4 Parallel arithmetic 71 4.1 Multiprecision integer arithmetic . 71 4.2 Parallel polynomial arithmetic . 77 4.3 Summary . 80 5 Systems of polynomial equations 83 5.1 Parallel Gro¨bner base algorithm . 83 5.2 Parallel sparse linear systems . 90 5 5.3 Summary . 97 6 Conclusions 99 6.1 Contributions . 99 6.2 Future directions . 100 6.3 Conclusions . 101 A CABAL: a short manual 103 A.1 CommSystem interface . 104 A.2 Polynomial interface . 105 A.3 Bignum interface . 106 A.4 Grobner interface . 106 B The Hitachi SR2201 109 B.1 The processor . 109 B.2 Pseudo vector processing . 110 B.3 Interconnection network . 111 B.4 Mapping of processors to topology . 113 B.5 Inter-process communication . 113 B.6 Memory hierarchy . 115 B.7 Programming environment . 115 C Network topologies 117 C.1 Fully connected network . 117 C.2 Ring network . 118 C.3 Star network . 118 C.4 Tree network . 118 C.5 Mesh network . 119 C.6 Hypercube . 119 C.7 The crossbar . 121 D Message passing interface 123 D.1 Features of MPI . 123 D.2 Communicators . 124 D.3 Point-to-point communication . 125 D.4 Type matching . 126 D.5 Collective communications . 126 D.6 Process groups and topologies . 127 D.7 Comparison of MPI and PVM . 127 D.8 Other communication libraries . 128 6 List of Figures 1.1 Growing memory requirements for a GCD calculation . 11 1.2 The comparative performance of memory and CPU . 12 1.3 Development of a parallel computer algebra system . 16 2.1 Resources in a parallel system . 20 2.2 Classes of distributed memory parallel machines . 22 2.3 Layering of parallel components . 23 2.4 Communication layers . 28 2.5 Partitioning of data with different granularity . 32 2.6 Block distribution . 36 2.7 The Bareiss determinant algorithm . 46 2.8 Buchberger’s algorithm for Gr¨obner basis . 49 3.1 Communication categories affecting data locality . 55 3.2 Fine granularity during multiplication . 56 3.3 Randomized memory allocation . 57 3.4 Dynamic partitioning at execution time . 59 3.5 Margin of imbalance in memory load (κ) . 60 3.6 Traffic during multiplication with 2 PEs . 64 3.7 Traffic during multiplication with 3 PEs . 65 3.8 Local process scheduling . 67 3.9 Architecture of a parallel computer algebra system . 68 4.1 Kernel components . 71 4.2 Multiprecision addition . 74 4.3 Parallel polynomial addition . 78 4.4 Identifying blocks of the same term . 78 4.5 Parallel polynomial multiplication . 79 4.6 Execution time when increasing buffer size . 80 5.1 Ordering s-pairs below the diagonal . 85 5.2 Row algorithm for selecting critical s-pairs . 85 5.3 Selecting leading term of a polynomial . 87 5.4 Distributed s-polynomial computation . 88 5.5 Reduction algorithm . 88 5.6 Parallel reduction algorithm . 89 5.7 Sparse matrix representation . 90 5.8 Recursion tree for minor expansion . 94 7 5.9 Costs at different levels of recursion tree . 94 5.10 Parallelizing recursive minor expansion . 96 A.1 Application programming interface . 104 B.1 Vector inner product . 109 B.2 Object code for vector inner product . 110 B.3 Sliding windows for pseudo-vector processing . 111 B.4 Some additional instructions in SR2201 . 111 B.5 3D crossbar switching network . 112 B.6 Message passing with system copy . 114 B.7 Remote direct memory access (rDMA) system . 114 B.8 Comparison of some message passing systems . 115 C.1 A fully connected network of 5 processors . 118 C.2 Ring, linear and star topologies . 118 C.3 A binary tree . 119 C.4 A 2D mesh network . 119 C.5 Hypercubes in 2,3,4 dimensions . 120 C.6 A 2D crossbar network . 121 D.1 A message passing interface (MPI) packet . ..
Load more

Recommended publications
  • Nonlinear Evolution Equations and Solving Algebraic Systems: the Importance of Computer Algebra
    - Y'M?S </#i/ ИНСТИТУТ ядерных исследований дубна Е5-89-62Ц V.P.Gerdt, N.A.Kostov, A.Yu.Zharkov* NONLINEAR EVOLUTION EQUATIONS AND SOLVING ALGEBRAIC SYSTEMS: THE IMPORTANCE OF COMPUTER ALGEBRA Submitted to International Conference "Solitons and its Applications", Dubna, August 25-27, 1989 * Saratov State University, USSR 1989 1.INTRODUCTION A wide-spread problem in the theory of nonlinear evolution equations (NEE) is to find the exact solutions of complicated algebraic systems. For example, let us consider the following evolution systems U - AU + F(U,U , ..U ), U=U(x,t)-(U1, . ,UH) , U=D1(U), D=d/dx IN 1 N-1 I /ii F«(F\ . -FH) , A==diag(A , . .A) , A *0, A *A , (i*j). The general symmetry approach to checking up the integrability and classification of integrable NEE (see the reviews [1,2] and references therein) allows to obtain the integrability conditions which are related to existence of higher order symmetries and conservation laws in a fully algorithmic way. The implementation of these algorithms in a form of FORMAC program FORMINT have been given for scalar equations (M=l in (1)) in [3 j and for the general case (M>1) in [4]. Using this program one can automatically obtain the equations in right hand side of (l) which follow from the necessary integrability conditions. The integrabiiity conditions for the right hand side with arbitrary functions have the form of systems of differential equations, the procedure of solving these equations is not generally algorithmic. But in the important particular cases when the right hand side F are polynomials and the integrability conditions are reduced to a system of nonlinear algebraic equations in coefficients of the polynomials.
    [Show full text]
  • A Weakly Stable Algorithm for General Toeplitz Systems
    A Weakly Stable Algorithm for General Toeplitz Systems∗ Adam W. Bojanczyk Richard P. Brent School of Electrical Engineering Computer Sciences Laboratory Cornell University Australian National University Ithaca, NY 14853-5401 Canberra, ACT 0200 Frank R. de Hoog Division of Mathematics and Statistics CSIRO, GPO Box 1965 Canberra, ACT 2601 Report TR-CS-93-15 6 August 1993 (revised 24 June 1994) Abstract We show that a fast algorithm for the QR factorization of a Toeplitz or Hankel matrix A is weakly stable in the sense that RT R is close to AT A. Thus, when the algorithm is used to solve the semi-normal equations RT Rx = AT b, we obtain a weakly stable method for the solution of a nonsingular Toeplitz or Hankel linear system Ax = b. The algorithm also applies to the solution of the full-rank Toeplitz or Hankel least squares problem min kAx − bk2. 1991 Mathematics Subject Classification. Primary 65F25; Secondary 47B35, 65F05, 65F30, 65Y05, 65Y10 Key words and phrases. Cholesky factorization, error analysis, Hankel matrix, least squares, normal equations, orthogonal factorization, QR factorization, semi-normal equa- tions, stability, Toeplitz matrix, weak stability. 1 Introduction arXiv:1005.0503v1 [math.NA] 4 May 2010 Toeplitz linear systems arise in many applications, and there are many algorithms which solve nonsingular n × n Toeplitz systems Ax = b in O(n2) arithmetic operations [2, 13, 44, 45, 49, 54, 55, 65, 67, 77, 78, 80, 81, 85, 86]. Some algorithms are restricted to symmetric systems (A = AT ) and others apply to general Toeplitz systems. Because of their recursive nature, most O(n2) algorithms assume that all leading principal submatrices of A are nonsingular, and break down if this is not the case.
    [Show full text]
  • Linear Algebra: Beware!
    LINEAR ALGEBRA: BEWARE! MATH 196, SECTION 57 (VIPUL NAIK) You might be expecting linear algebra to be a lot like your calculus classes at the University. This is probably true in terms of the course structure and format. But it’s not true at the level of subject matter. Some important differences are below. • Superficially, linear algebra is a lot easier, since it relies mostly on arithmetic rather than algebra. The computational procedures involve the systematic and correct application of processes for adding, subtracting, multiplying, and dividing numbers. But the key word here is superficially. • Even for the apparently straightforward computational exercises, it turns out that people are able to do them a lot better if they understand what’s going on. In fact, in past test questions, people have often made fewer errors when doing the problem using full-scale algebraic symbol manipulation rather than the synthetic arithmetic method. • One important difference between linear algebra and calculus is that with calculus, it’s relatively easy to understand ideas partially. One can obtain much of the basic intuition of calculus by un- derstanding graphs of functions. In fact, limit, continuity, differentaaition, and integration all have basic descriptions in terms of the graph. Note: these aren’t fully rigorous, which is why you had to take a year’s worth of calculus class to cement your understanding of the ideas. But it’s a start. With linear algebra, there is no single compelling visual tool that connects all the ideas, and conscious effort is needed even for a partial understanding. • While linear algebra lacks any single compelling visual tool, it requires either considerable visuo- spatial skill or considerable abstract symbolic and verbal skill (or a suitable linear combination thereof).
    [Show full text]
  • WORKSHOP on COMPUTATIONAL ASPECTS in the CONTROL of FLEXIBLE SYSTEMS
    NASA Technical Memorandum 10 1578, Part One WORKSHOP on COMPUTATIONAL ASPECTS in the CONTROL of FLEXIBLE SYSTEMS Held at the Royce Hotel in WiIIiamsburg, Virginia (\!~?~-~*-lolj?d-pt-l) P&~C~L~IN~Y-F THE <UL (; 8.; t\jo~-l?~Su ~~KY,~+~JDfl~CnHPUTATIONAL ASPFCTI IY TcF --T )~uI)-- ~f'4T1.1L dC Ei'XfrLt. >Y>TtU,, f'AQT i (.\A>A- r\eu-lP>CL L an-1-y Rps~arihCentnr) 492 p C5CL 2LR UnLl ~s 55 3 G;/i3 J,'i74qA Sponsored by the NASA Langley Research Center Proceedings Compiled by Larry Taylor Table of Contents Page Introduction Computational Aspects Workshop Call for Papers 1 Workshop Organizing Committee 5 Attendance List 7 --------___-----__-------------------------------------- Needs for Advanced CSI Software NASA's Control/Structures Interaction (CSI) Program Brantley R. Hanks, NASA Langley Research Center 21-, Computational Controls for Aerospace Systems Guy Man, Robert A. Laskin and A. Fernando Tolivar Jet Propulsion Laboratory 33. Additional Software Developments Wanted for Modeling and Control of Flexible Systems Jiguan G. Lin, Control Research Corporation 4 9 Survey of Available Software Flexible Structure Control Experiments Using a Real-Time Workstation for Computer-Aided Control Engineering Michael E. Steiber, Communications Research Centre 6 7 CONSOLE: A CAD Tandem for Optimizationl-Based Design Interacting with User-Supplied Simulators Michael K.H. Fan, Li-Shen Wang, Jan Koninckx and Andre L. Tits,University of Maryland, College Park 8 9 mLPAGE IS QUALfTY ORIGINAL PAGE IS OF POOR QUALITY The Application of TSIM Software to ACT Design and Analysis of Flexible Aircraft Ian W. Kaynes, Royal Aerospace Establishment, Farnborouth 109 - Control/Structure Interaction Methods for Space Statian Power Systems Paul Blelloch, Structural Dynamics Research Corporation 121L, Flexible Missile Autopilot Design Studies with PC-MATLAB386 Michael J.
    [Show full text]
  • Parallel Systems in Symbolic and Algebraic Computation
    UCAM-CL-TR-537 Technical Report ISSN 1476-2986 Number 537 Computer Laboratory Parallel systems in symbolic and algebraic computation Mantsika Matooane June 2002 15 JJ Thomson Avenue Cambridge CB3 0FD United Kingdom phone +44 1223 763500 http://www.cl.cam.ac.uk/ c 2002 Mantsika Matooane This technical report is based on a dissertation submitted August 2001 by the author for the degree of Doctor of Philosophy to the University of Cambridge, Trinity College. Technical reports published by the University of Cambridge Computer Laboratory are freely available via the Internet: http://www.cl.cam.ac.uk/TechReports/ Series editor: Markus Kuhn ISSN 1476-2986 Abstract This thesis describes techniques that exploit the distributed memory in massively parallel processors to satisfy the peak memory requirements of some very large com- puter algebra problems. Our aim is to achieve balanced memory use, which differen- tiates this work from other parallel systems whose focus is on gaining speedup. It is widely observed that failures in computer algebra systems are mostly due to mem- ory overload: for several problems in computer algebra, some of the best available algorithms suffer from intermediate expression swell where the result is of reason- able size, but the intermediate calculation encounters severe memory limitations. This observation motivates our memory-centric approach to parallelizing computer algebra algorithms. The memory balancing is based on a randomized hashing algorithm for dynamic distribution of data. Dynamic distribution means that the intermediate data is allocated storage space at the time that it is created and therefore the system can avoid overloading some processing elements.
    [Show full text]
  • Arxiv:1810.01634V2 [Cs.SC] 6 Nov 2020 Algebraic Number Fields And
    Algebraic number fields and the LLL algorithm M. J´anos Uray [email protected] ELTE – E¨otv¨os Lor´and University (Budapest) Faculty of Informatics Department of Computer Algebra Abstract In this paper we analyze the computational costs of various operations and algorithms in algebraic number fields using exact arithmetic. Let K be an algebraic number field. In the first half of the paper, we calculate the running time and the size of the output of many operations in K in terms of the size of the input and the parameters of K. We include some earlier results about these, but we go further than them, e.g. we also analyze some R-specific operations in K like less-than comparison. In the second half of the paper, we analyze two algorithms: the Bareiss algorithm, which is an integer-preserving version of the Gaussian elimination, and the LLL algorithm, which is for lattice basis reduction. In both cases, we extend the algorithm from Zn to Kn, and give a polynomial upper bound on the running time when the computations in K are performed exactly (as opposed to floating-point approximations). 1 Introduction Exact computation with algebraic numbers is an important feature that most computer algebra systems provide. They use efficient algorithms for the calculations, arXiv:1810.01634v2 [cs.SC] 6 Nov 2020 described in several papers and books, e.g. in [1, 2, 3]. However, the computational costs of these algorithms are often not obvious to calculate, because the bit complexity depends on how much the representing multi-precision integers grow during the computation.
    [Show full text]
  • (12) United States Patent (10) Patent No.: US 6,203,987 B1 Friend Et Al
    USOO6203987B1 (12) United States Patent (10) Patent No.: US 6,203,987 B1 Friend et al. (45) Date of Patent: Mar. 20, 2001 (54) METHODS FOR USING CO-REGULATED Garrels, JI et al “Quant exploration of the REF52 protein GENESETS TO ENHANCE DETECTION AND database: cluster analysis reveals major protein expression CLASSIFICATION OF GENE EXPRESSION profiles in resonses to growth regulation, Serum Stimulation, PATTERNS and viral transformation’, Electrophoresis, 12/90, 11(12): 1114–30.* (75) Inventors: Stephen H. Friend, Seattle, WA (US); Anderson et al., 1994, “Involvement of the protein tyrosine Roland Stoughton, San Diego, CA kinase p56' in T cell signaling and thymocyte develop (US) ment.” Adv. Immunol. 56:151-178. (73) Assignee: Rosetta Inpharmatics, Inc., Kirkland, Anderson, 1995, “Mutagenesis", Methods Cell. Biol. 48:31. WA (US) Baudin et al., 1993, “A simple and efficient method for direct gene deletion in Saccharomyces cerevisiae, ' Nucl. Acids (*) Notice: Subject to any disclaimer, the term of this ReS. 21:3329-3330. patent is extended or adjusted under 35 Belshaw et al., 1996, “Controlling protein association and U.S.C. 154(b) by 0 days. Subcellular localization with a Synthetic ligand that induces heterodimerization of proteins,” Proc. Natl. Acad. Sci. USA (21) Appl. No.: 09/179,569 93:46O4-46O7. (22) Filed: Oct. 27, 1998 Bernoist and Chambon, 1981, “In vivo sequence require ments of the SV40 early promoter region”, Nature 290: (51) Int. Cl. ............................ C12O 1/68; C12P 21/04; 304-310. GO1N 33/543 Biocca, 1995, “Intracellular immunization: antibody target ing to subcellular compartments,” Trends in Cell Biology (52) U.S.
    [Show full text]
  • A Review of Mathematica
    A Review of Mathematica RICHARD J. FATEMAN ∗ [email protected] Computer Science Division, University of California, Berkeley, CA 94720, USA 16 September 1991 Abstract The Mathematica computer system is reviewed from the perspective of its contributions to symbolic and algebraic computation, as well as its stated goals. Design and implementation issues are discussed. 1 Introduction The Mathematica1 computer program is a general system for doing mathematical computation [51][52]. It includes a command language, a programming language, and a calculation environment that is oriented toward symbolic as well as numeric mathematics. The back cover of the manual [52] provides excerpts from rave notices like “The importance of [Mathematica] cannot be overlooked. it so fundamentally alters the mechanics of mathematics.” —The New York Times. Fortune [47] says “. it will do, instantaneously, virtually all of applied mathematics. .” Hype aside, the program is without question interesting to mathematicians, computer scientists, and engineers because of its combination of a number of ∗This work has been supported in part by the following: the National Science Foundation under grant numbers CCR-8812843 and CDS-8922788, through the Center for Pure and Applied Mathe- matics and the Electronics Research Laboratory (ERL) at the University of California at Berkeley; the Defense Advanced Research Projects Agency (DoD) ARPA order #4871, monitored by Space & Naval Warfare Systems Command under contract N00039-84-C-0089, through ERL; and grants from the IBM Corporation, the State of California MICRO program, and Sun Microsystems. 1 Mathematica is a trademark of Wolfram Research Inc. (WRI). 1 technologies that have arisen in initially separate contexts—numerical and sym- bolic mathematics, graphics, and modern user interfaces.
    [Show full text]
  • Insight MFR By
    Manufacturers, Publishers and Suppliers by Product Category 11/6/2017 10/100 Hubs & Switches ASCEND COMMUNICATIONS CIS SECURE COMPUTING INC DIGIUM GEAR HEAD 1 TRIPPLITE ASUS Cisco Press D‐LINK SYSTEMS GEFEN 1VISION SOFTWARE ATEN TECHNOLOGY CISCO SYSTEMS DUALCOMM TECHNOLOGY, INC. GEIST 3COM ATLAS SOUND CLEAR CUBE DYCONN GEOVISION INC. 4XEM CORP. ATLONA CLEARSOUNDS DYNEX PRODUCTS GIGAFAST 8E6 TECHNOLOGIES ATTO TECHNOLOGY CNET TECHNOLOGY EATON GIGAMON SYSTEMS LLC AAXEON TECHNOLOGIES LLC. AUDIOCODES, INC. CODE GREEN NETWORKS E‐CORPORATEGIFTS.COM, INC. GLOBAL MARKETING ACCELL AUDIOVOX CODI INC EDGECORE GOLDENRAM ACCELLION AVAYA COMMAND COMMUNICATIONS EDITSHARE LLC GREAT BAY SOFTWARE INC. ACER AMERICA AVENVIEW CORP COMMUNICATION DEVICES INC. EMC GRIFFIN TECHNOLOGY ACTI CORPORATION AVOCENT COMNET ENDACE USA H3C Technology ADAPTEC AVOCENT‐EMERSON COMPELLENT ENGENIUS HALL RESEARCH ADC KENTROX AVTECH CORPORATION COMPREHENSIVE CABLE ENTERASYS NETWORKS HAVIS SHIELD ADC TELECOMMUNICATIONS AXIOM MEMORY COMPU‐CALL, INC EPIPHAN SYSTEMS HAWKING TECHNOLOGY ADDERTECHNOLOGY AXIS COMMUNICATIONS COMPUTER LAB EQUINOX SYSTEMS HERITAGE TRAVELWARE ADD‐ON COMPUTER PERIPHERALS AZIO CORPORATION COMPUTERLINKS ETHERNET DIRECT HEWLETT PACKARD ENTERPRISE ADDON STORE B & B ELECTRONICS COMTROL ETHERWAN HIKVISION DIGITAL TECHNOLOGY CO. LT ADESSO BELDEN CONNECTGEAR EVANS CONSOLES HITACHI ADTRAN BELKIN COMPONENTS CONNECTPRO EVGA.COM HITACHI DATA SYSTEMS ADVANTECH AUTOMATION CORP. BIDUL & CO CONSTANT TECHNOLOGIES INC Exablaze HOO TOO INC AEROHIVE NETWORKS BLACK BOX COOL GEAR EXACQ TECHNOLOGIES INC HP AJA VIDEO SYSTEMS BLACKMAGIC DESIGN USA CP TECHNOLOGIES EXFO INC HP INC ALCATEL BLADE NETWORK TECHNOLOGIES CPS EXTREME NETWORKS HUAWEI ALCATEL LUCENT BLONDER TONGUE LABORATORIES CREATIVE LABS EXTRON HUAWEI SYMANTEC TECHNOLOGIES ALLIED TELESIS BLUE COAT SYSTEMS CRESTRON ELECTRONICS F5 NETWORKS IBM ALLOY COMPUTER PRODUCTS LLC BOSCH SECURITY CTC UNION TECHNOLOGIES CO FELLOWES ICOMTECH INC ALTINEX, INC.
    [Show full text]
  • Using a Small Algebraic Manipulation System to Solve Differential and Integral Equations by Variational and Approximation Techniques
    J. Symbolic Computation (1987) 3, 291-301 Using a Small Algebraic Manipulation System to Solve Differential and Integral Equations by Variational and Approximation Techniques R. D. MILLS Computing Science Department. University of Glasgow, Glasgow, UK (Received 2 December 1985) The microcomputer algebraic manipulation system MUMATH is used to implement the classical variational, Galerkin and least-squares techniques for solving boundary-value problems in differential equations and also for solving Fredhohn integral equations. Examples are given which extend the precision of known results. The technique is presented as a general algorithm which can readily be implemented on other algebraic manipulation systems. 1. Introduction Computer programs have been in existence since the fifties to manipulate polynomials, differentiate functions and solve equations. By the late sixties and early seventies programs were written which could integrate functions analytically. The late seventies saw the development of programs for the symbolic solution of differential and integral equations (see Golden, 1977; Stoutemyer, 1977; Bogen, 1979). Large and complex algebraic manipulation systems were devised to implement the algorithms which effect these processes, the best known being FORMAC, REDUCE2, MACSYMA and SCRATCHPAD. There then opened up the exciting possibility of tackling many problems in applied mathematics and engineering which could be solved approximately by analytical methods but which require very large amounts of algebraic manipulation. The Rayleigh-Ritz, Galerkin and least-squares methods for solving boundary-value problems in differential equations are typical examples. For early work using computer algebra in these methods, see Miola (1974) and Andersen & Noor (1977). This application area involves the interaction of numerical and algebraic computation (see Ng, 1979).
    [Show full text]
  • Symbolic Software for Lie Symmetry Computations
    Invited Lecture 1 Symbolic Software for Lie Symmetry Computations Willy Hereman Dept. Mathematical and Computer Sciences Colorado School of Mines Golden, CO 80401-1887 U.S.A. ISLC Workshop Nordfjordeid, Norway Tuesday, June 18, 1996 16:00 I. INTRODUCTION Symbolic Software • Solitons via Hirota’s method (Macsyma & Mathematica) • Painlev´etest for ODEs or PDEs (Macsyma & Mathematica) • Conservation laws of PDEs (Mathematica) • Lie symmetries for ODEs and PDEs (Macsyma) Purpose of the programs • Study of integrability of nonlinear PDEs • Exact solutions as bench mark for numerical algorithms • Classification of nonlinear PDEs • Lie symmetries −→ solutions via reductions • Work in collaboration with Unal¨ G¨okta¸s Chris Elmer Wuning Zhuang Ameina Nuseir Mark Coffey Erik van den Bulck Tony Miller Tracy Otto Symbolic Software by Willy Hereman and Collaborators Software is freely available from anonymous FTP site: mines.edu Change to subdirectory: pub/papers/math cs dept/software Subdirectory Structure: – symmetry (Macsyma) systems of ODEs and PDEs systems of difference-differential equations – hirota (Macsyma) – painleve (Macsyma) single system – condens (Mathematica) – painmath (Mathematica) single system – hiromath (Mathematica) Computation of Lie-point Symmetries – System of m differential equations of order k ∆i(x, u(k)) = 0, i = 1, 2, ..., m with p independent and q dependent variables p x = (x1, x2, ..., xp) ∈ IR u = (u1, u2, ..., uq) ∈ IRq – The group transformations have the form x˜ = Λgroup(x, u), u˜ = Ωgroup(x, u) where the functions Λgroup
    [Show full text]
  • Efficient Solutions to Toeplitz-Structured Linear Systems for Signal Processing
    EFFICIENT SOLUTIONS TO TOEPLITZ-STRUCTURED LINEAR SYSTEMS FOR SIGNAL PROCESSING A Dissertation Presented to The Academic Faculty by Christopher Kowalczyk Turnes In Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the School of Electrical and Computer Engineering Georgia Institute of Technology May 2014 Copyright © 2014 by Christopher Kowalczyk Turnes EFFICIENT SOLUTIONS TO TOEPLITZ-STRUCTURED LINEAR SYSTEMS FOR SIGNAL PROCESSING Approved by: Dr. James McClellan, Committee Chair Dr. Jack Poulson School of Electrical and Computer School of Computational Science and Engineering Engineering Georgia Institute of Technology Georgia Institute of Technology Dr. Justin Romberg, Advisor Dr. Chris Barnes School of Electrical and Computer School of Electrical and Computer Engineering Engineering Georgia Institute of Technology Georgia Institute of Technology Dr. Monson Hayes Dr. Doru Balcan School of Electrical and Computer Quantitative Finance Engineering Bank of America Georgia Institute of Technology Date Approved: May 2014 Effort is one of the things that gives meaning to life. Effort means you care about something, that something is important to you and you are willing to work for it. It would be an impoverished existence if you were not willing to value things and commit yourself to working toward them. – CAROL S. DWECK, SELF-THEORIES To my wonderful, amazing, and patient parents Cynthia and Patrick, without whom none of this would be possible. Thank you for everything (but mostly for life). ACKNOWLEDGEMENTS Foremost, I wish to express my sincere gratitude to my advisor, Dr. Justin Romberg, for his support of my work and for his patience, time, and immense knowledge. Under his direction, my technical and communication skills improved dramatically.
    [Show full text]