Abstract
We introduce directional regularity, a new definition of anisotropy formultivariate functional data. Instead of taking the conventional view whichdetermines anisotropy as a notion of smoothness along a dimension, directionalregularity additionally views anisotropy through the lens of directions. Weshow that faster rates of convergence can be obtained through a change-of-basisby adapting to the directional regularity of a multivariate process. Analgorithm for the estimation and identification of the change-of-basis matrixis constructed, made possible due to the replication structure of functionaldata. Non-asymptotic bounds are provided for our algorithm, supplemented bynumerical evidence from an extensive simulation study. Possible applications ofthe directional regularity approach are discussed, and we advocate itsconsideration as a standard pre-processing step in multivariate functional dataanalysis.