11

Object Oriented MetricsObject Oriented Metrics

Ali Ali KamandiKamandiSharifSharif University of TechnologyUniversity of Technology

kamandi@ce.sharif.edukamandi@ce.sharif.edu

November 2006November 2006

22

OO Metrics: IntroductionOO Metrics: IntroductionMeasurement and metrics are key components of any Measurement and metrics are key components of any engineering discipline.engineering discipline.The use of metrics for OO systems has progresses much The use of metrics for OO systems has progresses much more slowly than the use of OO methods.more slowly than the use of OO methods.As OO systems become more pervasive, it is essential As OO systems become more pervasive, it is essential that software engineers have quantitative that software engineers have quantitative measurements for assessing the quality of designs at measurements for assessing the quality of designs at both the architectural and component levels.both the architectural and component levels.The intent of OO( or nonThe intent of OO( or non--OO) metrics:OO) metrics:•• To better understand the quality of the product.To better understand the quality of the product.•• To assess the effectiveness of the process.To assess the effectiveness of the process.•• To improve the quality of work performed at a To improve the quality of work performed at a

project level.project level.[Pressman 2001]

[Pressman]

33

Principles of Metrics Principles of Metrics

Metrics must be Metrics must be •• SimpleSimple•• ObjectiveObjective•• Easy to collect (automated)Easy to collect (automated)•• Easy to interpretEasy to interpret•• Hard to misinterpretHard to misinterpret

The selection of an extensive set of metrics will The selection of an extensive set of metrics will depend on the project's characteristics and depend on the project's characteristics and context context

[RUP]

44

Metrics for the OO Design ModelMetrics for the OO Design ModelWhitmire[WHI97] describes nine distinct and measurable Whitmire[WHI97] describes nine distinct and measurable characteristics of an OO design:characteristics of an OO design:• Size (Population, Volume, Length, Functionality)(Population, Volume, Length, Functionality)• Complexity (Interrelated classes)(Interrelated classes)• Coupling ( No. of Collaborations between classes or messages)( No. of Collaborations between classes or messages)• Sufficiency (the degree to which a design component possesses (the degree to which a design component possesses

features in its abstraction from the current application point ofeatures in its abstraction from the current application point of f view)view)

• Completeness ( same as sufficiency but consider multiple points ( same as sufficiency but consider multiple points of view for fully representation of problem domain)of view for fully representation of problem domain)

• Cohesion ( the degree to which the set of properties it possesses ( the degree to which the set of properties it possesses is part of the problem or design domain)is part of the problem or design domain)

• Primitiveness (the degree to which an operation is atomic)(the degree to which an operation is atomic)• Similarity (the degree to which two or more classes are similar)(the degree to which two or more classes are similar)• Volatility (the degree of change or churn in an artifact because (the degree of change or churn in an artifact because

of defects or changing requirements )of defects or changing requirements )

[Pressman]

55

Software MetricsSoftware Metrics

Some Traditional MetricsSome Traditional Metrics•• McCabe McCabe CyclomaticCyclomatic Complexity (CC)Complexity (CC)•• Lines of Code (LOC)Lines of Code (LOC)•• Comment Percentage (CP)Comment Percentage (CP)

Some ObjectSome Object--Oriented MetricsOriented Metrics•• Weighted Methods Per Class (WMC)Weighted Methods Per Class (WMC)•• Coupling Between Object Classes (CBO)Coupling Between Object Classes (CBO)•• Response for a Class (RFC)Response for a Class (RFC)•• Depth of Inheritance Tree (DIT)Depth of Inheritance Tree (DIT)•• Lack of Cohesion in Methods (LOCM)Lack of Cohesion in Methods (LOCM)

66

Researches on OO MetricsResearches on OO Metrics

ChidamberChidamber and and KemererKemerer: CK Metrics Suite, 1994: CK Metrics Suite, 1994CK suite validated by CK suite validated by BasiliBasili in 1996 and again by Tang in 1999in 1996 and again by Tang in 1999Many other objectMany other object--oriented metrics are derived from the CK suite oriented metrics are derived from the CK suite of objectof object--oriented metrics oriented metrics Lorenz and Kidd 1994Lorenz and Kidd 1994Harrison, Harrison, CounsellCounsell and and NithiNithi, MOOD Metric Suite, 1998, MOOD Metric Suite, 1998WhitmireWhitmire: OO Design Measurements, 1997: OO Design Measurements, 1997……

77

ClassClass--Oriented Metrics 1Oriented Metrics 1ChidamberChidamber and and KemererKemerer 1994:1994:•• WMC: weightedWMC: weighted methods per classmethods per class ((ΣΣ complexity of methods) complexity of methods)

(low)(low)Amount of effort required to implement and test a classAmount of effort required to implement and test a classComplexity of inheritance tree.Complexity of inheritance tree.Application specific and limiting potential reuse Application specific and limiting potential reuse

•• DIT: depth of the inheritance treeDIT: depth of the inheritance tree (Max length node to root) (Max length node to root) (low)(low)

Lower level classes inherit many methods. Lower level classes inherit many methods. Difficulties when attempting to predict the behavior of a class.Difficulties when attempting to predict the behavior of a class.Greater design complexity.Greater design complexity.Positive: many methods may be reusedPositive: many methods may be reused

•• NOC: number of childrenNOC: number of children ( subclasses that are immediately ( subclasses that are immediately subordinate to a class)subordinate to a class)

High: reuse, the amount of testHigh: reuse, the amount of test

[Pressman]

88

ClassClass--Oriented Metrics 2Oriented Metrics 2ChidamberChidamber and and KemererKemerer (Continue):(Continue):•• CBO:CBO: coupling between object classescoupling between object classes (No. of coll. in CRC) (No. of coll. in CRC)

(low)(low)High: reusability will decrease, complicated testing and High: reusability will decrease, complicated testing and modificationsmodificationsThis is consistent with the general This is consistent with the general guidlineguidline to reduce coupling.to reduce coupling.

•• RFC: response for a classRFC: response for a class (No. of methods in response set) (No. of methods in response set) (low)(low)

High: Testing and overall design complexity will increase.High: Testing and overall design complexity will increase.•• LCOM: lack of cohesion in methodsLCOM: lack of cohesion in methods (No of methods that (No of methods that

access one or more of the same attributes) (low)access one or more of the same attributes) (low)High: Complexity of class design, class might be better High: Complexity of class design, class might be better designed by breaking it into two or more separate classes.designed by breaking it into two or more separate classes.It is desirable to keep cohesion high; that is keep LCOM low.It is desirable to keep cohesion high; that is keep LCOM low.

[Pressman]

99

ClassClass--Oriented Metrics 3Oriented Metrics 3Lorenz and Kidd [LOR94]: Lorenz and Kidd [LOR94]: •• SizeSize--oriented:oriented: Counts of attributes and operations.Counts of attributes and operations.•• InheritanceInheritance--based:based: The manner in which operations are The manner in which operations are

reused through the class hierarchy.reused through the class hierarchy.•• InternalsInternals: Look at cohesion and code: Look at cohesion and code--oriented issues.oriented issues.•• Externals:Externals: Examine coupling and reuse.Examine coupling and reuse.A Sampling of Metrics:A Sampling of Metrics:•• CS:CS: class size (no. of operations) (no. of attributes) (low)class size (no. of operations) (no. of attributes) (low)

High: class has too much responsibility, reduce the High: class has too much responsibility, reduce the reusability, complicate implementation and testingreusability, complicate implementation and testing

•• NOO:NOO: number of operations overridden by a subclass (low)number of operations overridden by a subclass (low)•• NOA:NOA: number of operations added by a subclassnumber of operations added by a subclass•• SI:SI: specialization index = NOO * level / total no of methodsspecialization index = NOO * level / total no of methods

[Pressman]

1010

ClassClass--Oriented MetricsOriented Metrics

Method inheritance factor: inherited methods / total methodsMethod inheritance factor: inherited methods / total methodsCoupling factor: formulaCoupling factor: formulaPolymorphism factor: formulaPolymorphism factor: formula

The MOOD Metrics Suite: Harrison, The MOOD Metrics Suite: Harrison, CounsellCounsell and and NithiNithi, 1998, 1998

[Pressman]

1111

OperationOperation--Oriented MetricsOriented Metrics

average operation sizeaverage operation sizeoperation complexityoperation complexityaverage number of parameters per average number of parameters per operationoperation

Lorenz and Kidd [LOR94]:Lorenz and Kidd [LOR94]:

[Pressman]

1212

Testability MetricsTestability Metrics

encapsulation relatedencapsulation related•• lack of cohesion in methodslack of cohesion in methods•• percent public and protectedpercent public and protected•• public access to data memberspublic access to data members

inheritance relatedinheritance related•• number of root classesnumber of root classes•• fan in: multiple inheritancefan in: multiple inheritance•• number of children and depth of inheritance treenumber of children and depth of inheritance tree

Proposed by Binder [BIN94]:Proposed by Binder [BIN94]:

[Pressman]

1313

OO Project MetricsOO Project Metrics

number of scenario scriptsnumber of scenario scriptsnumber of key classesnumber of key classesnumber of subsystemsnumber of subsystems

Proposed by Lorenz and Kidd [LOR94]:Proposed by Lorenz and Kidd [LOR94]:

[Pressman]

1414

Req. and Use Case Model Metrics [RUP]Req. and Use Case Model Metrics [RUP]

Defects - number of defects, by severity (open, closed) Reported complexity (0-5, by analogy with COCOMO [BOE81])

Quality

Number of defects and change requests (open, closed) Volatility

Staff-time units (with production, change and repair separated) Effort

Number of requirementsNumber of Use Cases Number of Use Case Packages Number of scenarios, total and per use case Number of actors Length of Use Case (pages of event flow, for example)

Size

MetricsCharacteristic

1515

Requirements and Use Case Model Metrics (2)Requirements and Use Case Model Metrics (2)

Analysis • Scenarios realized in analysis model/total scenarios

Design • Scenarios realized in design model/total scenarios

Implementation • Scenarios realized in implementation model/total

scenarios Test

• Scenarios realized in test model (test cases)/total scenarios

Requirements-to-UC Traceability = Traceable to Use-Case Model/Total number of requirements

Traceability

Use Cases completed (reviewed and under configuration management with no defects outstanding)/use cases identified (orestimated number of use cases)

Completeness

MetricsCharacteristic

1616

OO Design Model Metrics [RUP]OO Design Model Metrics [RUP]

Number of defects and change requests (open, closed) Volatility

Staff-time units (with production, change and repair separated)

Effort

Number of classes Number of subsystems Number of subsystems of subsystems …Number of packages Methods per class, internal, external Attributes per class, internal, external Depth of inheritance tree Number of children

Size

MetricsCharacteristic

1717

OO Design Model Metrics (2) [RUP]OO Design Model Metrics (2) [RUP]

Number of classes in Implementation Model/number of classesTraceability

Number of classes completed/number of classes estimated (identified) Completeness

Number of defects, by severity (open, closed) Defects

Number of children Cohesion

Number of children Coupling between objects (class fan-out)

Coupling

Response For a Class (RFC): this may be difficult to calculate because a complete set of interaction diagrams is needed.

ComplexityQuality

MetricsCharacteristic

1818

OO Implementation Metrics (1) [RUP]OO Implementation Metrics (1) [RUP]

Number of defects and change requests (open, closed) Breakage for each corrective or perfective change, estimated (prior to fix) and actual (upon closure)

Volatility

Staff-time units (with production, change and repair separated) Effort

Number of classes Number of components Number of implementation subsystems Number of subsystems of subsystems …Number of packages Methods per class, internal, external Attributes per class, internal, external Size of methods, Size of attributes Depth of inheritance tree Number of children Estimated size at completion

Size

MetricsCharacteristic

1919

OO Implementation Metrics (2) [RUP]OO Implementation Metrics (2) [RUP]

Number of classes unit tested/number of classes in design model Number of classes integrated/number of classes in design model Active integration and system test time (accumulated from test process), that is, time with system operating (used for maturitycalculation)

Completeness

Number of defects, by severity, open, closed Defects

Number of children Lack of cohesion in methods (LCOM)

Cohesion

Number of children Coupling between objects (class fan-out) Message passing coupling (MPC) [Li and Henry 1993]

Coupling

Response For a Class (RFC)Cyclomatic complexity of methods*

ComplexityQuality

MetricsCharacteristic

2020

Test Metrics [RUP]Test Metrics [RUP]

Number of successful Test Cases in Test EvaluationSummary/Number of test cases

Traceability

Number of test cases written/number of test cases estimated Code coverage

Completeness

Defects — number of defects by severity, open, closed (these are defects raised against the test model itself)

Quality

Number of defects and change requests (open, closed)Volatility

Staff-time units for production of test cases, and so on Effort

Number of Test Cases, Test Procedures, Test Scripts Size

MetricsCharacteristic

2121

Change Management Metrics [RUP]Change Management Metrics [RUP]

Number of defects discovered/number of defects predictedCompleteness

Breakage (estimated, actual) for the implementation model subset.

Volatility

Defect repair effort, change implementation effort in staff-time units

Effort

Number of defects, change requests by severity and status, also categorized as number of perfective changes, number of adaptive changes and number of corrective changes.

Size

MetricsCharacteristic

2222

Process MetricsProcess Metrics

DurationDurationEffortEffortTraining timeTraining timeInspection timeInspection timeMeeting timeMeeting timeProcess ProblemProcess Problem

2323

Project MetricsProject Metrics

Cost/Schedule Control Cost/Schedule Control •• BCWS, Budgeted Cost for Work Scheduled BCWS, Budgeted Cost for Work Scheduled •• BCWP, Budgeted Cost for Work Performed BCWP, Budgeted Cost for Work Performed •• ACWP, Actual Cost of Work Performed ACWP, Actual Cost of Work Performed

ModularityModularity = average breakage (NCNB*) per perfective or = average breakage (NCNB*) per perfective or corrective change on implementation model corrective change on implementation model AdaptabilityAdaptability = average effort per perfective or corrective = average effort per perfective or corrective change on implementation model change on implementation model ……* NCNB is non* NCNB is non--comment, noncomment, non--blank code size.blank code size.

2424

Data Analysis: Example (1)Data Analysis: Example (1)

ProjectsProjects-- Project A: 46 Java Classes (Commercial Software)Project A: 46 Java Classes (Commercial Software)-- ObjectObject--Oriented ConstructsOriented Constructs

-- Project B: 1000 Java Classes (NASA Software)Project B: 1000 Java Classes (NASA Software)-- Excellent ObjectExcellent Object--Oriented ConstructsOriented Constructs

-- Project C: 1617 C++ Classes (NASA Software)Project C: 1617 C++ Classes (NASA Software)-- Good ObjectGood Object--Oriented ConstructsOriented Constructs

2525

Data Analysis: Example (2)Data Analysis: Example (2)

Correlation of Metrics over ProjectsCorrelation of Metrics over Projects

Metric 1 x Metric 2 = correlation between metric Metric 1 x Metric 2 = correlation between metric 1 and metric 2. 1 and metric 2.

0.830.830.260.260.350.35Project Project CC

0.750.750.110.110.280.28Project Project BB

0.470.470.430.430.830.83Project Project AA

RFCxWMCRFCxWMCCBOxWMCCBOxWMCCBOxRFCCBOxRFC

2626

Data Analysis: Example (3)Data Analysis: Example (3)

Regression ModelRegression Model-- RFC = RFC = ββ11WMC + WMC + ββ22CBO + ConstantCBO + Constant

0.700.7099

6.0396.0390.1480.1480.7920.792RFCRFCCC

0.600.6088

15.42715.4270.2000.2000.7320.732RFCRFCBB

0.700.7088

8.0368.0360.7770.7770.1310.131RFCRFCAA

RR22ConstantsConstantsββCBOCBOββWMCWMC

2727

ReferencesReferences

[PRE01] Pressman, R. S., Software Engineering, A Practitioners A[PRE01] Pressman, R. S., Software Engineering, A Practitioners Approach, pproach, 55thth Ed., McGrawEd., McGraw--Hill, 2001.Hill, 2001.[ROY98] Walker Royce 1998. [ROY98] Walker Royce 1998. Software Project Management: A Unified Software Project Management: A Unified Framework.Framework. Addison Wesley Longman.Addison Wesley Longman.[CHI94] [CHI94] ChidamberChidamber, S. R. and C. F. , S. R. and C. F. KemererKemerer, , ““A Metrics Suite for Object A Metrics Suite for Object Oriented DesignOriented Design””, IEEE Trans. Software Engineering, vol. SE, IEEE Trans. Software Engineering, vol. SE--20, no. 6, 20, no. 6, June 1994, pp. 476June 1994, pp. 476--493.493.[HAR98] Harrison, R. S.J. [HAR98] Harrison, R. S.J. CounsellCounsell, and R. V. , and R. V. NithiNithi, , ““An Evaluation of the An Evaluation of the MOOD set of ObjectMOOD set of Object--Oriented Software Metrics,Oriented Software Metrics,”” IEEE Trans. Software IEEE Trans. Software Engineering, vol. SEEngineering, vol. SE--24, no. 6, June 1998, pp. 49124, no. 6, June 1998, pp. 491--496.496.[LOR94] Lorenz, M. and J. Kidd, Object[LOR94] Lorenz, M. and J. Kidd, Object--Oriented Design Metrics, PrenticeOriented Design Metrics, Prentice--Hall, 1994.Hall, 1994.[WHI97] [WHI97] WhitmireWhitmire, S., Object, S., Object--Oriented Design Measurement, Wiley, 1997.Oriented Design Measurement, Wiley, 1997.[BIN94] Binder, R. V., [BIN94] Binder, R. V., ““ObjectObject--Oriented Software Testing,Oriented Software Testing,”” CACM, vol. 37, CACM, vol. 37, no. 9, no. 9, septemberseptember 1994, p. 29.1994, p. 29.

2828

Questions?Questions?