MATH0033 Numerical Methods 2019
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
MATH0033
Numerical Methods
2019
1. Consider the numerical solution of f(x) = 0 where f(x) = e/2 - x - 1, which has a unique non-zero root a satisfying a E [2, 3].
(a) Show that it is possible to approximate the root a using the bisection method with initial interval [2, 3], and calculate the number of iterations required for the absolute error to be smaller than 10-5 .
You may assume without proof that 24/3 < e < 3.
(b) Let <f>(x) = 2log (x + 1) for x > -l.
Show that x = <f>(x) is consistent with f(x) = 0, and, using an appropriate the orem from lectures, prove that the fixed point iteration xk+1 = <f>(xk) converges linearly to a for all initial guesses x0 E [2, 3], and satisfies an error bound
where O < Ck < 1 is a k-dependent constant you should specify.
(c) Let rp : IR ----+ IR be continuously differentiable and let (3 E IR be a fixed point of rp s.t. lrp'(/3)1 > 1. Prove that the iteration xk+1 = rp(xk), k = 0, 1, 2, ..., does not converge to (3 for any initial guess x0 E IR, unless xk = (3 for some k.
Hint: argue by contradiction, using the mean value theorem.
2. Consider the stationary iterative method
x lk+ = Bxk + c, k = 0,l,2,...,
for the numerical solution of the linear system
Ax=b.
Here A, B E ffi.n X n and x,b,C E ]Rn for some n E N.
(a) Given a vector norm II · II : ffi.n -----r IR, write down the definition of the induced matrix norm II · II : ffi.n x n -----r R
(b) Assuming the method is consistent, prove that if IIBII < 1 then llx - xk ll ------r 0
ask -----r oo.
(c) What does it mean for the matrix A to be strictly diagonally dominant (SDD) by rows?
(d) Write down the definition of the iteration matrix B for the Jacobi method.
Prove that if A is SDD then the Jacobi method converges.
You may use without proof the fact that the matrix norm induced by the vector norm llvlloo := maxi=l,...,n lvil satisfies
n
IIMlloo = i1a,x.,n L lmijl,
(e) Now consider the system of nonlinear equations
fi(x1 ,x2) = Xi - x2 = 0,
h(x1 ,x2) = x + x1 - 1 = 0,
which can be written as f(x) = 0 with f = (11 ,hf and x = (x1 ,x2f.
Write down the Newton iteration for the solution of the system f(x) = 0. Give a sufficient condition on the initial guess xC0) guaranteeing that the first Newton step can be computed using a Jacobi iteration.
One root of the system lies in the quadrant x1 0, x2 0. Does the Newton iteration converge linearly or quadratically to this root (for a sufficiently good initial guess)?
3. (a) Consider the linear system Mx =b where
M=( µ2)
µ -/- 0 is a real parameter, and b E 2 .
Write down the Jacobi and the Gauss-Seidel methods for the iterative solution of the linear system in the form Pxk+1 =Nxk + b where P and N are matrices you should specify. Without computing the iteration matrices, give a sufficient condition on the parameter µ so that the Jacobi and the Gauss-Seidel methods are both convergent.
(b) Compute the iteration matrices BJ and Bes for the Jacobi and the Gauss
Seidel methods respectively. Using information from the iteration matrices, determine a necessary and sufficient condition onµ for the Jacobi and Gauss Seidel methods to be convergent. Which method is expected to converge faster?
(c) Let A be a symmetric positive definite matrix. Consider the gradient method for the solution of Ax = b, defined by xk+1 =xk + akrk , where rk =b - Axk and ak is the step length. Prove that the optimal choice of ak , minimising llek+1IIA, where ek =x - xk and llvll :=vTAv= (Av, v), is
4. For the solution of the Cauchy problem
consider the following family of numerical methods, parametrized by O 0 1:
n = 1,2,....
(a) For which 0 is the method explicit?
(b) State the values of 0 for which (*) coincides with the forward Euler, backward
Euler and Crank-Nicolson methods respectively.
(c) State the definition of truncation error for a general one-step method. Show that the method (*) is of second order if 0 = 1/2, and of first order otherwise. State clearly the smoothness assumptions under which your analysis is valid.
(d) State the definition of absolute stability for a general one-step method. Prove that the method (*) is unconditionally absolutely stable if 0 1/2 and condi tionally absolutely stable if 0 < 1/2. In the latter case, determine a necessary and sufficient condition on h for the method to be absolutely stable.
5. (a) Define the 2-norm condition number K2 (A) of an invertible matrix A E ffi.nxn _ Show that if Ax = b with b -/- 0 then for any x E ]Rn ,
(b) Write down a definition involving eigenvalues of what it means for a matrix A E ffi.nxn to be symmetric positive definite (SPD).
Prove that if A is SPD then A is invertible, A-1 is SPD, and
where Amax and Amin are the maximum and minimum eigenvalues of A.
(c) Consider the second order boundary value problem
where µ 0 and f is continuous on [0, 1], and the finite difference method
-5,Un+l + µun fn , n = l, ..., N - l,
where N E N, h = l/N, Xn = nh, fn = f(xn) and Un is the approximation of y(xn), for each n = 0, ..., N.
Write the method as an (N - 1)-by-(N - 1) linear system of the form Au= f, specifying the system matrix A and the vectors u and f.
(d) In the special case µ= 0 the eigenvalues of A are given by the formula
An = :2 (1 + COS ( n;)) , n = l, ..., N - l.
Use this fact to prove that
4N2
K2(A) = -2 + 0(1), N---+ oo.
In light of (a), what does this tell you about the usefulness of a residual-based stopping criterion when solving the system Au = f using an iterative method, when N is large?
2022-05-11