next up previous
Next: Locality of Data ( Up: Coupling Previous: Data Abstraction Coupling (


Instability ($ I$ )

Works with all instances of a common meta-model, regardless if they where produced with the Java or the UML front-end. It does not relay on relations since it uses existing metric values for its calculation. By defining a view containing only packages as elements, the metric definition can ignore classes, methods and fields as part of its description, since the relations originating from them are not relevant.
Description
Instability between packages (I) measures the ratio between the outgoing and the total number of in- and outgoing couplings from classes inside the package from/to classes outside the package (coupling to classes external classes of the package, uses I definition of coupling). Each class counts only once. Zero if the package does not contain any classes or if external classes are not used by the package's classes. I is primarily applicable to object-oriented systems.
Scope
Package
View
$ {V}^{I} = ({G}^{I}, {R}^{I})$
  • Grammar $ {G}^{I} = (\{\textrm{package}^{I}\}, \emptyset, \textrm{package}^{I})$
  • Relations $ {R}^{I} = \{\emptyset\}$
  • Mapping $ \alpha^{I}$:
    $\displaystyle \alpha^{I}(\textrm{Package})$ $\displaystyle \mapsto$ $\displaystyle \textrm{package}^{I}$  

Definition
The $ I$value of a package $ p \in \textrm{package}^{I}$is defined:
$\displaystyle I(p)$ $\displaystyle =$ $\displaystyle \frac{p.Ce}{p.Ca + p.Ce}$  

Scale
Absolute.
Domain
Rationale in $ 0..1.0\infty$.
Highly Related Software Quality Properties
 
Re-Usability 2.4
is negatively influenced by coupling.
Understandability for Reuse 2.4.1:
A part of a system that has a high (outgoing) efferent coupling may be highly inversely related to understandability, since it uses other parts of the system which need to be understood as well.

Understandability decreases with increasing I.

Attractiveness 2.4.4:
Parts that have a high (outgoing) efferent coupling may be highly inversely related to attractiveness, since they are using other parts of the system which need to be understood as well, and represent dependencies.

Attractiveness decreases with increasing I.

Maintainability 2.6
decreases with increasing I.
Analyzability 2.6.1:
Parts that have a high (outgoing) efferent coupling may be highly inversely related to analyzability, since they are using other parts of the system which need to be analyzed as well.

Analyzability decreases with increasing I.

Changeability 2.6.2:
Parts that have a high (outgoing) efferent coupling may be inversely related to changeability, since they are using other parts of the system which might need to be changed as well.

Changeability decreases with increasing I.

Stability 2.6.3:
Parts of a system showing a high afferent (ingoing) coupling from other system parts may be inversely related to stability, since other parts are affected by them. Parts showing a high (outgoing) efferent coupling may be inversely related to stability, since they are using other parts of the system, which are can affect them.

Stability decreases with increasing I.

Testability 2.6.4:
Parts that have a high (outgoing) efferent coupling may be highly inversely related to testability, since they are using other parts of the system which increase the number of possible test paths.

Testability decreases with increasing I.

Portability 2.7
decreases with increasing I.
Adaptability 2.7.1:
Parts that have a high (outgoing) efferent coupling may be inversely related to adaptability, since they are using other parts of the system which might need to be adapted as well.

Adaptability decreases with increasing I.

Related Software Quality Properties
 
Functionality 2.1
is both negatively and positively influenced by coupling.
Interoperability 2.1.3:
Parts that have a high (outgoing) efferent coupling may be directly related to interoperability, since they are using/interacting with other parts of the system.

Interoperability might increase with increasing I.

Security 2.1.4:
Parts that have a high (outgoing) efferent coupling may be inversely related to security, since they can be affected by security problems in other parts of the system.

Security might decrease with increasing I.

Reliability 2.2
might decrease with increasing I.
Fault-tolerance 2.2.2:
Parts that have a high (outgoing) efferent coupling may be inversely related to fault-tolerance, since they can be affected by faults in other parts of the system.

Fault-Tolerance might decrease with increasing I.

Recoverability 2.2.3:
Parts that have a high (outgoing) efferent coupling may be inversely related to recoverability, since their data is distributed in other parts of the system making their recovery difficult.

Recoverability might decrease with increasing I.

Re-Usability 2.4
might decrease with increasing I.
Learnability for Reuse 2.4.2:
Parts that have a high (outgoing) efferent coupling may be inversely related to learnability, since they are using other parts of the system which need to be understood as well.

Learnability might decrease with increasing I.

Operability for Reuse - Programmability 2.4.3:
Parts that have a high (outgoing) efferent coupling may be inversely related to learnability, since they are using other parts of the system, which represent dependencies.

Programmability might decrease with increasing I.

Efficiency 2.5
might decrease with increasing I.
Time Behavior 2.5.1:
Parts that have a high (outgoing) efferent coupling may be inversely related to time behavior, since they are using other parts of the system, thus execution during test or operation does not stay local, but might involve huge parts of the system.

Time behavior might get worse with increasing I.

Resource Utilization 2.5.2:
Parts that have a high (outgoing) efferent coupling may be inversely related to resource utilization, since they are using other parts of the system, thus execution during test or operation does not stay local, but might involve huge parts of the system.

Resource utilization might get worse with increasing I.

References
 
  • I is discussed in [16],
  • it is implemented in the VizzAnalyzer Metrics Suite.
Since
Compendium 1.0

next up previous
Next: Locality of Data ( Up: Coupling Previous: Data Abstraction Coupling (
Copyright © 2008-2009, ARiSA AB. Phone: +46-(0)470 70 8495 Email: info@arisa.se Valid CSS! Valid XHTML 1.0 Transitional