FloatingPointNumbersandArithmetic

v r elative error is not defined for v 0 loosely

Info iconThis preview shows page 1. Sign up to view the full content.

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

Unformatted text preview: error is v v , and (unless v 0 ) the r elative error is .( v R elative error is not defined for v 0. ) Loosely, precision refers to the number of correct v v v v digits. More precisely it is log 10 decimal digits (or log 2 binary bits). v v N otice the errors in the e xamples above relative to the exact result: 1. A n absolute error of .00000043 relative to 133.333243 is about 3.2 109 or about 8.5 digits. 2. A n absolute error.000057 relative to 113.580157 is about 5.0 107 or about 6.3 digits. 3. A n absolute error.0004061881 relative to 1219.3254061881 is about 3.3 107 or about 6.5 digits. 4. A n absolute error.000001140624811... relative to 12.499991140624811... is about 9.1 108 or about 7.0 digits....
View Full Document

Ask a homework question - tutors are online