Quantifying internal attributes of functional programs

Abstract

There has been considerable interest in the use of metrics to quantify various quality attributes of software systems 1–4. However, much of the reported research has been concerned with systems which are implemented in imperative and (to a lesser extent) object-oriented languages. The aim of the research reported here was to determine a suite of metrics which could be used to assess the quality attributes of code written in functional languages. This is a necessary preliminary step to assessing programs written in functional languages in a quantitative manner. In this paper we identify a set of appropriate measures and suggest additional measures to cope with the more novel aspects of functional languages. Experiments to determine whether or not these suggested measures do have any correlation with quality attributes have been performed with imperative languages 5. Similar experiments with functional languages will also need to be performed.