NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
The netCDF Operators NCO version 4.7.0 have escaped! http://nco.sf.net (Homepage, Mailing lists) http://github.com/nco (Source Code, Releases, Developers) What's new? Version 4.7.0 adds spit and polish throughout the toolkit. Nothing major stands out, except perhaps for easy access to compression algorithms in ncclimo and ncremap. Work on NCO 4.7.1 has commenced. Planned changes include better diagnosis and workarounds for the netCDF CDF5 bug, and parallel weight generation by ncremap. Enjoy, Charlie NEW FEATURES (full details always in ChangeLog): A. ncclimo and ncremap accept fl_fmt options for output file format. The option behaves the same as it does for other operators: ncclimo [-3 -4 -5 -6 -7] ... ncremap [-3 -4 -5 -6 -7] ... ncclimo --fl_fmt=netcdf4 ... ncremap --fl_fmt=cdf5 ... http://nco.sf.net/nco.html#fl_fmt http://nco.sf.net/nco.html#autoconversion B. ncclimo and ncremap accept the dfl_lvl option for compression. The value of dfl_lvl is an integer fro 0..9 that specifies the DEFLATE (= Lempel-Ziv + Huffman) compression level for netCDF4. The option behaves the same as it does for other operators: ncclimo --dfl_lvl=3 ... ncremap --dfl_lvl=3 ... ncclimo -L 3 ... ncremap -L 3 ... http://nco.sf.net/nco.html#deflate C. ncclimo accepts the ppc option for Precision-Preserving Compression. Other operators allow multiple, per-variable PPC specification. ncclimo allows only one PPC option, equivalent to specifying --ppc default=<val> with the compiled operators. NB: regridding does not yet accept PPC. ncclimo --ppc=3 ... http://nco.sf.net/nco.html#ppc http://nco.sf.net/nco.html#bitgrooming D. ncrename is now quiet by default, like all other operators. The old summary line that tallied the numbers of things renamed can still be seen by invoking ncrename with -D 1. http://nco.sf.net#ncrename E. CMake: Pedro Vicente contributed a CMake build-engine in 4.6.9. The CMake engine now correctly links ncap2 to intrinsic math functions like erf(), gamma(), ... Please give us feedback on any wrinkles in the CMake build. To build with CMake and install in /usr/local: cd nco/cmake cmake .. -DCMAKE_INSTALL_PREFIX=/usr/local make sudo make install Additional examples in cmake/build.bat http://nco.sf.net#bld F. The automatic CDF5 bug checker in 4.6.9 has been turned-off. It produced too much distracting output in workflows affected by the CDF5 bug. To activate this check, build NCO with, e.g., CPPFLAGS='-DNCO_CDF5_BUG_CHK' ./configure ... The checker itself has been improved so that, when turned on, operators are now quiet when sniffing CDF5 files for corruption, until they detect a "large" (> 4 GB) variable. In 4.6.9 the sniffing algorithm was noisier, and printed an INFO upon closing any CDF5 file. http://nco.sf.net#bug_cdf5 BUG FIXES: A. Fix bug where ncra and ncrcat treated negative record hyperslab indices with old (pre-2014) convention, and were off-by-one relative to the 2014+ Python-based convention. Previously "ncrcat -d time,-1" extracted the penultimate record, now it extracts the last. This bug only affected the record dimension of the multi-file operators ncra and ncrcat. Full release statement at http://nco.sf.net/ANNOUNCE -- Charlie Zender, Earth System Sci. & Computer Sci. University of California, Irvine 949-891-2429 )'(
netcdfgroup
archives: