Cookies help us deliver our services. By using our services, you agree to our use of cookies. More information

Difference between revisions of "Talk:Outlier-skewness index"

From Bioblast
Line 1: Line 1:
== Problems or solutions ==
== Problems or solutions ==


::::* ''SIAM'' and Pearson's 2 coefficient are zero when average equals median.
::::* ''OSI'' and Pearson's 2 coefficient are zero when average equals median.
::::* ''SIAM'' and Pearson's 2 coefficient are invariant when the data set is multiplied by a constant factor.
::::* ''OSI'' and Pearson's 2 coefficient are invariant when the data set is multiplied by a constant factor.
::::* ''SIAM'' but not Pearson's 2 coefficient correlates with coefficient of variation when the data set is baseline corrected by a constant factor.
::::* ''OSI'' but not Pearson's 2 coefficient correlates with the coefficient of variation (average/SD) when the data set is baseline corrected by a constant factor. ''OSI'' changes after a data set is baseline corrected by a constant factor, which changes ''x'' at a given SD. This does not affect Pearson's 2 index.
::::* ''SIAM'' changes after a data set is baseline corrected by a constant factor, which changes ''x'' at a given SD. This does not affect Pearson's 2 index.
::::* ''OSI'' changes from flux control ratio ''Y''/''Z'' to flux control efficiency 1-''Y''/''Z'', which changes ''x'' at a given SD. This does not affect Pearson's 2 index. In both cases the sign is changed.
::::* ''SIAM'' changes from flux control ratio ''Y''/''Z'' to flux control efficiency 1-''Y''/''Z'', which changes ''x'' at a given SD. This does not affect Pearson's 2 index. In both cases the sign is changed.

Revision as of 07:47, 8 June 2021

Problems or solutions

  • OSI and Pearson's 2 coefficient are zero when average equals median.
  • OSI and Pearson's 2 coefficient are invariant when the data set is multiplied by a constant factor.
  • OSI but not Pearson's 2 coefficient correlates with the coefficient of variation (average/SD) when the data set is baseline corrected by a constant factor. OSI changes after a data set is baseline corrected by a constant factor, which changes x at a given SD. This does not affect Pearson's 2 index.
  • OSI changes from flux control ratio Y/Z to flux control efficiency 1-Y/Z, which changes x at a given SD. This does not affect Pearson's 2 index. In both cases the sign is changed.