Software equation
From Wikipedia, the free encyclopedia
In the study of software project estimation, the Software Equation is a model with multiple variables based on assumptions of a specific distribution of effort throughout the entire length of a software development project. The models basis was formed through analysis of productivity data collected from over 4000 modern day software development projects.[1] The software equation was derived from the Putnam-Norden-Rayleigh Curve which can be used to show the non-linear correlation between time to complete the project and applied human effort.[2]
Contents |
[edit] Model Form
The estimation model can be described in the following form.
where
- E=Project effort measured in person-months or person-years
- LOC=a Lines of Code estimate for the project
- t=Length of project measured in months or years
- B=a "special skills factor"[1]
- P=a "Productivity Parameter"[1]
[edit] Defining B
B, the special skills factor, is related to the size of the product.[3]
| B Value | Size of Project |
|---|---|
| .16 | 5-15K |
| .18 | 20K |
| .28 | 30K |
| .34 | 40K |
| .37 | 50K |
| .39 | >70K |
[edit] Defining P
Research from the collected productivity data supplies initial values for variable P determined by the type of software being developed. Some examples are listed below.[1] These numbers to not apply in all situations and P can often be calculated from previous metrics information.
| P Value | Description |
|---|---|
| 2,000 | Real time embedded software |
| 10,000 | Telecommunications software |
| 12,000 | Scientific software |
| 28,000 | Business system applications |
[edit] See also
[edit] References
- ^ a b c d Pressman, Roger S. [1982] (2005). Software Engineering - A Practitioner's Approach, 6th, Newyork, NY: McGraw-Hill, 662-663. ISBN 0-07-285318-2.
- ^ Pressman, Roger S. [1982] (2005). Software Engineering - A Practitioner's Approach, 6th, Newyork, NY: McGraw-Hill, 679. ISBN 0-07-285318-2.
- ^ The Computational Software Equation. Retrieved on 2007-09-06.
