You are on page 1of 1

Q UIZ 2

120202: ESM4A - N UMERICAL M ETHODS

Spring 2016

Prof. Dr. Lars Linsen


School of Engineering and Science
Jacobs University

February 17, 2016

Problem Q.2: Gaussian Elimination. (2+8+3+2=15 points)


Given a system of linear equations by
x + 10, 000y = 20, 000
10, 000x + 30, 000y = 70, 000
we want to solve it assuming a normalized floating point representation in a decimal
system with precision k = 4, i.e., base b = 10 and the mantissa uses up to k = 4 digits.
(a) Write the system in normalized floating point representation (with k = 4 and
b = 10).
(b) Apply Naive Gaussian Elimination to solve the linear equation system in the given
floating point representation. Show all computation steps and explain why we
have a loss of significant digits.
(c) Compute the residual error of the solution computed in (b).
(d) How would one compute the relative error of the solution?
Duration: 15 min.

You might also like