Keywords
component-based modelling, environmental modeling frameworks, invasiveness, framework, software metrics
Start Date
1-7-2010 12:00 AM
Abstract
Environmental modeling frameworks support scientific model development by providing an Application Programming Interface (API) which model developers use to implement models. This paper presents results of an investigation on the framework invasiveness of environmental modeling frameworks. Invasiveness is defined as the quantity of dependencies between model code and the modeling framework. This research investigates relationships between invasiveness and the quality of modeling code. Additionally, we investigate the relationship between invasiveness and two common framework designs (lightweight vs. heavyweight). Five metrics to measure framework invasiveness were proposed and applied to measure invasiveness between model and framework code of several implementations of Thornthwaite and the Precipitation-Runoff Modeling System (PRMS), two hydrological models. Framework invasiveness measurements were compared with existing software metrics including size (lines of code), cyclomatic complexity, and object-oriented coupling with generally positive correlations being found. We found that models with lower framework invasiveness tended to be smaller, less complex, and have lower coupling. In addition, the lightweight framework implementations of the Thornthwaite and PRMS models were less invasive than the heavyweight framework model implementations. Our initial results suggest that framework invasiveness is undesirable for model code quality and that lightweight frameworks may help reduce invasiveness.
Environmental Modeling Framework Invasiveness: Analysis and Implications
Environmental modeling frameworks support scientific model development by providing an Application Programming Interface (API) which model developers use to implement models. This paper presents results of an investigation on the framework invasiveness of environmental modeling frameworks. Invasiveness is defined as the quantity of dependencies between model code and the modeling framework. This research investigates relationships between invasiveness and the quality of modeling code. Additionally, we investigate the relationship between invasiveness and two common framework designs (lightweight vs. heavyweight). Five metrics to measure framework invasiveness were proposed and applied to measure invasiveness between model and framework code of several implementations of Thornthwaite and the Precipitation-Runoff Modeling System (PRMS), two hydrological models. Framework invasiveness measurements were compared with existing software metrics including size (lines of code), cyclomatic complexity, and object-oriented coupling with generally positive correlations being found. We found that models with lower framework invasiveness tended to be smaller, less complex, and have lower coupling. In addition, the lightweight framework implementations of the Thornthwaite and PRMS models were less invasive than the heavyweight framework model implementations. Our initial results suggest that framework invasiveness is undesirable for model code quality and that lightweight frameworks may help reduce invasiveness.
