A Deflated Version of the Conjugate Gradient Algorithm

Abstract

Abstract. We present a deflated version of the conjugate gradient algorithm for solving linear systems. The new algorithm can be useful in cases when a small number of eigenvalues of the iteration matrix are very close to the origin. It can also be useful when solving linear systems with multiple right-hand sides, since the eigenvalue information gathered from solving one linear system can be recycled for solving the next systems and then updated.Key words. conjugate gradient, deflation, multiple right-hand sides, Lanczos algorithm AMS subject classifications. 65F10, 65F15PII. S10648295983397611. Introduction. A number of recent articles have established the benefits of using eigenvalue deflation when solving nonsymmetric linear systems with Krylov subspace methods. It has been observed that significant improvements in convergence rates can be achieved from Krylov subspace methods by adding to these subspaces a few approximate eigenvectors associated with the eigenvalues closest to zero [2,4,7,8,13,14]. In practice, approximations to the eigenvectors closest to zero are obtained from the use of a certain Krylov subspace; then these approximations are dynamically updated using the new Krylov subspace. Results of experiments obtained from these variations indicate that the improvement in convergence over standard Krylov subspaces of the same dimension can sometimes be substantial, especially when the convergence of the original scheme is hampered by a small number of eigenvalues near zero; see, e.g., [2,8].In this paper we consider extensions of this idea to the conjugate gradient (CG) algorithm for the symmetric case. Our starting point is an algorithm recently proposed by Erhel and Guyomarc'h [5]. This is an augmented subspace CG method aimed at linear systems with several right-hand sides. Erhel and Guyomarc'h [5] propose an algorithm which adds one specific vector obtained from a subspace related to a previous right-hand side. We first extend this algorithm to one which handles an arbitrary block W of vectors. We note that introducing an arbitrary W into the Krylov subspace of CG has already been considered by Nicolaides in [9]. The algorithm introduced in this paper is mathematically equivalent to the one in [9]. Nicolaides's algorithm is directly derived from a deflated Lanczos procedure and uses the 3-term recurrence version of the conjugate gradient algorithm. The algorithm in this paper exploits the link between the Lanczos algorithm and the standard CG algorithm. The LDL T factorization of the projected system obtained from the same

An Efficient Preconditioned CG Method for the Solution of a Class of Layered Problems with Extreme Contrasts in the Coefficients

“…We specify an implementation such that the basis of our subspace consist of vectors with many zero elements. Related work has recently been presented in [21,35].…”

Alya: Multiphysics engineering simulation toward exascale

“…On one hand, the explicit schemes, which can be viewed as the most simple iterative method with a unique simple (Richardson) iteration per time step. On the other hand, iterative schemes such as GMRES, BiCGSTAB, CG [27] or Deflated CG [28,19] are chosen depending on the case. Both explicit and implicit schemes are illustrated in Figure 1.…”

Iterative numerical methods for sampling from high dimensional Gaussian distributions

“…. , s. This has the effect of improving the conditioning of the system as per a preconditioner Saad et al (1999). Specifically, let W = [w 1 w 2 · · · w s ] be the given orthonormal eigenvectors, and λ i , i = 1, .…”

“…The first of these two is the more natural, and in this one it is possible to address the question on whether we should deflate or not. In the article Saad et al (1999) a detailed analysis on how deflating is related to preconditioning is presented. So suppose that K m (Q, r 0 ) ⊥ w 1 , w 2 , .…”