Navigation ...
|
|
|
Scientific Writing ...
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Other Writing ...
|
|
|
|
|
|
Reference details
Author(s)
| Year
| Title
| Reference
| View/Download
|
Tim Hopkins , Les Hatton | 2008b | Exploring defect correlations in a major Fortran numerical library | Submitted for publication | NAG01_01-08.pdf |
Synopsis and invited feedback
This work was or is being reviewed by domain-specific experts appointed independently.
If you would like to provide feedback just e-mail me here.
Synopsis
| Invited Feedback
| Importance (/10, author rated :-) )
|
This is an analysis of a mature numerical Fortran software library comparing its defect
history against a wide variety of well-known static measurements
searching for significant correlation. We found that most of these
static measurements individually correlate moderately well with a high degree of
significance with the number of defects in a particular component but
principle component analysis reveals that defect prediction is influenced by most to some modest extent but no particular factor dominated.
The analysis revealed a number of interesting properties however including the existence of negative correlations which may suggest that
programmers take more care in certain scenarios and less care in others.
We go on to show that when static measurements are averaged across components, strong evidence of a logarithmic relationship with defect emerges. | None yet | 10 |
Related links
Related papers and links
|
Sorry, no links registered in database yet. |
Auto-generated: $Revision: 1.63 $, $Date: 2020/01/25 16:18:09 $, Copyright Les Hatton 2001-
|