Comparative measurements (on a much slower machine then I've tested on before)
Mind you: were grouping N groups ~ N nodes.
I just finished *comparing* the examples:
The first example I tried with 1000 (83 sec) 2000 (320 sec) and 4000 (1200 sec)
The second (recursive) example I tried with 1000 nodes and XALAN ran out of
stack space.
The third (binary tree) example I tried with 1000 (34 sec) 2000 (65 sec) and
4000 (150 sec)
So the first example is quadratic
The second does not apply
The third is linear but probably O(log(n)*n)
Cheers,
Robbert
----- Original Message -----
From: "Robbert van Dalen" <juicer(_at_)xs4all(_dot_)nl>
To: <xsl-list(_at_)lists(_dot_)mulberrytech(_dot_)com>
Sent: Sunday, March 23, 2003 1:22 AM
Subject: Re: [xsl] How efficient is DVC? - A grouping example
I've done some testing on 1000, 2000, 4000, and 8000 nodes and it seems that
it's growing linear (apart from the sorted step which is probably O(log(N)*N).
However sorting is, much much quicker (sort), so that doesn't show up in the
totals.
The timings include building the binary tree and getting the ranges of nodes.
Cheers,
Robbert
----- Original Message -----
From: "Michael Kay" <mhk(_at_)mhk(_dot_)me(_dot_)uk>
To: <xsl-list(_at_)lists(_dot_)mulberrytech(_dot_)com>
Sent: Saturday, March 22, 2003 11:38 PM
Subject: RE: [xsl] How efficient is DVC? - A grouping example
This is fascinating stuff, but the proof of the pudding is in the
eating: have you made any comparative performance measurements, using a
non-trivial input file?
Michael Kay
Software AG
home: Michael(_dot_)H(_dot_)Kay(_at_)ntlworld(_dot_)com
work: Michael(_dot_)Kay(_at_)softwareag(_dot_)com
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list