We added an option to the Regrid module that puts the data value at the
closest grid point. If more than one scattered data point share the same
grid point, then the values are averaged. It is very quick if you want to
give it a shot. Just set the radius = 0 for this option. Its documented
in 4.3 and I believe 4.2.
Jeff
At 12:55 PM 7/4/2003 -0400, you wrote:
Actually, that's a pretty small data set. The problem here is not the
number of points per se, it's the speed of the spatial sorting in Connect
and Regrid, if I recall. However, there was a discussion on the mail group
a year or so ago, about reimplementing some of that code and/or adding some
additional capability to address this problem. Jeff, David: do you know
what ever happened to those ideas?
FYI, I and others have used DX on reasonable new Intel/Linux workstations
with millions of gridded data points.
"å×ÇÅÎÉÊ óÁ×ÅÎËÏ×" <[EMAIL PROTECTED]>@opendx.watson.ibm.com on
07/04/2003 11:01:16 AM
Please respond to [email protected]
Sent by: [EMAIL PROTECTED]
To: <[email protected]>
cc:
Subject: [opendx-users] large data sets
Hello,
I've tried to visualize with OpenDX 2d scalar scattered data with large
number of points -- over 150000. I've used "Connect" and "AutoGrid" modules
and found out that it takes unacceptable time to construct grid structure
--
more then several hours. Does anybody knows efficient way to work with such
a large data sets? I'm using OpenDX 4.2.0, RedHat Linux 7.3.
Thank you,
Eugene Savenkov
