scientific computing seminar
DESCRIPTION
Scientific Computing Seminar. AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues *. * Work supported by NSF under grants NSF/ITR-0082094, NSF/ACI-0305120 and by the Minnesota Supercomputing Institute . C. Bekas: SC Seminar . Introduction and Motivation. Target Problem - PowerPoint PPT PresentationTRANSCRIPT
Scientific Computing Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues*
C. Bekas Y. SaadComp. Science & Engineering Dept.University of Minnesota, Twin Cities
* Work supported by NSF under grants NSF/ITR-0082094, NSF/ACI-0305120 and by the Minnesota Supercomputing Institute
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Introduction and Motivation
C. Bekas: SC Seminar
Target ProblemCompute a large number of the smallest eigenvalues of large sparse
matrices Numerous important applications, including:
• structural engineering• computational materials science (electronic structure calculations)• Signal/Image processing and Control
Shift and Invert techniques can be very successful(Grimes et al 94, MSC.NASTRAN) …
BUT quickly become impractical to use:
For very large problem sizes (…N>106) a supercomputer is needed
When we need to compute several hundreds or thousands of eigenvalues (deep in the spectrum) reorthogonalization costs dominate and become prohibitive!
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Introduction and Motivation
Component Mode Synthesis (CMS) (Hurty ’60, Graig-Bampton ’68)Well known alternative. Used for many years in Structural Engineering.But it too suffers from limitations due to problem size…
AMLS, (Bennighof, Lehoucq, Kaplan and collaborators) Multilevel CMS method (solves the dimensionality problem) Automatic computation of substructures (easy application) Approximation: Truncated Congruence Transformation Builds very large projection basis without reorthogonalization Successful in computing thousands of eigenvalues in vibro-acoustic analysis (N>107) in a few hours on workstations (Kropp–Heiserer, 02)
Accuracy issues AMLS accuracy is adequate in Structural Eng. (in the order of the discretization error) , but higher accuracy is needed in other applications, (i.e. electronic structure calculations) AMLS is an one shot approach: no (iterative) refinement is done
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Recent Advances
The success of AMLS in Struct. Eng. and its potential for other applications has sparked several new research initiatives.
Some include:
- Bekas and SaadPurely algebraic analysis of AMLS1) Approximation to a nonlinear (Schur) eigenvalue problem2) Approximation of the resolvent (A- I)-1 by a careful projection3) Improvements: a) 2nd order expansions, b) Krylov projections and combinations
- Yang et alAlgebraic substructuring. Careful selection of added eigenvectors for improved accuracy.
- Elssel and VossA priori error bounds for algebraic substructuring.
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
In this talk
We will describe
- Approximation mechanism behind AMLS. We will review our recent purely algebraic analysis of AMLS (Bekas – Saad, 04).
- Iterative application of AMLS1) Refinement of approximated eigenvalues2) Multiple shifts in AMLS3) Computation of eigenvalues in an interval
- Numerical examples
C. Bekas: SC Seminar
1 2
Subdivide into 2 subdomains: 1 and 2
Component Mode Synthesis: a model problem
Y
X
Consider the model problem:
on the unit square . We wish to compute smallest eigenvalues.
Component Mode Synthesis
• Solve problem on each i
• “Combine” partial solutions
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
CMS: Approximation procedure. Ignore coupling!
Component Mode Synthesis: Approximation
CMS: Approximation procedure (2)
• Solve B v = v • Remember that B is block diagonal• Thus: we have to solve smaller decoupled eigenproblems
• Then, CMS approximates the coupling among the subdomains by the application of a carefully selected operator on the interface unknowns
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
Recent CMS method: AMLS
Block Gaussian eliminator matrix: such that:
Where S = C – E> B-1 E is the Schur Complement
Equivalent Generalized Eigenproblem : UTAUu= UTU u
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
AMLS Approximation
AMLS: Approximation procedure. Ignore coupling!
Solve decoupled problems Form basis
FINAL APPROXIMATION:
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
AMLS: Multilevel application
S1
E1
E1*
B2
B1
S2
S3
Scheme applied recursively.Resulting to thousands of subdomains. Successful in computing thousands smallest eigenvalues in vibro-acoustic analysis with problem size N>107
In the following we analyze the approximation mechanism of one step of AMLS, adopting a purely algebraic setting.
This will naturally lead to improved versions of the method
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
AMLS: approximation to a nonlinear eigenproblem!
AMLS: Approximation procedure. DO NOT ignore coupling this time
Leads to:
Substitute equation (1) in equation (2) to give equivalent non-linear problem:
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
AMLS: approximation to a nonlinear eigenproblem!
Define:
Resolvent:
Basic Property
We can show that:
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
AMLS: approximation to a nonlinear eigenproblem!
Neumann Series of the Resolvent:
Truncate the series. Keep the first two terms only. Then…
Approximate problem:
Remember:
AMLS:
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
AMLS: The projection view-point
Initial problem:
We can show that if: Then, is eigenvalue of (1) and
Respective eigenvector
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
AMLS: The projection view-point
1. AMLS solves approximately (truncation)
4. Remedy: augment the space of approximants by eigenvectors of B. Why?
2. AMLS, change of basis:
3. Thus:
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
AMLS: The projection view-point
We examine the difference:
Expansion in terms of eigenvectors of B:
Then, this difference is:
Thus: the difference is large in eigenvectors of B with eigenvalues close to the smallest eigenvalues of A.
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
AMLS: The projection view-point
Therefore: augmenting the space of approximants with ‘’smallest’’eigenvectors of B is well justified. Can we bound the error?
Let: mB “smallest” eigenvectors of B
XB: restriction to the B-part (upper part) of the space of approximants. Then:
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Improvements
The algebraic framework we have described leads to improvements:
Nonlinear Schur complement problem: Introduce an additional term of the truncated Neumann series Utilize corresponding 2nd order projection
Approximation of the resolvent (B- I)-1: Approximate with Krylov subspaces on B-1 and Utilize corresponding projection
Combinations of the above improvements lead to hybrid algorithms with enhanced stability and robustness
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Augmenting with Krylov Subspaces
We need to approximate
Therefore it is natural to consider the Krylov subspace
Vk is an orthonormal block Krylov basis
The columns of US are eigenvectors of the nonlinear Schur complement problem
or in general the block Krylov subspace many j
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Second Order Approximation
Neumann Series of the Resolvent:
AMLS approximation
Can we do better? 2nd order approximation
Quadratic Eigenvalue Problem:
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Second Order Approximation: Solving the QED
Quadratic Eigenvalue Problem:
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Linearization leads to equivalent generalized eigenproblem:
Second Order Approximation: Projection view point
Eigenvector of A:Better approx. by the QEP
We can add more eigenvectors of B…or add “second order” vectors:
ADD THE VECTORS
Construct basis
Solve
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Second Order Approximation: Error bounds
Augmenting the space of approximants with eigenvectors of B and “second order vectors” leads to quadratic error bounds compared to adding just eigenvectors
Let: mB smallest eigenvectors of B
XB: restriction to the B-part (upper part) of the space of approximants. Then:
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Improvements
The algebraic framework we have described leads to improvements:
Our previous workNonlinear Schur complement problem:
Introduce an additional term of the truncated Neumann seriesApproximation of the resolvent (B- I)-1:
Approximate with Krylov subspaces on B-1 and Utilize corresponding projection
Recent Work Framework for the iterative refinement of the AMLS approximations Utilization of many different shifts (A-k I)-1 that can be used for… …computation of eigenvalues deep in the spectrum More robust method. Currently examining connections with shift-invert Lanczos (Grimes et al) and Rational Krylov (Ruhe)
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
Iterative Refinement (1/3)
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
OR WITH SOME REARRANGEMENT…
Iterative Refinement (2/3)
NON-LINEAR SCHUR COMPLEMENT
AMLS APPROXIMATIONR: AMLS REMAINDER
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
Iterative Refinement (3/3)
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
Using Multiple Shifts in AMLS (1/3)
MATRIX: BCSSTK111st shift: 3, close to eigenvalues in the interval [2.9,3]2nd shift: 69, close to eigenvalues in the interval [68,70]
CAN WE COMBINE THESE TWO SHIFTS?
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
Using Multiple Shifts in AMLS (2/3)
Smallest eigenvectors of
Smallest eigenvectors of
PROJECTED PROBLEM
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
Using Multiple Shifts in AMLS (3/3)
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
Numerical Experiments
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
Numerical Experiments: Standard v.s. Krylov
Thus: Small /favors the Krylov version
Matrix: BCSSTK11(N=1473), Reorder using Nes. Disec., size of Schur complement NS=94
Bef. reord./
After reord./
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Numerical Experiments: Standard v.s. Krylov
Number of Schur eigenvectors: 5Number of added vectors:
AMLS: 30 eigenvectors of B (left) and 40 eigenvectors of B (right) Krylov: 30=5 x 6 (left) and 40=5 x 8 (right)
REORDERED VERSION
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Numerical Experiments: Standard v.s. Krylov
Number of Schur eigenvectors: 5Number of added vectors:
AMLS: 30 eigenvectors of B (left) and 40 eigenvectors of B (right) Krylov: 30=5 x 6 (left) and 40=5 x 8 (right)
NOT REORDERED VERSION
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Second order method
Matrix: BCSSTK11Approximate S() u= u with a Quadratic eigenvalue problem instead of the original generalized eigenvalue problem of AMLS
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Second order method
Matrix: BCSSTK11Compare 2nd order AMLS (20 added vectors) v.s. standard AMLS with increasing number of added eigenvectors of B (mB=20,40,60,100)
1 2 3 4 5 6 7 8 9 10
10-8
10-6
10-4
10-2
Eigenvalues: smallest to larger
abs.
rela
tive
erro
r
AMLS, mB = 100AMLS, mB = 60AMLS, mB = 40AMLS, mB = 202nd order AMLS, mS = 10
2nd order: 10 Schur vectors: 20 added Krylov vectors. 30 vectors in total
AMLS: 10 Schur vectors: 20, 40, 60 and 100 added eigenvectors of B
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Combine AMLS with Krylov AMLS
1 2 3 4 5 6 7 8 9 1010
-6
10-4
10-2
100
102
Eigenvalues: smallest to larger
abs.
rela
tive
erro
r
AMLS, k=20Krylov AMLS, m=2
Matrix: 5-point stencil discretization of the LaplacianUse only the 2 smallest Schur vectors. Then, for larger eigenvalues the Krylov version can fail.
REMEDY: Augment subspace with eigenvectors of B too!
VB: eigenvectors of BVK: Krylov basisUS: Schur eigenvectors
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Combine AMLS with Krylov AMLS
Matrix: 5-point stencil discretization of the LaplacianUse only 2 smallest Schur vectors. Then, for larger eigenvalues the Krylov version can fail.
REMEDY: Augment subspace with eigenvectors of B too!
VB: eigenvectors of BVK: Krylov subspaceUS: Schur eigenvectors
1 2 3 4 5 6 7 8 9 1010
-6
10-5
10-4
10-3
10-2
10-1
Eigenvalues: smallest to larger
abs.
rela
tive
erro
r
AMLS, k=30Krylov - Standard AMLS, m=2, k=20
C. Bekas: SC Seminar
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
Iterative Refinement
Matrix: 5-point stencil discretization of the Laplacian
Next shift k is 1st smallest approximate eigenvalue at step k-1
Next shift k is 5th smallest approximate eigenvalue at step k-1
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
Multiple Shifts
Dimension of each Qi : 40
Matrix: 5-point stencil discretization of the Laplacian
Dimension of each Qi : 60
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar
Conclusions
AMLS is a promising alternative to Shift-Invert methods for very large problems
In this work we have presented an analysis of AMLS based on a completely algebraic framework:
• AMLS is a nonlinear (spectral) Schur complement method that utilizes…• Projection on carefully selected eigenspaces to approximate the solution
Improvements• Based on the algebraic framework we have proposed Krylov projection subspaces and second order approximations (and combinations) with significant improvements.
Iterative Refinement• Approximations can be iteratively refined, allowing for very good accuracy• Many different shifts are combined and thus we can compute eigenvalues deep in the spectrum
Current Work• Investigate strategies to compute all eigenvalues in a (large) interval [a,b]
AMLS, Spectral Schur Complements and Iterative Computation of Eigenvalues
C. Bekas: SC Seminar