A program complexity metric based on data flow information in control graphs
Kuo‐Chung Tai · International Conference on Software Engineering · 1984
This paper presents a new approach to measuring program complexity with the use of data flow information in programs. A complexity metric, called DU, is defined for the control graph of a structured program. This new metric is different from other control-graph based metrics in that it is based on “representative” data flow information in a control graph. An algorithm for computing the value of DU(G) for a control graph G is given. The lower and upper bounds of DU(G) are provided. The DU metric is shown to have several advantages over other control-graph based complexity metrics.