The netCDF operators NCO version 3.1.0 are ready.
http://nco.sf.net (Homepage)
http://dust.ess.uci.edu/nco (Homepage "mirror")
Version 3.1.0 achieves two major milestones:
1. All arithmetic operators work and are bug-free in MPI mode.
2. netcdf4-alpha support works well and is numerically correct.
Using MPI, NCO can now distribute computations among multiple nodes
which share disk space. This is the first step toward truly efficient
distributed versions of NCO toward which we are working.
These more efficient versions will utilize parallel I/O techniques
from netcdf4 and/or pnetcdf.
We may refer to NCO version 3.1.0+ as SDO, Scientific Data Operators.
For now it is just the same old netCDF data, but with the HDF files
it now supports via netCDF4, and other upcoming features, the names
they are a changin'.
1. MPI-enabled operators have zero-known bugs! ZKB! (Daniel Wang)
2. Autotools builds (Harry Mangalam): fixes to configure for
--enable-netcdf4 and --enable-mpi
Benchmarks also work in these modes--try 'em out.
You'll be surprised---we were!
3. Supports netcdf4-alpha10. Currently reads/writes HDF, no other
netCDF4 features supported yet.
4. ncwa most rapidly varying dimension (MRV) averaging speed-up:
Complaints about ncwa being a slow averager (Brian, Byron... :))
are as old as NCO itself.
I've finally implemented an MRV-averaging acceleration algorithm.
This improves zonal- (longitude) and area- (latitude/longitude)
averaging throughput by a factor of five--ten on files which
adhere to CF conventions (t,z,y,x).
Improvement is seen in all MRV averages with no downside for
non-MRV averages. Should have done this long ago!
Recommend upgrade if you use ncwa a lot.
5. ncap has a random number generator: rnd_nbr()
Enjoy,
Charlie
--
Charlie Zender, surname@xxxxxxx, Department of Earth System Science
3228 Croul Hall, UC Irvine, Irvine CA 92697-3100. (949) 824-2987 :)
From owner-netcdfgroup@xxxxxxxxxxxxxxxx 19 Fri, Jun
Date: Fri, 19 Jun 1992 14:01:07 -0400 (EDT)
From: GRIFFIOEN@xxxxxxxxxxxxxxxxxxxx (Hank Griffioen 301-513-1669)
To: HDF-netcdf@xxxxxxxxxxxxx, owner-netcdfgroup@xxxxxxxxxxxxxxxx,
Subject: FWD: IDL support for CDF format
Received: by unidata.ucar.edu id AA12374
(5.65c/IDA-1.4.4 for netcdfgroup-send); Fri, 19 Jun 1992 12:00:58 -0600
Received: from nssdca.gsfc.nasa.gov by unidata.ucar.edu with SMTP id AA12369
(5.65c/IDA-1.4.4 for <netcdfgroup@xxxxxxxxxxxxxxxx>); Fri, 19 Jun 1992
12:00:55 -0600
Message-Id: <920619140107.20c01ac9@xxxxxxxxxxxxxxxxxxxx>
netcdfgroup@xxxxxxxxxxxxxxxx
X-Vmsmail-To:
SMTP%"HDF-netcdf@xxxxxxxxxxxxx",SMTP%"owner-netcdfgroup@xxxxxxxxxxxxxxxx"
From: SMTP%"davidf@xxxxxxxxxxxxxxxxxxxxxx" 19-JUN-1992 13:40:53.13
To: GRIFFIOEN@xxxxxxxxxxxxxxxxxxxx
CC:
Subj: IDL support for CDF format
Date: Fri, 19 Jun 1992 11:39:12 -0600
From: davidf@xxxxxxxxxxxxxxxxxxxxxx (David Fanning)
Message-Id: <9206191739.AA19364@xxxxxxxxxxxxxxxxxxxxxx>
To: GRIFFIOEN@xxxxxxxxxxxxxxxxxxxx
Subject: IDL support for CDF format
Hank
I wanted to let you know that David Stern is going to JPL
next week and one of the things he is announcing is IDL
support for the CDF and Net-CDF data formats. (We hope
CDF extensions will take care of HDF, but if not, we will
do HDF as well.) Our current plans are to have these formats
ready to go for a big release in September that will even
have an IDL debugger in it!!
David Fanning
IDL
