Software Metrics
0 likes447 views
1. Software metrics are used to quantify attributes of software objects by mapping symbols to those objects. There are different types of metrics that can be calculated. 2. Important metrics include McCabe's cyclomatic complexity metric, which measures the number of linearly independent paths through a program, and Halstead's software science metrics which measure properties like program volume and difficulty. 3. Product metrics measure attributes of the software independent of how it was produced, like lines of code. Process metrics measure attributes relating to the development process.
Software Metrics
- 1. Software MetricsEhsan hessamiIslamic Azad University OF Qazvin
- 2. What ?Software measurement is the mapping of symbols to objects
- 3. What ?Software measurement is the mapping of symbols to objectsWhy ?!!!!!!quantify some attribute of the objects
- 4. Criteria
- 5. CriteriaA metric must allow differententitiesto be distinguished.A metric must obey a representationcondition
- 6. Criteria…Each unit of the attribute must contribute an equivalent amount to the metricDifferent entities can have thesameattributevalue
- 7. Many timesthe attribute of interest is notdirectly measurable . An indirect measure involves ameasure and a predictionformula
- 9. Software Measurement Theory2 Kinds :numerical relation systemempirical relation system
- 10. The empirical relation systemtwo parts:A set of entities, EA set of relationships, R
- 11. The numerical relation systemtwo parts:A set of entities, NEXP : Integers , RealA set of relationships, PEXP : Less , Equal
- 12. MappingM, maps (E, R) to (N, P)related in the empirical systemrelated in the numberical system
- 13. MappingM, maps (E, R) to (N, P)related in the empirical systemrelated in the numberical systemEXP : x rel y if M(x) rel M(y)
- 14. EXP . Mappingmeasure, BIGX height=< Y height and X weight =< Y weight
- 15. MONOTONICITYIF increase the value of attribute , the object Will not Reverse.
- 16. MONOTONICITYIF increase the value of attribute , The object will not Reverse. Exp : 2y = XMonotonicityX>0 And X<52y = 5x - xIncreaesX>=5 And X<10DecreaseNot Monotonicity
- 17. MEASUREMENT SCALESThere are five different scale types :nominalordinalintervalratioabsolute
- 18. NormalThis type basically assigns numbers or symbols without regard to any quantity
- 19. NormalThis type basically assigns numbers or symbols without regard to any quantityEXP:
- 20. Ordinalthere is an implied ordering of the entities by the numbers assigned to the entity
- 21. Ordinalthere is an implied ordering of the entities by the numbers assigned to the entityEXP: First – SecondNot Equal1Second - third23
- 22. Intervalthe amount of the difference is constant
- 23. Intervalthe amount of the difference is constantEXP: converting from a Celsius to a FahrenheitFormule : 9.5 x +32
- 24. Ratiothe amount of the difference is constant and Well Zero
- 25. absoluteThe absolute is a counting scale measureEXP:
- 26. Product MetricsProduct metrics are metrics thatcan be calculated from the document independent of how it was produced . The most basic metric for size is thelinesof code metric .
- 27. McCABE’S CYCLOMATIC NUMBERintroduced in 1976 . Use the Gragh Theory : C = E – N + 2C = Cyclomatic number .E = number of Edges .N = number of Eodes .
- 28. McCABE’S CYCLOMATIC NUMBEREXP : 51432E = 11 N = 8C = 11 – 8 + 2 C = 5
- 29. McCABE’S CYCLOMATIC NUMBERCalculating the cyclomatic number from control flow graphs is time-consuming . McCabe found a more directMethod of calculating his measure .
- 30. McCABE’S CYCLOMATIC NUMBERRestriction : IN source Code : 1 decision IF , For , While Case or multiple Branch 1 less decision
- 31. McCABE’S CYCLOMATIC NUMBERRestriction … : Control flow graphs are required to have a distinct starting node and a distinct stoppingnodeFormula : C = D + 1D = Sum of Decisions
- 32. ExampleE = 2D = 1E = 2D = 1E = 3D = 2D = 4 C = 5Complexity
- 33. Threshold ValueIf the Complexity is greater than 10, efforts should be made to reduce the value or to split the module
- 34. HALSTEAD’S SOFTWARE SCIENCEHe did his work in the late 1960s and 1970s.He empirically looked for measuresof intrinsic size
- 35. HALSTEAD’S SOFTWARE SCIENCEvariablesOperandsconstantscommasparenthesesOperatorarithmetic operatorsbrackets
- 36. Basic Measures—η1 and η2η = η1 + η2 η =total count of unique tokensCount of unique Operatorη1 =Count of unique Operands η2 =
- 37. Basic Measures—η1 and η2Exp :z = 0;While x> 0 z = z + y ; x = x + 1;End –while ;Print(z) ;Operator = ; > + - Print () While-endwhileOperand z x y 0 1η1 =8 η2 = 5 η = 13
- 38. Length(N)N = N1 + N2 N =total count of tokensCount of OperatorN1 =Count of OperandsN2 =
- 39. LengthEXP : Operators= 3; 5While 1> 1+ 1- 1Print 1() 1OperandsZ 4X 3Y 20 21 1z = 0;While x> 0 z = z + y ; x = x + 1;End –while ;Print(z) ;N = N1 + N2 = 14+12 = 26
- 40. Estimate of the Length(est N)η1η2est N = η1*log + η2*log22
- 41. Estimate of the Length(est N)η1η2est N = η1*log + η2*log22Exp :η1=8 , η2=5est N = 8*3 +5*2.32 = 35.6
- 42. Volume ( V )Volume = number of bits(η1+ η2)V = N * Log2
- 43. Volume ( V )Volume = number of bits(η1+ η2)V = N * Log2Exp: N = 26 , η1=8 , η2=513V = 26 * Log = 96.22
- 44. Potential Volume, V*Potential Volume = Min number of bitsV* = (2 + η2*)log(2+ η2*)2η2* = The Minimal number of Operands
- 45. Other ParameterImplementation Level : L = V* / V
- 46. Other ParameterImplementation Level : L = V* / VEffort : E = V / L unit of Effors is emd
- 47. Other ParameterImplementation Level : L = V* / VEffort : E = V / L unit of Effors is emdTime :T = E / S18 emd / sec
- 48. HENRY–KAFURA INFORMATION FLOWmeasure the intermodule complexity of source code . HK = Weight *(out * in )iiiiIn = into the moduleiOut = out the module iWeight = weight the module i
- 49. HENRY–KAFURA INFORMATION FLOWExp : Weight = 1 HK = 1110