# Comparison Of Rate Of Convergence Of Iterative Methods Philosophy Essay

The term “iterative method” refers to a wide range of techniques that use successive approximations to obtain more accurate solutions to a linear system at each step… In numerical analysis it attempts to solve a problem by finding successive approximations to the solution starting from an initial guess. This approach is in contrast to direct methods which attempt to solve the problem by a finite sequence of operations, and, in the absence of rounding errors, would deliver an exact solution Iterative methods are usually the only choice for non linear equations. However, iterative methods are often useful even for linear problems involving a large number of variables (sometimes of the order of millions), where direct methods would be prohibitively expensive (and in some cases impossible) even with the best available computing power.

Tell us what you need to have done now!

order now

Stationary methods are older, simpler to understand and implement, but usually not as effective Stationary iterative method are the iterative methods that performs in each iteration the same operations on the current iteration vectors.Stationary iterative methods solve a linear system with an operator approximating the original one; and based on a measurement of the error in the result, form a “correction equation” for which this process is repeated. While these methods are simple to derive, implement, and analyze, convergence is only guaranteed for a limited class of matrices. Examples of stationary iterative methods are the Jacobi method,gauss seidel method and the successive overrelaxation method.

The Nonstationary methods are based on the idea of sequences of orthogonal vectors Nonstationary methods are a relatively recent development; their analysis is usually harder to understand, but they can be highly effective These are the Iterative method that has iteration-dependent coefficients.It include Dense matrix: Matrix for which the number of zero elements is too small to warrant specialized algorithms. Sparse matrix: Matrix for which the number of zero elements is large enough that algorithms avoiding operations on zero elements pay off. Matrices derived from partial differential equations typically have a number of nonzero elements that is proportional to the matrix size, while the total number of matrix elements is the square of the matrix size.

The rate at which an iterative method converges depends greatly on the spectrum of the coefficient matrix. Hence, iterative methods usually involve a second matrix that transforms the coefficient matrix into one with a more favorable spectrum. The transformation matrix is called a preconditioner. A good preconditioner improves the convergence of the iterative method, sufficiently to overcome the extra cost of constructing and applying the preconditioner. Indeed, without a preconditioner the iterative method may even fail to converge.

Rate of Convergence

In numerical analysis, the speed at which a convergent sequence approaches its limit is called the rate of convergence. Although strictly speaking, a limit does not give information about any finite first part of the sequence, this concept is of practical importance if we deal with a sequence of successive approximations for an iterative method as then typically fewer iterations are needed to yield a useful approximation if the rate of convergence is higher. This may even make the difference between needing ten or a million iterations.Similar concepts are used for discretization methods. The solution of the discretized problem converges to the solution of the continuous problem as the grid size goes to zero, and the speed of convergence is one of the factors of the efficiency of the method. However, the terminology in this case is different from the terminology for iterative methods.

The rate of convergence of an iterative method is represented by mu (I?) and is defined as such:

Suppose the sequence{xn} (generated by an iterative method to find an approximation to a fixed point) converges to a point x, then

limn->[infinity] = |xn+1-x|/|xn-x|[alpha]=I?, where I?a‰?0 and I±(alpha)=order of convergence.

In cases where I±=2 or 3 the sequence is said to have quadratic and cubic convergence respectively. However in linear cases i.e. when I±=1, for the sequence to converge I? must be in the interval (0,1). The theory behind this is that for En+1a‰¤I?En to converge the absolute errors must decrease with each approximation, and to guarantee this, we have to set 0

In cases where I±=1 and I?=1 and you know it converges (since I?=1 does not tell us if it converges or diverges) the sequence {xn} is said to converge sublinearly i.e. the order of convergence is less than one. If I?>1 then the sequence diverges. If I?=0 then it is said to converge superlinearly i.e. it’s order of convergence is higher than 1, in these cases you change I± to a higher value to find what the order of convergence is. In cases where I? is negative, the iteration diverges.

Stationary iterative methods

Stationary iterative methods are methods for solving a linear system of equations. Ax=B. where is a given matrix and is a given vector. Stationary iterative methods can be expressed in the simple form

where neither nor depends upon the iteration count . The four main stationary methods are the Jacobi Method,Gauss seidel method, successive overrelaxation method (SOR), and symmetric successive overrelaxation method (SSOR).

1.Jacobi method:- The Jacobi method is based on solving for every variable locally with respect to the other variables; one iteration of the method corresponds to solving for every variable once. The resulting method is easy to understand and implement, but convergence is slow. The Jacobi method is a method of solving a matrix equation on a matrix that has no zeros along its main diagonal . Each diagonal element is solved for, and an approximate value plugged in. The process is then iterated until it converges. This algorithm is a stripped-down version of the Jacobi transformation method of matrix diagnalization.

The Jacobi method is easily derived by examining each of the equations in the linear system of equations in isolation. If, in the th equation

solve for the value of while assuming the other entries of remain fixed. This gives

which is the Jacobi method.

In this method, the order in which the equations are examined is irrelevant, since the Jacobi method treats them independently. The definition of the Jacobi method can be expressed with matrices as

where the matrices , , and represent the diagnol, strictly lower triangular, and strictly upper triangular parts of , respectively

Convergence:- The standard convergence condition (for any iterative method) is when the spectral radius of the iteration matrix

I?(D a?’ 1R) < 1.

D is diagonal component,R is the remainder.

The method is guaranteed to converge if the matrix A is strictly or irreducibly diagonally dominant. Strict row diagonal dominance means that for each row, the absolute value of the diagonal term is greater than the sum of absolute values of other terms:

The Jacobi method sometimes converges even if these conditions are not satisfied.

2. Gauss-Seidel method:- The Gauss-Seidel method is like the Jacobi method, except that it uses updated values as soon as they are available. In general, if the Jacobi method converges, the Gauss-Seidel method will converge faster than the Jacobi method, though still relatively slowly. The Gauss-Seidel method is a technique for solving the equations of the linear system of equations one at a time in sequence, and uses previously computed results as soon as they are available,

There are two important characteristics of the Gauss-Seidel method should be noted. Firstly, the computations appear to be serial. Since each component of the new iterate depends upon all previously computed components, the updates cannot be done simultaneously as in the Jacobi method. Secondly, the new iterate depends upon the order in which the equations are examined. If this ordering is changed, the components of the new iterates (and not just their order) will also change. In terms of matrices, the definition of the Gauss-Seidel method can be expressed as

where the matrices , , and represent the diagonal, strictly lower triangular, and strictly upper triangular parts of A, respectively.

The Gauss-Seidel method is applicable to strictly diagonally dominant, or symmetric positive definite matrices A.

Convergence:-

Given a square system of n linear equations with unknown x:

The convergence properties of the Gauss-Seidel method are dependent on the matrix A. Namely, the procedure is known to converge if either:

A is symmetric positive definite, or

A is strictly or irreducibly diagonally dominant.

The Gauss-Seidel method sometimes converges even if these conditions are not satisfied.

3.Successive Overrelaxation method:-

The successive overrelaxation method (SOR) is a method of solving a linear system of equations derived by extrapolating the gauss-seidel method. This extrapolation takes the form of a weighted average between the previous iterate and the computed Gauss-Seidel iterate successively for each component,

where denotes a Gauss-Seidel iterate and is the extrapolation factor. The idea is to choose a value for that will accelerate the rate of convergence of the iterates to the solution.

In matrix terms, the SOR algorithm can be written as

where the matrices , , and represent the diagonal, strictly lower-triangular, and strictly upper-triangular parts of , respectively.

If , the SOR method simplifies to the gauss-seidel method. A theorem due to Kahan shows that SOR fails to converge if is outside the interval .

In general, it is not possible to compute in advance the value of that will maximize the rate of convergence of SOR. Frequently, some heuristic estimate is used, such as where is the mesh spacing of the discretization of the underlying physical domain.

Convergence:-

Successive Overrelaxation method may converge faster than Gauss-Seidel by an order of magnitude. We seek the solution to set of linear equations

In matrix terms, the successive over-relaxation (SOR) iteration can be expressed as

where , , and represent the diagonal, lower triangular, and upper triangular parts of the coefficient matrix , is the iteration count, and is a relaxation factor. This matrix expression is not usually used to program the method, and an element-based expression is used

Note that for that the iteration reduces to the gauss-seidel iteration. As with the Gauss seidel method, the computation may be done in place, and the iteration is continued until the changes made by an iteration are below some tolerance.

The choice of relaxation factor is not necessarily easy, and depends upon the properties of the coefficient matrix. For symmetric, positive definite matrices it can be proven that will lead to convergence, but we are generally interested in faster convergence rather than just convergence.

4.Symmetric Successive overrelaxation:- Symmetric Successive Overrelaxation (SSOR) has no advantage over SOR as a stand-alone iterative method; however, it is useful as a preconditioner for nonstationary methods The symmetric successive overrelaxation (SSOR) method combines two successive overrelaxation method (SOR) sweeps together in such a way that the resulting iteration matrix is similar to a symmetric matrix it the case that the coefficient matrix of the linear system is symmetric. The SSOR is a forward SOR sweep followed by a backward SOR sweep in which the unknowns are updated in the reverse order. The similarity of the SSOR iteration matrix to a symmetric matrix permits the application of SSOR as a preconditioner for other iterative schemes for symmetric matrices. This is the primary motivation for SSOR, since the convergence rate is usually slower than the convergence rate for SOR with optimal ..

Non-Stationary Iterative Methods:-

1.Conjugate Gradient method:- The conjugate gradient method derives its name from the fact that it generates a sequence of conjugate (or orthogonal) vectors. These vectors are the residuals of the iterates. They are also the gradients of a quadratic functional, the minimization of which is equivalent to solving the linear system. CG is an extremely effective method when the coefficient matrix is symmetric positive definite, since storage for only a limited number of vectors is required. Suppose we want to solve the following system of linear equations

Ax = b

where the n-by-n matrix A is symmetric (i.e., AT = A), positive definite (i.e., xTAx > 0 for all non-zero vectors x in Rn), and real.

We denote the unique solution of this system by x*.

We say that two non-zero vectors u and v are conjugate (with respect to A) if

Since A is symmetric and positive definite, the left-hand side defines an inner product

So, two vectors are conjugate if they are orthogonal with respect to this inner product. Being conjugate is a symmetric relation: if u is conjugate to v, then v is conjugate to u.

Convergence:- Accurate predictions of the convergence of iterative methods are difficult to make, but useful bounds can often be obtained. For the Conjugate Gradient method, the error can be bounded in terms of the spectral condition number of the matrix . ( if and are the largest and smallest eigenvalues of a symmetric positive definite matrix , then the spectral condition number of is . If is the exact solution of the linear system , with symmetric positive definite matrix , then for CG with symmetric positive definite preconditioner , it can be shown that

where , and . From this relation we see that the number of iterations to reach a relative reduction of in the error is proportional to .

In some cases, practical application of the above error bound is straightforward. For example, elliptic second order partial differential equations typically give rise to coefficient matrices with (where is the discretization mesh width), independent of the order of the finite elements or differences used, and of the number of space dimensions of the problem . Thus, without preconditioning, we expect a number of iterations proportional to for the Conjugate Gradient method.

Other results concerning the behavior of the Conjugate Gradient algorithm have been obtained. If the extremal eigenvalues of the matrix are well separated, then one often observes so-called; that is, convergence at a rate that increases per iteration. This phenomenon is explained by the fact that CG tends to eliminate components of the error in the direction of eigenvectors associated with extremal eigenvalues first. After these have been eliminated, the method proceeds as if these eigenvalues did not exist in the given system, i.e., the convergence rate depends on a reduced system with a smaller condition number. The effectiveness of the preconditioner in reducing the condition number and in separating extremal eigenvalues can be deduced by studying the approximated eigenvalues of the related Lanczos process.

2. Biconjugate Gradient Method-The Biconjugate Gradient method generates two CG-like sequences of vectors, one based on a system with the original coefficient matrix , and one on . Instead of orthogonalizing each sequence, they are made mutually orthogonal, or “bi-orthogonal”. This method, like CG, uses limited storage. It is useful when the matrix is nonsymmetric and nonsingular; however, convergence may be irregular, and there is a possibility that the method will break down. BiCG requires a multiplication with the coefficient matrix and with its transpose at each iteration.

Convergence:- Few theoretical results are known about the convergence of BiCG. For symmetric positive definite systems the method delivers the same results as CG, but at twice the cost per iteration. For nonsymmetric matrices it has been shown that in phases of the process where there is significant reduction of the norm of the residual, the method is more or less comparable to full GMRES (in terms of numbers of iterations). In practice this is often confirmed, but it is also observed that the convergence behavior may be quite irregular , and the method may even break down . The breakdown situation due to the possible event that can be circumvented by so-called look-ahead strategies. This leads to complicated codes. The other breakdown situation, , occurs when the -decomposition fails, and can be repaired by using another decomposition.

Sometimes, breakdown or near-breakdown situations can be satisfactorily avoided by a restart at the iteration step immediately before the breakdown step. Another possibility is to switch to a more robust method, like GMRES.

3. Conjugate Gradient Squared (CGS ).

The Conjugate Gradient Squared method is a variant of BiCG that applies the updating operations for the -sequence and the -sequences both to the same vectors. Ideally, this would double the convergence rate, but in practice convergence may be much more irregular than for BiCG, which may sometimes lead to unreliable results. A practical advantage is that the method does not need the multiplications with the transpose of the coefficient matrix.

often one observes a speed of convergence for CGS that is about twice as fast as for BiCG, which is in agreement with the observation that the same “contraction” operator is applied twice. However, there is no reason that the “contraction” operator, even if it really reduces the initial residual , should also reduce the once reduced vector . This is evidenced by the often highly irregular convergence behavior of CGS . One should be aware of the fact that local corrections to the current solution may be so large that cancelation effects occur. This may lead to a less accurate solution than suggested by the updated residual. The method tends to diverge if the starting guess is close to the solution.

4 Biconjugate Gradient Stabilized (Bi-CGSTAB ).

The Biconjugate Gradient Stabilized method is a variant of BiCG, like CGS, but using different updates for the -sequence in order to obtain smoother convergence than CGS. Bi-CGSTAB often converges about as fast as CGS, sometimes faster and sometimes not. CGS can be viewed as a method in which the BiCG “contraction” operator is applied twice. Bi-CGSTAB can be interpreted as the product of BiCG and repeatedly applied GMRES. At least locally, a residual vector is minimized , which leads to a considerably smoother convergence behavior. On the other hand, if the local GMRES step stagnates, then the Krylov subspace is not expanded, and Bi-CGSTAB will break down . This is a breakdown situation that can occur in addition to the other breakdown possibilities in the underlying BiCG algorithm. This type of breakdown may be avoided by combining BiCG with other methods, i.e., by selecting other values for One such alternative is Bi-CGSTAB2 ; more general approaches are suggested by Sleijpen and Fokkema.

5..Chebyshev Iteration.

The Chebyshev Iteration recursively determines polynomials with coefficients chosen to minimize the norm of the residual in a min-max sense. The coefficient matrix must be positive definite and knowledge of the extremal eigenvalues is required. This method has the advantage of requiring no inner products. Chebyshev Iteration is another method for solving nonsymmetric problems . Chebyshev Iteration avoids the computation of inner products as is necessary for the other nonstationary methods. For some distributed memory architectures these inner products are a bottleneck with respect to efficiency. The price one pays for avoiding inner products is that the method requires enough knowledge about the spectrum of the coefficient matrix that an ellipse enveloping the spectrum can be identified ; however this difficulty can be overcome via an adaptive construction developed by Manteuffel , and implemented by Ashby . Chebyshev iteration is suitable for any nonsymmetric linear system for which the enveloping ellipse does not include the origin.

Convergence:-

In the symmetric case (where and the preconditioner are both symmetric) for the Chebyshev Iteration we have the same upper bound as for the Conjugate Gradient method, provided and are computed from and (the extremal eigenvalues of the preconditioned matrix ).

There is a severe penalty for overestimating or underestimating the field of values. For example, if in the symmetric case is underestimated, then the method may diverge; if it is overestimated then the result may be very slow convergence. Similar statements can be made for the nonsymmetric case. This implies that one needs fairly accurate bounds on the spectrum of for the method to be effective (in comparison with CG or GMRES).

Acceleration of convergence

Many methods exist to increase the rate of convergence of a given sequence, i.e. to transform a given sequence into one converging faster to the same limit. Such techniques are in general known as “series acceleration”. The goal of the transformed sequence is to be much less “expensive” to calculate than the original sequence. One example of series acceleration is Aitken’s delta -squared process.