That's right.  The bounding box is maintained and the original data are
(weighted-average) interpolated to the Reduce'd resolution grid.

For subsampling, you can use Slab/Slice/Stack to do this doing some
book-keeping in the process.  Here's an example of this idea with a macro.

(See attached file: SubSample.net)

I haven't used it in awhile and I wrote it over 4 years ago, so I can't
vouch for it working perfectly with OpenDX.  But it should help you get
started.

--------------------------
Lloyd A. Treinish
Deep Computing Institute
IBM Thomas J. Watson Research Center
P. O. Box 704
Yorktown Heights, NY 10598
914-784-5038 (voice)
914-784-7667 (facsimile)
[EMAIL PROTECTED]
http://www.research.ibm.com/people/l/lloydt/


"Yuzhong Shen" <[EMAIL PROTECTED]>@opendx.watson.ibm.com on 03/02/2000
03:08:32 AM

Please respond to [email protected]

Sent by:  [EMAIL PROTECTED]


To:   <[email protected]>
cc:
Subject:  [opendx-users] Sampling grids



The Reduce module computes a reduced-version of a field, and it uses box
filtering. However, I want to only subsample the field, and don't need the
filtering. What should I do? Thanks.

yuzhong


Attachment: SubSample.net
Description: Binary data

Reply via email to