From: predictr@bellatlantic.net   
      
   Dephased wrote:   
   > Thanks for the answers. My variables are, for most of them, nominal.   
   > Originally, they were continuous numerical values but I discretized   
   > them using a fixed-witdh bin strategy.   
   >   
   > Finding a distance with nominal/discrete variables is what puzzles me   
   > especially since I am not trying to find a distance between two   
   > variables but between two vectors. I don't see how euclidean,   
   > mahalanobis or any other distance I heard of could make sense with   
   > this. (if it does, I would be more than happy to hear an explanation ;)   
   > )   
      
   If the original variables are continuous, why not bin by replacing with   
   the mean (or some other central value) of each bin?  This would permit   
   further calculation, such as distance computations.   
      
   -Will Dwinnell   
   http://will.dwinnell.com   
      
   [ comp.ai is moderated ... your article may take a while to appear. ]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)