Right-invariant generalized metrics applied to rank correlation coefficients

In this paper we investigate properties of rank correlation coefficients that can be derived from right-invariant generalized metrics on the symmetric group. We prove some new inequalities between a number of generalized metrics, and we characterize the sample sizes for which several (right-invariant) rank correlation coefficients can equal zero. Using the Hausdorff generalized metric, we show how to construct circular rank correlation coefficients from regular rank correlation coefficients. In addition, we show how generalized triangle inequalities satisfied by generalized metrics on the symmetric group can be used to create new partial rank correlation coefficients (that measure the association between two variables controlling for the effect of a third one).

[1]  P. Sprent,et al.  Statistical Analysis of Circular Data. , 1994 .

[2]  J. Dieudonne.  Treatise on Analysis , 1969 .

[3]  Vladimir Estivill-Castro,et al.  Generating Nearly Sorted Sequences - The use of measures of disorder , 2004, CATS.

[4]  Richard W. Hamming,et al.  Error detecting and error correcting codes , 1950 .

[5]  Judith M. Tanur,et al.  A Note on the Partial Correlation Coefficient , 1971 .

[6]  G. Guilbaud.  Relation entre les deux coefficients de corrélation de rangs , 1980 .

[7]  Bernard Monjardet,et al.  Concordance between two linear orders: The Spearman and Kendall coefficients revisited , 1997 .

[8]  Elena Deza,et al.  Encyclopedia of Distances , 2014 .

[9]  Nicholas I. Fisher,et al.  Statistical Analysis of Circular Data , 1993 .

[10]  M. Kendall,et al.  Rank Correlation Methods , 1949 .

[11]  A. Sampson,et al.  Metrics on permutations useful for positive dependence , 1997 .

[12]  Derick Wood,et al.  Right Invariant Metrics and Measures of Presortedness , 1993, Discret. Appl. Math..

[13]  Bernard Monjardet,et al.  On the comparison of the Spearman and Kendall metrics between linear orders , 1998, Discret. Math..

[14]  P. Sen.  Estimates of the Regression Coefficient Based on Kendall's Tau , 1968 .

[15]  R. Forthofer,et al.  Rank Correlation Methods , 1981 .

[16]  D. Critchlow.  Metric Methods for Analyzing Partially Ranked Data , 1986 .

[17]  J. Gibbons.  Nonparametric measures of association , 1993 .

[18]  R. Shibata,et al.  PARTIAL CORRELATION AND CONDITIONAL CORRELATION AS MEASURES OF CONDITIONAL INDEPENDENCE , 2004 .

[19]  N. Fisher,et al.  A correlation coefficient for circular data , 1983 .

[20]  A. .. Lawrance.  On Conditional and Partial Correlation , 1976 .

[21]  Vladimir Estivill-Castro.  Sorting and measures of disorder , 1992 .

[22]  K. B. Lakshmanan,et al.  Measures of disorder and straight insertion sort with erroneous comparisons , 2011, Ars Comb..

[23]  Nicholas I. Fisher,et al.  Nonparametric measures of angular-angular association , 1981 .

[24]  A. Cayley,et al.  LXXVII. Note on the theory of permutations , 1849 .

[25]  B. Monjardet,et al.  Concordance et consensus d'ordres totaux: Les coefficients K et W , 1985 .

[26]  P. Diaconis.  Group representations in probability and statistics , 1988 .

[27]  R. Graham,et al.  Spearman's Footrule as a Measure of Disarray , 1977 .

[28]  M. Chiani.  Error Detecting and Error Correcting Codes , 2012 .

[29]  K. B. Lakshmanan,et al.  Bubble sort with erroneous comparisons , 2005, Australas. J Comb..

[30]  Mehdi Hassani,et al.  Derangements and Applications , 2003 .

[31]  K. B. Lakshmanan,et al.  Recursive merge sort with erroneous comparisons , 2011, Discret. Appl. Math..