Complexity
Measuring complexity
Halstead Complexity Measures Measures complexity at the function level like so:
- number of distinct operators (function calls, syntax elements)
- number of distinct operands (variables)
- total number of operators (how often you use those operands)
- total number of operands (how oftern you use those variables)
- By refactoring code into smaller methods, you can reduce the number of distinct operators and operands in a function, moving that complexity to another location. However do this too much and you would end up increasing the number of distinct operators.
Types of complexity in technology:
- branching complexity - how many different outcomes are there?
- data access complexity - are there permissions
Measuring complexity in object-oriented programming
- methods per class
- coupling between object classes
- response for a class(?)
- number of children
- depth of inheritance tree
- lack of cohesion of methods
(Chidamber, S.R.; Kemerer, C.F. IEEE Transactions on Software Engineering Volume 20, Issue 6, Jun 1994 Page(s):476 - 493)