CSC_349_HANDOUT#16

# CSC_349_HANDOUT#16 - COMPUTER SCIENCE 349A Handout Number...

This preview shows pages 1–2. Sign up to view the full content.

COMPUTER SCIENCE 349A Handout Number 16 GAUSSIAN ELIMINATION WITH PARTIAL PIVOTING (Section 9.4) The Naïve Gaussian elimination algorithm will fail if any of the pivots K , , , ) 2 ( 33 ) 1 ( 22 11 a a a is equal to 0. Mathematically speaking, the algorithm only requires that the pivots be nonzero. However, as the algorithm fails when a pivot is exactly equal to 0, it often gives very poor numerical results (using floating-point arithmetic) when a pivot is close to 0. Example Consider the n = 2 linear system with augmented matrix 78 . 46 13 . 6 291 . 5 17 . 59 14 . 59 003 . 0 . The exact solution is = 1 10 x . However, using 4 decimal-digit floating-point rounding arithmetic and Naïve Gaussian elimination, a very inaccurate solution is computed: 6 2 2 ) 1 ( 2 6 2 2 ) 1 ( 22 10 1044 . 0 or 10 1044 )) 10 1044 ( 78 . 46 ( )) 17 . 59 1764 ( 78 . 46 ( 10 1043 . 0 or 10 1043 )) 10 1043 ( 13 . 6 ( )) 14 . 59 1764 ( 13 . 6 ( . 1764 ) 666 . 1763 ( ) 003 . 0 / 291 . 5 ( mult × × = × = × × × = × = × = = l l l l l l K l l f f f b f f f a f f That is, the reduced linear system (in which the coefficient matrix has been reduced to upper triangular form) is × × 2 2 10 1044 10 1043 0 . 0 17

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 12/10/2011 for the course CSC 349 taught by Professor Oadje during the Spring '11 term at University of Victoria.

### Page1 / 5

CSC_349_HANDOUT#16 - COMPUTER SCIENCE 349A Handout Number...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online