Estimation of Functional Relationship Model for Circular Variables and Its Application in Measurement Problems

This paper discusses on the functional relationship model for circular variables which may be used in the comparing two sets of circular data subjected to measurements or unobservable errors, for instance the measurements of wind direction by two different techniques or at two different levels. As opposed to the functional relationship model for linear or real line data, the proposed model is rather complicated. However, parameter estimation may be obtained numerically using iteration procedure. Further differences observed between the functional model for real line data sets are highlighted in the estimation process. Model verification and its behavior will be examined via simulation. The model is illustrated with an application to the analysis of wind direction data that have been collected from Bayan Lepas Airport, Malaysia measured at two different levels.