Computational Statistics

, Volume 32, Issue 4, pp 1395–1410 | Cite as

Fast implementation of the Tukey depth

Original Paper
  • 141 Downloads

Abstract

Tukey depth function is one of the most famous multivariate tools serving robust purposes. It is also very well known for its computability problems in dimensions \(p \ge 3\). In this paper, we address this computing issue by presenting two combinatorial algorithms. The first is naive and calculates the Tukey depth of a single point with complexity \(O\left( n^{p-1}\log (n)\right) \), while the second further utilizes the quasiconcave of the Tukey depth function and hence is more efficient than the first. Both require very minimal memory and run much faster than the existing ones. All experiments indicate that they compute the exact Tukey depth.

Keywords

Tukey depth Quasiconcave Combinatorial property Fast computation 

Notes

Acknowledgments

The author thanks Prof. Mosler, K. and Dr. Mozharovskyi, P. for their valuable discussions during the preparation of this manuscript. The author also greatly appreciates two anonymous reviewers for their careful reading and insightful comments, which led to many improvements in this paper. This research is supported by NSFC of China (No. 11601197, 11461029, 71463020), NSF of Jiangxi Province (No. 20161BAB201024, 20151BAB211016), and the Key Science Fund Project of Jiangxi provincial education department (No. GJJ150439).

Compliance with ethical standards

Conflict of interest

I am the sole author of this manuscript. This research involves no human participants and/or animals, and has no conflict of interest.

References

  1. Cuesta-Albertos J, Nieto-Reyes A (2008) The random Tukey depth. Comput Stat Data Anal 52:4979–4988CrossRefMATHMathSciNetGoogle Scholar
  2. Donoho DL, Gasko M (1992) Breakdown properties of location estimates based on halfspace depth and projected outlyingness. Ann Stat 20:1808–1827MATHMathSciNetGoogle Scholar
  3. Edelsbrunner H (1987) Algorithms in combinatorial geometry. Springer, HeidelbergCrossRefMATHGoogle Scholar
  4. Hallin M, Paindaveine D, Šiman M (2010) Multivariate quantiles and multiple-output regression quantiles: from \(L_{1}\) optimization to halfspace depth. Ann Stat 38:635–669CrossRefMATHGoogle Scholar
  5. Härdle W, Simar L (2007) Applied multivariate statistical analysis. Springer, HeidelbergMATHGoogle Scholar
  6. Johnson T, Kwok I, Ng R (1998) Fast computation of 2-dimensional depth contours. In: Agrawal R, Stolorz P (eds) Proceedings of the 4th international conference on knowledge discovery and data mining. AAAI Press, New York, pp 224–228Google Scholar
  7. Kong L, Mizera I (2012) Quantile tomography: using quantiles with multivariate data. Stat Sin 22:1589–1610MATHMathSciNetGoogle Scholar
  8. Kong L, Zuo Y (2010) Smooth depth contours characterize the underlying distribution. J Multivar Anal 101:2222–2226CrossRefMATHMathSciNetGoogle Scholar
  9. Koshevoy H, Mosler K (1997) Zonoid trimming for multivariate distributions. Ann Stat 25:1998–2017CrossRefMATHMathSciNetGoogle Scholar
  10. Li J, Cuesta-Albertos JA, Liu RY (2012) DD-classifier: nonparametric classification procedure based on DD-plot. J Am Stat Assoc 107(498):737–753CrossRefMATHMathSciNetGoogle Scholar
  11. Liu X, Zuo Y (2014a) Computing halfspace depth and regression depth. Commun Stat-Simul Comput 43:969–985CrossRefMATHMathSciNetGoogle Scholar
  12. Liu X, Zuo Y (2014b) Computing projection depth and its associated estimators. Stat Comput 24(1):51–63CrossRefMATHMathSciNetGoogle Scholar
  13. Mosler K (2013) Depth statistics. In Robustness and Complex Data Structures. Springer, Berlin, pp 17–34Google Scholar
  14. Mosler K, Lange T, Bazovkin P (2009) Computing zonoid trimmed regions of dimension \(d > 2\). Comput Stat Data Anal 53:2500–2510CrossRefMATHMathSciNetGoogle Scholar
  15. Mozharovskyi P (2014) Contributions to depth-based classification and computation of the Tukey depth. PhD thesis. University of CologneGoogle Scholar
  16. Rousseeuw PJ, Ruts I (1996) Algorithm AS 307: bivariate location depth. J Appl Stat 45:516–526CrossRefMATHGoogle Scholar
  17. Rousseeuw PJ, Struyf A (1998) Computing location depth and regression depth in higher dimensions. Stat Comput 8:193–203CrossRefGoogle Scholar
  18. Tsay RS (2010) Analysis of financial time series, 3rd edn. Wiley, HobokenCrossRefMATHGoogle Scholar
  19. Tukey JW (1975) Mathematics and the picturing of data. In Proceedings of the international congress of mathematicians on canadian mathematical congress, Montreal, pp 523–531Google Scholar
  20. Yeh A, Singh K (1997) Balanced confidence regions based on Tukey’s depth and the bootstrap. J R Stat Soc Ser B 59:639–652CrossRefMATHMathSciNetGoogle Scholar
  21. Zuo YJ (2003) Projection based depth functions and associated medians. Ann Stat 31:1460–1490CrossRefMATHMathSciNetGoogle Scholar
  22. Zuo YJ, Serfling R (2000) General notions of statistical depth function. Ann Stat 28:461–482CrossRefMATHMathSciNetGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2016

Authors and Affiliations

  1. 1.School of StatisticsJiangxi University of Finance and EconomicsNanchangChina
  2. 2.Research Center of Applied StatisticsJiangxi University of Finance and EconomicsNanchangChina

Personalised recommendations