[netcdfgroup] netCDF Operators NCO version 4.6.8

• To: netCDF Mail List <netcdfgroup@xxxxxxxxxxxxxxxx>, NCO Announce Mail List <nco-announce@xxxxxxxxxxxxxxxxxxxxx>
• Subject: [netcdfgroup] netCDF Operators NCO version 4.6.8
• From: Charlie Zender <zender@xxxxxxx>
• Date: Wed, 16 Aug 2017 11:13:42 -0700

The netCDF Operators NCO version 4.6.8 are released.

http://nco.sf.net (Homepage, Mailing lists)
http://github.com/nco (Source Code, Releases, Developers)

What's new?
Many new features in 4.6.8 are, as usual, related to climatologies and
regridding. ncclimo can now be given an explicit list of seasons to
compute instead of, or in addition to, MAM,JJA,SON,DJF. ncremap has
improved mask and format handling. There are also a collection of
minor bugfixes and improvements, including JSON tweaks, ISO8601
printing, and GCC7 and netCDF 4.5.x compatibility.

This release includes an important ncwa fix that prevents incorrect
answers when hyperslabs are used with masks and/or weights.
Upgrade if you use ncwa with -d hyperslabs.

PLEASE READ THIS IF YOU USE NCKS TO PRINT DATA:
The ncks default printing will change to CDL in NCO 4.7.0.
For 20+ years ncks has, by default, printed the text representation
of a file in what we now call "traditional" mode. This mode best
facilitates meticulous data examination in line-by-line format.
ncdump produces CDL format that is more useful for most NCO users.
ncdump has always printed clean CDL and for many years there was
little point in defaulting ncks printed output to CDL.
However, ncks CDL mode now rivals ncdump in many ways.
In particular, ncks --cln now prints times as human-readable calendar
dates, the last ncdump feature that I used which ncks lacked.
Hence in NCO 4.7.0 ncks default printed output will change to CDL.
Then one will type simply "ncks" instead of "ncks --cdl".
We just added a new "--trd" option to print traditional output.
Add --trd to your scripts and their behavior will not change.
Otherwise your printing scripts will start to print CDL.
You have been warned :)

Work on NCO 4.7.0 has commenced. Planned changes include printing
CDL by default, and planned improvements include support for conda
installs on MS Windows, and more ncclimo and ncremap features.

Enjoy,
Charlie

NEW FEATURES (full details always in ChangeLog):

A. ncap2 now fully implements NCO chunking maps and policies.
   Previously ncap2 preserved existing chunking in variables,
   but could not be told to do anythin different. Now all command-line
   chunking behavior supported by NCO works in ncap2.
   ncap2 --cnk_plc=unchunk -S cnk.nco in.nc4 out.nc4
   http://nco.sf.net/nco.html#cnk

B. ncks CDL printing supports finer-grained control of date formats,
   including an ISO 8601 "T" option. Previously ncks printed
   UDUnits-compliant times as dates when invoked with the --cdl and
   --cal options. A third option, --dt_fmt, now exposes finer
   control of the format with short, regular, and ISO8601 options:
   ncks -H -m -v time_bnds -C --cdl --dt_fmt=1 ~/nco/data/in.nc
   ncks -H -m -v time_bnds -C --cdl --dt_fmt=2 ~/nco/data/in.nc
   ncks -H -m -v time_bnds -C --cdl --dt_fmt=3 ~/nco/data/in.nc
   dt_fmt:   Output:
   0,1       1964-03-13 09:08:16 (Default, short format)
   2         1964-03-13 09:08:16.000000 (regular format)
   3         1964-03-13T09:08:16.000000 (ISO8601 "T" format)
   Note that --dt_fmt automatically implies --cal.
   http://nco.sf.net/nco.html#cln

C. Some data are best evaluated with custom-defined seasons, e.g., JFM
   instead of DJF, or two-month seasons such as FM or ON. ncclimo now
   supports up to eleven (and counting) seasons, although by default it
   only computes MAM, JJA, SON, and DJF. As of NCO 4.6.8, use the
   seasons option to specify additional or alternate seasons:
   ncclimo --seasons=jfm,jas,ann -s 1980 -e 1983 -i drc_in -o drc_out
   Use "–seasons=none" to completely turn-off seasonal and annual-mean
   climatologies:
   ncclimo --seasons=none -s 1980 -e 1983 -i drc_in -o drc_out
   http://nco.sf.net/nco.html#ncclimo

D. ncremap --msk_src and --msk_dst options now accept the value 'none'
   to prevent the regridder from interpreting any variable as a mask
   in files from which source or destination grids are inferred.
   ncremap -i ~/ed.nc --msk_src=none -d ~/ed.nc -o ~/foo.nc
   ncremap -i ~/ed.nc -d ~/ed.nc --msk_dst=none -o ~/foo.nc
   http://nco.sf.net/nco.html#ncremap

E. ncremap now accepts standard NCO arguments for output file type.
   The -3, -4, -5, -6, and -7 flags, or their long-option equivalents
   --fl_fmt=fmt or --file_format=fmt where fmt is classic, netcdf4,
   64bit_data, 64bit_offset, or netcdf4_classic (or some synonyms)
   causes the output files created by ncremap, e.g., the regridded
   files and the grid files, to have the requested format.
   The ERWG and TempestRemap weight-generators support only a subset
   of these formats.
   ncremap -7 -i ~/deepthroat.nc -d ~/wapo.nc -o ~/foo.nc
   http://nco.sf.net/nco.html#fl_fmt_ncremap

BUG FIXES:

A. The ncap2 implementation of mibs()/mabs()/mebs() is fixed.
   Thanks to Dominique Briand for reporting this.

B. JSON now prints "null" instead of "NaN" for non-normal
   floating-point values like NaN and Infinity.
   This makes output fully JSON-compliant, although it simultaneously
   makes it impossible for JSON parsers to determine whether a "null"
   floating-point value is NaN, or +/-Infinity.
   Thanks to Bob Simons for noticing this.

C. Fix longstanding (since ~4.2.3) ncwa issue where hyperslabs
   were not correctly handled for masks and weights. Hyperslabs
   that started after the first array element could lead to incorrect
   answers. Thanks to Tony Leboissetier for reporting the problem,
   and to Pedro Vicente for fixing it.          
   http://nco.sf.net#bug_ncwa_hyp_msk_wgt

D. Ensure ncclimo and ncremap load netCDF libraries 4.4.x+ on
   certain clusters so that CDF5 capabilities always available.
   Thanks to Milena Veneziani for reporting this.


--
Charlie Zender, Earth System Sci. & Computer Sci.
University of California, Irvine 949-891-2429 )'(