A Note on QR-Based Model Reduction: Algorithm, Software, and Gravitational Wave Applications

While the proper orthogonal decomposition (POD) is optimal under certain norms, its also expensive to compute. For large matrix sizes, the QR decomposition provides a tractable alternative. Under the assumption that it is a rank-revealing QR (RRQR), the approximation error incurred is similar to the POD error; furthermore, the authors show the existence of an RRQR with exactly the same error estimate as POD. To numerically realize an RRQR decomposition, they discuss the (iterative) modified Gram Schmidt with pivoting and reduced basis methods. They show that these two seemingly different approaches are equivalent. They then describe an MPI/OpenMP parallel code that implements one of the QR-based model reduction algorithms analyzed, and document the codes scalability for large problems, such as gravitational waves, and demonstrate excellent scalability up to 32,768 cores and, for complex dense matrices, as large as 10,000x3,276,800.

[1]  Rory Smith,et al.  Accelerated gravitational wave parameter estimation with reduced order modeling. , 2014, Physical review letters.

[2]  Enrique S. Quintana-Ortí,et al.  Parallel Algorithms for Computing Rank-Revealing QR Factorizations , 1997 .

[3]  B. A. Boom,et al.  GW170104: Observation of a 50-Solar-Mass Binary Black Hole Coalescence at Redshift 0.2. , 2017, Physical review letters.

[4]  David F. Gleich,et al.  Model Reduction With MapReduce-enabled Tall and Skinny Singular Value Decomposition , 2013, SIAM J. Sci. Comput..

[5]  Christian H. Bischof,et al.  A Parallel QR Factorization Algorithm with Controlled Local Pivoting , 1991, SIAM J. Sci. Comput..

[6]  N. Nguyen,et al.  A general multipurpose interpolation procedure: the magic points , 2008 .

[7]  G. Golub,et al.  Linear least squares solutions by householder transformations , 1965 .

[8]  Å. Björck.  Solving linear least squares problems by Gram-Schmidt orthogonalization , 1967 .

[9]  Chad R. Galley,et al.  Towards beating the curse of dimensionality for gravitational waves using reduced basis , 2012, 1205.6009.

[10]  Michael Purrer,et al.  Fast and accurate inference on gravitational waves from precessing compact binaries , 2016, 1604.08253.

[11]  A. Patera,et al.  A PRIORI CONVERGENCE OF THE GREEDY ALGORITHM FOR THE PARAMETRIZED REDUCED BASIS METHOD , 2012 .

[12]  Walter Hoffmann,et al.  Iterative algorithms for Gram-Schmidt orthogonalization , 1989, Computing.

[13]  Wolfgang Dahmen,et al.  Convergence Rates for Greedy Algorithms in Reduced Basis Methods , 2010, SIAM J. Math. Anal..

[14]  Daniel B. Szyld,et al.  The many proofs of an identity on the norm of oblique projections , 2006, Numerical Algorithms.

[15]  The Ligo Scientific Collaboration,et al.  GW170817: Observation of Gravitational Waves from a Binary Neutron Star Inspiral , 2017, 1710.05832.

[16]  B. A. Boom,et al.  GW170814: A Three-Detector Observation of Gravitational Waves from a Binary Black Hole Coalescence. , 2017, Physical review letters.

[17]  P. Strevens.  Iii , 1985 .

[18]  W. Marsden.  I and J , 2012 .

[19]  Joel A. Tropp,et al.  Column subset selection, matrix factorization, and eigenvalue optimization , 2008, SODA.

[20]  Efstratios Gallopoulos,et al.  Parallelism in Matrix Computations , 2016 .

[21]  Danny C. Sorensen,et al.  Nonlinear Model Reduction via Discrete Empirical Interpolation , 2010, SIAM J. Sci. Comput..

[22]  Frank Ohme,et al.  Twist and shout: A simple model of complete precessing black-hole-binary gravitational waveforms , 2013, 1308.3271.

[23]  Ilse C. F. Ipsen,et al.  On Rank-Revealing Factorisations , 1994, SIAM J. Matrix Anal. Appl..

[24]  Ronald DeVore,et al.  Greedy Algorithms for Reduced Bases in Banach Spaces , 2012, Constructive Approximation.

[25]  L. Trefethen,et al.  Numerical linear algebra , 1997 .

[26]  Gene H. Golub,et al.  Numerical methods for solving linear least squares problems , 1965, Milestones in Matrix Computation.

[27]  Per Christian Hansen,et al.  Some Applications of the Rank Revealing QR Factorization , 1992, SIAM J. Sci. Comput..

[28]  C. Pan,et al.  Rank-Revealing QR Factorizations and the Singular Value Decomposition , 1992 .

[29]  Michael Purrer,et al.  Frequency-domain gravitational waves from nonprecessing black-hole binaries. II. A phenomenological model for the advanced detector era , 2015, 1508.07253.

[30]  A. Pinkus.  n-Widths in Approximation Theory , 1985 .

[31]  Michael Purrer,et al.  Frequency-domain gravitational waves from nonprecessing black-hole binaries. I. New numerical waveforms and anatomy of the signal , 2015, 1508.07250.

[32]  Robert H. Halstead,et al.  Matrix Computations , 2011, Encyclopedia of Parallel Computing.

[33]  Michael Purrer,et al.  Parametrized tests of the strong-field dynamics of general relativity using gravitational wave signals from coalescing binary black holes: Fast likelihood calculations and sensitivity of the method , 2017, 1712.08772.

[34]  Axel Ruhe.  Numerical aspects of gram-schmidt orthogonalization of vectors , 1983 .

[35]  James Demmel,et al.  Communication Avoiding Rank Revealing QR Factorization with Column Pivoting , 2015, SIAM J. Matrix Anal. Appl..

[36]  Zvonimir Bujanovic,et al.  On the Failure of Rank-Revealing QR Factorization Software -- A Case Study , 2008, TOMS.

[37]  Y. Wang,et al.  First search for gravitational waves from known pulsars with advanced LIGO , 2017, 1701.07709.

[38]  Ludmil T. Zikatanov,et al.  Some observations on Babu\vs}ka and Brezzi theories , 2003, Numerische Mathematik.

[39]  Julien Langou,et al.  Rounding error analysis of the classical Gram-Schmidt orthogonalization process , 2005, Numerische Mathematik.

[40]  Sjsu ScholarWorks,et al.  Rank revealing QR factorizations , 2014 .

[41]  Ricardo H. Nochetto,et al.  Two-Step Greedy Algorithm for Reduced Order Quadratures , 2012, J. Sci. Comput..

[42]  Gerhard Wellein,et al.  Introduction to High Performance Computing for Scientists and Engineers , 2010, Chapman and Hall / CRC computational science series.