MATH 5610/6860PRACTICE MIDTERM EXAMProblem 1. Let αn→ 0, xn= O(αn) and yn= O(αn).Show that xnyn= o(αn).Problem 2.(a) Write the Taylor expansion of ln(1+x) about x = 0 with the Lagrangeform of the residual term.(b) Assume the Taylor series for ln(1 + x) is truncated after the terminvolving x10and is used to approximate the number ln 2. What boundon the error can be given?Problem 3. Consider the number x = 26+ 2−16+ 2−19.(a) Write x in scientific base 2 (binary) notation of the form (1.b1b2b3. . . bn)×2S.(b) If IEEE single precision is used, the number of bits above is limited to23. What is x+(floating point number immediately above x) and x−(floating point number immediately below x).(c) What is fl(x) (the floating point representation of x, assuming roundto nearest)(d) What is machine precision in this system?(e) Verify that the relative error between x and fl(x) is less than machineprecision.Problem 4. Halley’s method for solving f(x) = 0 uses the iteration formulaxn+1= xn−fnf0n(f0n)2− (fnf00n)/2,where fn= f(xn) and so on. Show that this formula results from applyingNewton’s method to the function f /√f0.Problem 5. Consider the polynomial p(z) = z4+ 2x3+ 3x2+ 3z + 2.(a) Compute p(2) using Horner’s method(b) Compute p0(2) using Horner’s method(c) Write p(z) in the form p(z) = (z − 2)q(z) + r, specifying q(z) and r.Problem 6. Consider a smooth function f with the following values.x -1 0 1 2f(x) -1 0 3 -1(a) Write the polynomial interpolating f (x) at the first three nodes inLagrange form.(b) Use divided differences to find the polynomial p(x) interpolating f(x)in Newton form.(c) Assuming all derivatives of f are available, give an expression of theinterpolation error f(t) − p(t) for some t ∈ [−1, 2].12 MATH 5610/6860 PRACTICE MIDTERM EXAMProblem 7. Let f (x) be a function of x and x0, . . . , xnbe n + 1 distinctnodes. For k = 0, . . . , n, let pkbe the polynomial interpolating f at thenodes x0, x1, . . . , xk. Let q be the polynomial interpolating f at the nodesx1, . . . , xn. Show that:pn(x) = q(x) +x − xnxn− x0(q(x) −
View Full Document