Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!elroy.jpl.nasa.gov!decwrl!pa.dec.com!shlump.nac.dec.com!decuac!haven!ni.umd.edu!uc780.umd.edu!cs450a03
From: cs450a03@uc780.umd.edu
Newsgroups: comp.lang.apl
Subject: RE: JKT replies to EEM re: histograms
Message-ID: <4APR91.20043819@uc780.umd.edu>
Date: 4 Apr 91 20:04:38 GMT
References: <1991Apr4.002052.24661@yrloc.ipsa.reuter.COM>
Sender: usenet@ni.umd.edu (USENET News System)
Organization: The University of Maryland University College
Lines: 25
Nntp-Posting-Host: uc780.umd.edu

Joey Tuttle writes:
>
>     #/.~,y
> 
>is what I have asked for in the past in APL.  By making a special
>performance case of (#/.~), a problem which has always been
>"inappropriate for APL because of performance" could be solved very
>nicely.

what's wrong with

    v =. /:, y
    k =. k # i. $k=. 1, (_1}.v) ~: 1}. v
 hist =. k - ($k){. 0,k

here, /:  is O(n log n)  (where n is the length of v)
and everything is is O(n)

Also note that k{v  is the nub of ,y

Yeah, this version will blow up if y is empty, but that is easily
handled (either as a special case, by coercing k, or by choosing a
different algorithm to check adjecency when computing k).

Raul Rockwell