NetNews Usenet Archive 1992 #16/NN_1992_16.iso/spool/sci/math/symbolic/2038

Text File | 1992-07-21 | 1.5 KB | 36 lines

Newsgroups: sci.math.symbolic
Path: sparky!uunet!sun-barr!decwrl!access.usask.ca!kakwa.ucs.ualberta.ca!news
From: [e-mail]  (John Manuel)
Subject: I need help with slow Mma code
Message-ID: <[e-mail]>
Sender: [e-mail]
Nntp-Posting-Host: deimos.space.ualberta.ca
Reply-To: [e-mail]  (John Manuel)
Organization: University Of Alberta, Edmonton Canada
Date: Tue, 21 Jul 1992 22:50:45 GMT
Lines: 23

I've got a particularly slow section of Mma code which I've GOT to speed up.
Here's the offending section:

xyz = Select[xyz, ((MemberQ[xSubSampleA, #[[1,1]]]
                    &&
                    MemberQ[ySubSampleA, #[[1,2]]])
                   ||
                   (MemberQ[xSubSampleB, #[[1,1]]]
                    &&
                    MemberQ[ySubSampleB, #[[1,2]]])) &];

xyz is a 20,000 X 2 X 2 array of real numbers and the SubSample arrays are
each <100 elements in length and are also composed of real numbers. The code
extracts a subset of values from xyz (usually about 1,000 elements of the
20,000) for later processing, but does so quite slowly (~4 minutes in 2.0 on
a NeXT).

Does anyone know of a more efficient, but still concise, way to do this?

--
John R. Manuel                 Department of Physics, University of Alberta
[e-mail]                 Edmonton, Alberta, CANADA, T6G 2J1
Office: [phone]          (NeXTMail OK)           Fax: [phone]