Aggregation Techniques on Software Metrics: A Study

Metrics are usually defined on a micro level like methods, classes and packages. These are failed to provide an adequate picture of the entire system effectively. By combine different metrics with varying output values and ranges to get insight in the evolution of the macro level system. We listed various metrics like Product, Project & Process and also various aggregation techniques in Traditional methods such as mean, median, sum, and cardinality; in Distribution fittings such as Log-Normal, Exponential, Negative binomial; and in Inequality Indices such as Theil, Gini, Kolm and Atkinson. The theoretical criteria in Domain, Range, and Invariance & Decomposability of various metrics are discussed. The Aggregation Techniques from simple mathematical operations to more complex operations to get Macro level system would be helpful as the outliers get pulled into the larger amounts of data. The Developers or Managers have an understanding of the parts of system are still needed to make sure that the metrics not misused or misunderstood. Metris are powerful tools that need to be used with care. We wish to understand, the aggregation techniques influence the strength of the relations among metrics to asses software quality.