[Scipy-tickets] [SciPy] #1486: 'sokalmichener' and 'rogerstanimoto' distances are the same

SciPy Trac scipy-tickets@scipy....
Sun Aug 7 12:30:13 CDT 2011


#1486: 'sokalmichener' and 'rogerstanimoto' distances are the same
---------------------------+------------------------------------------------
 Reporter:  muellner       |       Owner:  peridot    
     Type:  defect         |      Status:  new        
 Priority:  normal         |   Milestone:  Unscheduled
Component:  scipy.spatial  |     Version:  0.9.0      
 Keywords:                 |  
---------------------------+------------------------------------------------

Comment(by warren.weckesser):

 A bit of googling suggests that Sokal-Michener is the same as "simple
 matching", so the current implementation of this dissimilarity is
 incorrect.  It should be the same as spatial.distance.matching.  The only
 reference that I've found that shows Sokal-Michener dissimilarity being
 the same as Rogers-Tanimoto dissimilarity is the manuscript "Properties of
 Binary Vector Dissimilarity Measures" by Bin Zhang and Sargur N. Srihari
 (it is currently in the top five hits of a google search for "sokal-
 michener dissilimarity").  That paper could easily have a cut-and-paste
 error; all other references that I've seen refer to Sokal-Michener as
 "simple matching", and where formulas are given, the Sokal-Michener
 formula is the simple matching formula.

 The usual reference provided for Sokal-Michener dissimilarity is

 Sokal RR, Michener CD. 1958. A statistical method for evaluating
 systematic
     relationships. University of Kansas Science Bulletin 38: 1409-1438.

 The paper can be read online here:

 http://www.biodiversitylibrary.org/item/23745#page/387/mode/1up

 I haven't read it, so I can't say for sure that the reference is actually
 the correct citation for the Sokal-Michener dissimilarity.

-- 
Ticket URL: <http://projects.scipy.org/scipy/ticket/1486#comment:3>
SciPy <http://www.scipy.org>
SciPy is open-source software for mathematics, science, and engineering.


More information about the Scipy-tickets mailing list