Adv Data Anal Classif (2014) 8:321–338
DOI 10.1007/s11634-014-0166-6
REGULAR ARTICLE
Simplicial band depth for multivariate functional data
Sara López-Pintado · Ying Sun · Juan K. Lin · Marc G. Genton
Received: 26 April 2013 / Revised: 26 January 2014 / Accepted: 10 February 2014 /
Published online: 5 March 2014
© Springer-Verlag Berlin Heidelberg 2014
Abstract We propose notions of simplicial band depth for multivariate functional
data that extend the univariate functional band depth. The proposed simplicial band
depths provide simple and natural criteria to measure the centrality of a trajectory
within a sample of curves. Based on these depths, a sample of multivariate curves can
be ordered from the center outward and order statistics can be defined. Properties of the
proposed depths, such as invariance and consistency, can be established. A simulation
study shows the robustness of this new definition of depth and the advantages of
using a multivariate depth versus the marginal depths for detecting outliers. Real data
examples from growth curves and signature data are used to illustrate the performance
and usefulness of the proposed depths.
Keywords Band depth · Functional boxplot · Functional and image data ·
Modified band depth · Multivariate · Simplicial
Mathematics Subject Classification 62F07 · 62M10
S. López-Pintado
Department of Biostatistics, Columbia University, New York, NY 10032, USA
e-mail: sl2929@columbia.edu
Y. Sun
Department of Statistics, The Ohio State University,
Columbus, OH 43210, USA
e-mail: sunwards@stat.osu.edu
J. K. Lin
SearchForce, Inc., San Mateo, CA 94403, USA
e-mail: juan.k.lin@gmail.com
M. G. Genton (B )
CEMSE Division, King Abdullah University of Science and Technology,
Thuwal 23955-6900, Saudi Arabia
e-mail: marc.genton@kaust.edu.sa
123