Abstract: Longitudinal data can be viewed as a type of functional data. The functional viewpoint is not typical for most analysts of longitudinal data, but provides a route for powerful new insights. The potential of this approach is demonstrated through an analysis of periodicities in a microarray gene expression data set.
Key words and phrases: Classification, Fourier subspace, functional data analysis, longitudinal data, SiZer analysis, time-course micro-array experiments, yeast cell cycle.