Last 30 days

• Dec 21, 2009

  issue 33 (Run-time errors after (maybe-not-so) sucessful netcdf4-pytho...) commented on by whitaker.jeffrey   -   I'm not able to reproduce this error with current SVN trunk, hdf5 1.8.4 and netcdf-4.1rc1. I would suggest upgrading your netcdf lib to 4.1rc1. Also, please report any test failures you see when running test/run_all.py. -Jeff

  I'm not able to reproduce this error with current SVN trunk, hdf5 1.8.4 and netcdf-4.1rc1. I would suggest upgrading your netcdf lib to 4.1rc1. Also, please report any test failures you see when running test/run_all.py. -Jeff
• Dec 21, 2009

  issue 33 (Run-time errors after (maybe-not-so) sucessful netcdf4-pytho...) reported by amy.macf...@noaa.gov   -   What steps will reproduce the problem? Various run-time errors occur, for many different cases -- for example, in trying to write NETCDF3_CLASSIC format (the following code works without the format='NETCDF3_CLASSIC' flag): from netCDF4 import Dataset rootgrp = Dataset('test.nc','w',format='NETCDF3_CLASSIC') rootgrp.createDimension('time',None) rootgrp.createDimension('lat', 73) rootgrp.createDimension('lon', 144) latitudes = rootgrp.createVariable('lat','f4',('lat',)) longitudes = rootgrp.createVariable('lon','f4',('lon',)) somevar = rootgrp.createVariable('somevar','f4',('time','lat','lon')) import numpy lats = numpy.arange(-90,91,2.5) lons = numpy.arange(-180,180,2.5) latitudes[:] = lats[:] longitudes[:] = lons[:] f = numpy.zeros((1,len(lats),len(lons)), dtype='f4') somevar[:] = f[:] rootgrp.close() What is the expected output? What do you see instead? Expected output, a netcdf file with variables lat,lon,somevar Instead: Runtime error Traceback (most recent call last): File "test.py", line 18, in <module> somevar[:] = f[:] File "netCDF4.pyx", line 2453, in netCDF4.Variable.__setitem__ (netCDF4.c:16793) File "netCDF4.pyx", line 2577, in netCDF4.Variable._put (netCDF4.c:17771) RuntimeError: Protocol error What version of the product are you using? On what operating system? Ubuntu Karmic netcdf4 0.8.2 Please provide any additional information below. I built the HDF5 (1.8.4) and netCDF (4.0.1) following the instructions on this site (ie with --enable-shared, etc) Neither seemed to have any problems. I don't get any errors on the import statements in Python and I have some functionality -- ie the error in the above code comes about when trying to assign values to the 3-d variable, everything else works okay. However, I get all sorts of run-time errors when I do various things, so clearly I have done something wrong :(

  What steps will reproduce the problem? Various run-time errors occur, for many different cases -- for example, in trying to write NETCDF3_CLASSIC format (the following code works without the format='NETCDF3_CLASSIC' flag): from netCDF4 import Dataset rootgrp = Dataset('test.nc','w',format='NETCDF3_CLASSIC') rootgrp.createDimension('time',None) rootgrp.createDimension('lat', 73) rootgrp.createDimension('lon', 144) latitudes = rootgrp.createVariable('lat','f4',('lat',)) longitudes = rootgrp.createVariable('lon','f4',('lon',)) somevar = rootgrp.createVariable('somevar','f4',('time','lat','lon')) import numpy lats = numpy.arange(-90,91,2.5) lons = numpy.arange(-180,180,2.5) latitudes[:] = lats[:] longitudes[:] = lons[:] f = numpy.zeros((1,len(lats),len(lons)), dtype='f4') somevar[:] = f[:] rootgrp.close() What is the expected output? What do you see instead? Expected output, a netcdf file with variables lat,lon,somevar Instead: Runtime error Traceback (most recent call last): File "test.py", line 18, in <module> somevar[:] = f[:] File "netCDF4.pyx", line 2453, in netCDF4.Variable.__setitem__ (netCDF4.c:16793) File "netCDF4.pyx", line 2577, in netCDF4.Variable._put (netCDF4.c:17771) RuntimeError: Protocol error What version of the product are you using? On what operating system? Ubuntu Karmic netcdf4 0.8.2 Please provide any additional information below. I built the HDF5 (1.8.4) and netCDF (4.0.1) following the instructions on this site (ie with --enable-shared, etc) Neither seemed to have any problems. I don't get any errors on the import statements in Python and I have some functionality -- ie the error in the above code comes about when trying to assign values to the 3-d variable, everything else works okay. However, I get all sorts of run-time errors when I do various things, so clearly I have done something wrong :(
• Dec 19, 2009

  r811 (update requirements ) committed by whitaker.jeffrey   -   update requirements

  update requirements
• Dec 18, 2009

  issue 32 (AttributeError: 'module' object has no attribute '_StartCoun...) commented on by sourish.basu   -   The problem's gone! For some reason, "python setup.py install" wasn't overwriting the old netcdf4_utils.pyc. So deleting that file manually, regenerating netCDF4.c using cython, and re-doing the "install" step solved the problem :) -Sourish

  The problem's gone! For some reason, "python setup.py install" wasn't overwriting the old netcdf4_utils.pyc. So deleting that file manually, regenerating netCDF4.c using cython, and re-doing the "install" step solved the problem :) -Sourish
• Dec 18, 2009

  issue 32 (AttributeError: 'module' object has no attribute '_StartCoun...) commented on by whitaker.jeffrey   -   Sourish: A normal user account should do it. I'll also need to know where you installed the hdf5 and netcdf4 libs. Thanks. -Jeff

  Sourish: A normal user account should do it. I'll also need to know where you installed the hdf5 and netcdf4 libs. Thanks. -Jeff
• Dec 18, 2009

  issue 32 (AttributeError: 'module' object has no attribute '_StartCoun...) commented on by sourish.basu   -   Jeff, There is a module netCDF4_utils, but that doesn't have the function _StartCountStride. [1]: import netCDF4_utils [2]: dir(netCDF4_utils) [2]: ['__builtins__', '__doc__', '__file__', '__name__', '_buildStartCountStride', '_find_dim', '_quantize', 'ma', 'numpy', 'types'] I regenerated netCDF4.c using cython, but that didn't help either. Regarding access to a machine running snow leopard, do you just need a normal user account? If you want, I can create an account for you on my machine. Since I need the netCDF4 module for my research, and you need a snow leopard machine to debug the module, I think that would be a fair exchange :) -Sourish

  Jeff, There is a module netCDF4_utils, but that doesn't have the function _StartCountStride. [1]: import netCDF4_utils [2]: dir(netCDF4_utils) [2]: ['__builtins__', '__doc__', '__file__', '__name__', '_buildStartCountStride', '_find_dim', '_quantize', 'ma', 'numpy', 'types'] I regenerated netCDF4.c using cython, but that didn't help either. Regarding access to a machine running snow leopard, do you just need a normal user account? If you want, I can create an account for you on my machine. Since I need the netCDF4 module for my research, and you need a snow leopard machine to debug the module, I think that would be a fair exchange :) -Sourish
• Dec 17, 2009

  issue 32 (AttributeError: 'module' object has no attribute '_StartCoun...) commented on by whitaker.jeffrey   -   Sourish: Unfortunately, I don't have a snow leopard machine to test on. I'm a bit concerned about "/usr/include/AvailabilityMacros.h:108:14: warning: #warning Building for Intel with Mac OS X Deployment Target < 10.4 is invalid." - but I don't know what to suggest to get rid of that. The function _StartCountStride is in the module netCDF4_utils, so to see if that module was installed correctly, you could try >>> import netCDF4_utils >>> netCDF4_utils._StartCountStride <function _StartCountStride at 0x1146b30> If you have cython installed, you might try re-generating the C source code (netCDF4.c) by running "cython netCDF4.pyx". Other than that, I really don't have any ideas. I'll see if I can find someone who has snow leopard installed so I can try it myself. -Jeff

  Sourish: Unfortunately, I don't have a snow leopard machine to test on. I'm a bit concerned about "/usr/include/AvailabilityMacros.h:108:14: warning: #warning Building for Intel with Mac OS X Deployment Target < 10.4 is invalid." - but I don't know what to suggest to get rid of that. The function _StartCountStride is in the module netCDF4_utils, so to see if that module was installed correctly, you could try >>> import netCDF4_utils >>> netCDF4_utils._StartCountStride <function _StartCountStride at 0x1146b30> If you have cython installed, you might try re-generating the C source code (netCDF4.c) by running "cython netCDF4.pyx". Other than that, I really don't have any ideas. I'll see if I can find someone who has snow leopard installed so I can try it myself. -Jeff
• Dec 17, 2009

  issue 32 (AttributeError: 'module' object has no attribute '_StartCoun...) commented on by sourish.basu   -   Hi, Has this issue been resolved yet? Recently I installed netcdf-4.1-rc1, and I still get this error. -Sourish

  Hi, Has this issue been resolved yet? Recently I installed netcdf-4.1-rc1, and I still get this error. -Sourish
• Dec 12, 2009

  r810 (coerce to string before using PyString_AsString ) committed by whitaker.jeffrey   -   coerce to string before using PyString_AsString

  coerce to string before using PyString_AsString
• Dec 10, 2009

  issue 32 (AttributeError: 'module' object has no attribute '_StartCoun...) commented on by sourish.basu   -   P.S. - I can see that the python module is linked to the correct libraries because: ======================== $ otool -L /Library/Frameworks/Python.framework/Versions/5.0.0/lib/python2.5/site-packages/netCDF4.so /Library/Frameworks/Python.framework/Versions/5.0.0/lib/python2.5/site-packages/netCDF4.so: /usr/local/lib/libnetcdf.5.dylib (compatibility version 6.0.0, current version 6.0.0) /usr/local/lib/libhdf5.6.dylib (compatibility version 7.0.0, current version 7.2.0) /usr/local/lib/libhdf5_hl.6.dylib (compatibility version 7.0.0, current version 7.2.0) /opt/local/lib/libz.1.dylib (compatibility version 1.0.0, current version 1.2.3) /opt/local/lib/libsz.2.dylib (compatibility version 3.0.0, current version 3.0.0) /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 125.0.0) ========================

  P.S. - I can see that the python module is linked to the correct libraries because: ======================== $ otool -L /Library/Frameworks/Python.framework/Versions/5.0.0/lib/python2.5/site-packages/netCDF4.so /Library/Frameworks/Python.framework/Versions/5.0.0/lib/python2.5/site-packages/netCDF4.so: /usr/local/lib/libnetcdf.5.dylib (compatibility version 6.0.0, current version 6.0.0) /usr/local/lib/libhdf5.6.dylib (compatibility version 7.0.0, current version 7.2.0) /usr/local/lib/libhdf5_hl.6.dylib (compatibility version 7.0.0, current version 7.2.0) /opt/local/lib/libz.1.dylib (compatibility version 1.0.0, current version 1.2.3) /opt/local/lib/libsz.2.dylib (compatibility version 3.0.0, current version 3.0.0) /usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 125.0.0) ========================
• Dec 10, 2009

  issue 32 (AttributeError: 'module' object has no attribute '_StartCoun...) commented on by sourish.basu   -   Sure, here's the output: ======================== $ python setup.py build /Library/Frameworks/Python.framework/Versions/5.0.0/lib/python2.5/distutils/dist.py:263: UserWarning: Unknown distribution option: 'summary' warnings.warn(msg) running build running config_cc unifing config_cc, config, build_clib, build_ext, build commands --compiler options running config_fc unifing config_fc, config, build_clib, build_ext, build commands --fcompiler options running build_src building py_modules sources building extension "netCDF4" sources running build_py creating build creating build/lib.macosx-10.3-i386-2.5 copying netCDF4_utils.py -> build/lib.macosx-10.3-i386-2.5 creating build/lib.macosx-10.3-i386-2.5/netcdftime copying netcdftime/__init__.py -> build/lib.macosx-10.3-i386-2.5/netcdftime copying netcdftime/netcdftime.py -> build/lib.macosx-10.3-i386-2.5/netcdftime running build_ext customize UnixCCompiler customize UnixCCompiler using build_ext building 'netCDF4' extension compiling C sources C compiler: gcc -fno-strict-aliasing -fno-common -dynamic -no-cpp-precomp -mno-fused-madd -DNDEBUG -arch i386 -g -O3 -I/usr/local/include -I/opt/local/include creating build/temp.macosx-10.3-i386-2.5 compile options: '-I/usr/local/include -I/usr/local/include -I/opt/local/include -I/Library/Frameworks/Python.framework/Versions/5.0.0/lib/python2.5/site-packages/numpy/core/include -I/Library/Frameworks/Python.framework/Versions/5.0.0/include/python2.5 -c' gcc: netCDF4.c In file included from /usr/include/architecture/i386/math.h:626, from /usr/include/math.h:28, from /Library/Frameworks/Python.framework/Versions/5.0.0/include/python2.5/pyport.h:231, from /Library/Frameworks/Python.framework/Versions/5.0.0/include/python2.5/Python.h:57, from netCDF4.c:4: /usr/include/AvailabilityMacros.h:108:14: warning: #warning Building for Intel with Mac OS X Deployment Target < 10.4 is invalid. netCDF4.c: In function ‘__pyx_pf_7netCDF4_getlibversion’: netCDF4.c:23811: warning: assignment discards qualifiers from pointer target type gcc -g -arch i386 -L/usr/local/lib -L/opt/local/lib -bundle -undefined dynamic_lookup build/temp.macosx-10.3-i386-2.5/netCDF4.o -L/usr/local/lib -L/usr/local/lib -L/opt/local/lib -L/usr/local/lib -L/usr/local/lib -L/opt/local/lib -lnetcdf -lhdf5 -lhdf5_hl -lz -lsz -o build/lib.macosx-10.3-i386-2.5/netCDF4.so running scons running build_scripts creating build/scripts.macosx-10.3-i386-2.5 copying and adjusting utils/nc3tonc4 -> build/scripts.macosx-10.3-i386-2.5 copying and adjusting utils/nc4tonc3 -> build/scripts.macosx-10.3-i386-2.5 copying and adjusting utils/grib2nc4 -> build/scripts.macosx-10.3-i386-2.5 changing mode of build/scripts.macosx-10.3-i386-2.5/nc3tonc4 from 644 to 755 changing mode of build/scripts.macosx-10.3-i386-2.5/nc4tonc3 from 644 to 755 changing mode of build/scripts.macosx-10.3-i386-2.5/grib2nc4 from 644 to 755 ======================== -Sourish

  Sure, here's the output: ======================== $ python setup.py build /Library/Frameworks/Python.framework/Versions/5.0.0/lib/python2.5/distutils/dist.py:263: UserWarning: Unknown distribution option: 'summary' warnings.warn(msg) running build running config_cc unifing config_cc, config, build_clib, build_ext, build commands --compiler options running config_fc unifing config_fc, config, build_clib, build_ext, build commands --fcompiler options running build_src building py_modules sources building extension "netCDF4" sources running build_py creating build creating build/lib.macosx-10.3-i386-2.5 copying netCDF4_utils.py -> build/lib.macosx-10.3-i386-2.5 creating build/lib.macosx-10.3-i386-2.5/netcdftime copying netcdftime/__init__.py -> build/lib.macosx-10.3-i386-2.5/netcdftime copying netcdftime/netcdftime.py -> build/lib.macosx-10.3-i386-2.5/netcdftime running build_ext customize UnixCCompiler customize UnixCCompiler using build_ext building 'netCDF4' extension compiling C sources C compiler: gcc -fno-strict-aliasing -fno-common -dynamic -no-cpp-precomp -mno-fused-madd -DNDEBUG -arch i386 -g -O3 -I/usr/local/include -I/opt/local/include creating build/temp.macosx-10.3-i386-2.5 compile options: '-I/usr/local/include -I/usr/local/include -I/opt/local/include -I/Library/Frameworks/Python.framework/Versions/5.0.0/lib/python2.5/site-packages/numpy/core/include -I/Library/Frameworks/Python.framework/Versions/5.0.0/include/python2.5 -c' gcc: netCDF4.c In file included from /usr/include/architecture/i386/math.h:626, from /usr/include/math.h:28, from /Library/Frameworks/Python.framework/Versions/5.0.0/include/python2.5/pyport.h:231, from /Library/Frameworks/Python.framework/Versions/5.0.0/include/python2.5/Python.h:57, from netCDF4.c:4: /usr/include/AvailabilityMacros.h:108:14: warning: #warning Building for Intel with Mac OS X Deployment Target < 10.4 is invalid. netCDF4.c: In function ‘__pyx_pf_7netCDF4_getlibversion’: netCDF4.c:23811: warning: assignment discards qualifiers from pointer target type gcc -g -arch i386 -L/usr/local/lib -L/opt/local/lib -bundle -undefined dynamic_lookup build/temp.macosx-10.3-i386-2.5/netCDF4.o -L/usr/local/lib -L/usr/local/lib -L/opt/local/lib -L/usr/local/lib -L/usr/local/lib -L/opt/local/lib -lnetcdf -lhdf5 -lhdf5_hl -lz -lsz -o build/lib.macosx-10.3-i386-2.5/netCDF4.so running scons running build_scripts creating build/scripts.macosx-10.3-i386-2.5 copying and adjusting utils/nc3tonc4 -> build/scripts.macosx-10.3-i386-2.5 copying and adjusting utils/nc4tonc3 -> build/scripts.macosx-10.3-i386-2.5 copying and adjusting utils/grib2nc4 -> build/scripts.macosx-10.3-i386-2.5 changing mode of build/scripts.macosx-10.3-i386-2.5/nc3tonc4 from 644 to 755 changing mode of build/scripts.macosx-10.3-i386-2.5/nc4tonc3 from 644 to 755 changing mode of build/scripts.macosx-10.3-i386-2.5/grib2nc4 from 644 to 755 ======================== -Sourish
• Dec 10, 2009

  issue 32 (AttributeError: 'module' object has no attribute '_StartCoun...) commented on by whitaker.jeffrey   -   Can you please capture all the output of "python setup.py build" and post it here? This is unrelated to your problem, but you should probably be using a more recent netcdf (netcdf-4.1 rc is available).

  Can you please capture all the output of "python setup.py build" and post it here? This is unrelated to your problem, but you should probably be using a more recent netcdf (netcdf-4.1 rc is available).
• Dec 10, 2009

  issue 32 (AttributeError: 'module' object has no attribute '_StartCoun...) reported by sourish.basu   -   What steps will reproduce the problem? 1. Install netcdf-4.0 from source with the following configure options ./configure --prefix=/usr/local --enable-shared --disable-f77 --enable-static --with-hdf5=/usr/local --with-zlib=/opt/local --with-szlib=/opt/local --enable-netcdf-4 CC=gcc CFLAGS='-arch i386' LDFLAGS='-arch i386' FCFLAGS='-arch i386' CXXFLAGS='-arch i386' HDF5 was installed from source, zlib and szlib were installed from Macports. 2. Install netCDF4-0.8.2 (python module for netcdf4) after setting proper SZIP_DIR, HDF5_DIR and NETCDF4_DIR, using the command python setup.py build sudo python setup.py install 3. Try to import netCDF4 from python (ipython shell): [1]: import netCDF4 What is the expected output? What do you see instead? I expect the module to be imported without any error(s). Instead, I see the following: [1]: import netCDF4 --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) /Users/basu/ResearchCode/Test IO/<ipython console> in <module>() /Users/basu/ResearchCode/Test IO/netCDF4.pyx in netCDF4 (netCDF4.c:30879)() AttributeError: 'module' object has no attribute '_StartCountStride' What version of the product are you using? On what operating system? netCDF4-0.8.2 on Mac OSX 10.6 (snow leopard) Please provide any additional information below. $ gcc -v Using built-in specs. Target: i686-apple-darwin10 Configured with: /var/tmp/gcc/gcc-5646.1~2/src/configure --disable-checking --enable-werror --prefix=/usr --mandir=/share/man --enable-languages=c,objc,c++,obj-c++ --program-transform-name=/^[cg][^.-]*$/s/$/-4.2/ --with-slibdir=/usr/lib --build=i686-apple-darwin10 --with-gxx-include-dir=/include/c++/4.2.1 --program-prefix=i686-apple-darwin10- --host=x86_64-apple-darwin10 --target=i686-apple-darwin10 Thread model: posix gcc version 4.2.1 (Apple Inc. build 5646) (dot 1)

  What steps will reproduce the problem? 1. Install netcdf-4.0 from source with the following configure options ./configure --prefix=/usr/local --enable-shared --disable-f77 --enable-static --with-hdf5=/usr/local --with-zlib=/opt/local --with-szlib=/opt/local --enable-netcdf-4 CC=gcc CFLAGS='-arch i386' LDFLAGS='-arch i386' FCFLAGS='-arch i386' CXXFLAGS='-arch i386' HDF5 was installed from source, zlib and szlib were installed from Macports. 2. Install netCDF4-0.8.2 (python module for netcdf4) after setting proper SZIP_DIR, HDF5_DIR and NETCDF4_DIR, using the command python setup.py build sudo python setup.py install 3. Try to import netCDF4 from python (ipython shell): [1]: import netCDF4 What is the expected output? What do you see instead? I expect the module to be imported without any error(s). Instead, I see the following: [1]: import netCDF4 --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) /Users/basu/ResearchCode/Test IO/<ipython console> in <module>() /Users/basu/ResearchCode/Test IO/netCDF4.pyx in netCDF4 (netCDF4.c:30879)() AttributeError: 'module' object has no attribute '_StartCountStride' What version of the product are you using? On what operating system? netCDF4-0.8.2 on Mac OSX 10.6 (snow leopard) Please provide any additional information below. $ gcc -v Using built-in specs. Target: i686-apple-darwin10 Configured with: /var/tmp/gcc/gcc-5646.1~2/src/configure --disable-checking --enable-werror --prefix=/usr --mandir=/share/man --enable-languages=c,objc,c++,obj-c++ --program-transform-name=/^[cg][^.-]*$/s/$/-4.2/ --with-slibdir=/usr/lib --build=i686-apple-darwin10 --with-gxx-include-dir=/include/c++/4.2.1 --program-prefix=i686-apple-darwin10- --host=x86_64-apple-darwin10 --target=i686-apple-darwin10 Thread model: posix gcc version 4.2.1 (Apple Inc. build 5646) (dot 1)
• Dec 02, 2009

  issue 30 (date2index is buggy when used with dates outside the support...) commented on by whitaker.jeffrey   -   Sounds good to me - thanks David.

  Sounds good to me - thanks David.
• Dec 02, 2009

  issue 31 (szip library included by default) reported by david.huard   -   sz in the libs list, even if SZIP_DIR is not in the environment.

  sz in the libs list, even if SZIP_DIR is not in the environment.
• Dec 02, 2009

  issue 30 (date2index is buggy when used with dates outside the support...) reported by david.huard   -   Using a date before the first date in the time variable returns index -1. Using a date after the last date in the time variable raises an error. I've fixed this by replacing indices == -1 by 0, and those equal to N by N-1, where N is the length of the time variable. If this is OK with you, I'll commit those changes and the accompanying tests.

  Using a date before the first date in the time variable returns index -1. Using a date after the last date in the time variable raises an error. I've fixed this by replacing indices == -1 by 0, and those equal to N by N-1, where N is the length of the time variable. If this is OK with you, I'll commit those changes and the accompanying tests.
• Dec 01, 2009

  r809 (preparing for 0.9 release ) committed by whitaker.jeffrey   -   preparing for 0.9 release

  preparing for 0.9 release
• Dec 01, 2009

  r808 (regenerate with Cython 0.12 ) committed by whitaker.jeffrey   -   regenerate with Cython 0.12

  regenerate with Cython 0.12
• Dec 01, 2009

  r807 (try to link szip by default. ) committed by whitaker.jeffrey   -   try to link szip by default.

  try to link szip by default.

Earlier this year

• Nov 22, 2009

  issue 29 (netCDF4.c(230) : error C2026: string too big, trailing chara...) commented on by cjgohlke   -   Sure, I will rebuild when 4.1 final is released and send the installers to you. netCDF 4.1 RC1 seems to have some problems with the VS2008 build: I get test failures in nc_open and a crash in ncsync. Maybe I contact the netCDF group. Thanks for your help!

  Sure, I will rebuild when 4.1 final is released and send the installers to you. netCDF 4.1 RC1 seems to have some problems with the VS2008 build: I get test failures in nc_open and a crash in ncsync. Maybe I contact the netCDF group. Thanks for your help!
• Nov 22, 2009

  r806 (update docs online. ) committed by whitaker.jeffrey   -   update docs online.

  update docs online.
• Nov 22, 2009

  issue 29 (netCDF4.c(230) : error C2026: string too big, trailing chara...) commented on by whitaker.jeffrey   -   Christoph: OK, looks as though that's as good as it's going to get on Windows. Thanks - I will post the installers on the google code site. 4.1 final should be released in a week or so, would you mind rebuilding them then? -Jeff

  Christoph: OK, looks as though that's as good as it's going to get on Windows. Thanks - I will post the installers on the google code site. 4.1 final should be released in a week or so, would you mind rebuilding them then? -Jeff
• Nov 22, 2009

  issue 29 (netCDF4.c(230) : error C2026: string too big, trailing chara...) commented on by cjgohlke   -   The Visual Studio solution/project provided with netCDF 4.1 does not have DAP enabled (there is no configure script in the MS toolchain). Manually enabling USE_DAP and trying to adjust the project files failed so far. I linked against HDF5 1.8.4 and netcdf-4.1-rc1, compiled from sources using project default settings. Using the HDF5 binaries/libraries provided by the HDF group makes no difference.

  The Visual Studio solution/project provided with netCDF 4.1 does not have DAP enabled (there is no configure script in the MS toolchain). Manually enabling USE_DAP and trying to adjust the project files failed so far. I linked against HDF5 1.8.4 and netcdf-4.1-rc1, compiled from sources using project default settings. Using the HDF5 binaries/libraries provided by the HDF group makes no difference.
• Nov 22, 2009

  issue 29 (netCDF4.c(230) : error C2026: string too big, trailing chara...) commented on by whitaker.jeffrey   -   Christoph - seems like there are three errors here. I don't know what could be causing the file lock error, but I suspect it's a windows thing. The DAP error indicates you didn't compile with --enable-dap - is that correct? The HDF5 error in tst_slicing is the most troubling. Did you link against HDF5 1.8.4 and netcdf-4.1-rc1?

  Christoph - seems like there are three errors here. I don't know what could be causing the file lock error, but I suspect it's a windows thing. The DAP error indicates you didn't compile with --enable-dap - is that correct? The HDF5 error in tst_slicing is the most troubling. Did you link against HDF5 1.8.4 and netcdf-4.1-rc1?
• Nov 21, 2009

  issue 29 (netCDF4.c(230) : error C2026: string too big, trailing chara...) commented on by cjgohlke   -   The slow write performance in bench_compress.py was due to an unoptimized zlib1.dll.

  The slow write performance in bench_compress.py was due to an unoptimized zlib1.dll.
• Nov 21, 2009

  issue 29 (netCDF4.c(230) : error C2026: string too big, trailing chara...) commented on by cjgohlke   -   Thanks. You were right. It works now, after reloading setup.py from svn. Test/run_all.py returns the following 5 errors in both 32- and 64-bit versions. In bench_compress.py writing seems slow (I don't have anything to compare to): D:\Dev\Compile\NetCDF\netcdf4-svn\test>python26-x64 run_all.py ...EE...........E....EE................ ====================================================================== ERROR: testing zlib and shuffle compression filters ---------------------------------------------------------------------- Traceback (most recent call last): File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_compression.py", line 41, in tearDown os.remove(file) WindowsError: [Error 32] The process cannot access the file because it is being used by another process: 'c:\\users\\gohlke\\appdata\\local\\temp\\tmp5r0ztl.nc' ====================================================================== ERROR: testing access of data over http using opendap ---------------------------------------------------------------------- Traceback (most recent call last): File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_dap.py", line 28, in runTest file = netCDF4.Dataset(URL) File "netCDF4.pyx", line 1289, in netCDF4.Dataset.__init__ (netCDF4.c:6640) RuntimeError: Invalid argument ====================================================================== ERROR: test_select_nc (tst_netcdftime.TestDate2index) ---------------------------------------------------------------------- Traceback (most recent call last): File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_netcdftime.py", line 214, in tearDown os.remove(self.file) WindowsError: [Error 32] The process cannot access the file because it is being used by another process: 'c:\\users\\gohlke\\appdata\\local\\temp\\tmprz0vdd.nc' ====================================================================== ERROR: test_1d (tst_slicing.VariablesTestCase) ---------------------------------------------------------------------- Traceback (most recent call last): File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_slicing.py", line 39, in tear Down os.remove(self.file) WindowsError: [Error 32] The process cannot access the file because it is being used by another process: 'c:\\users\\gohlke\\appdata\\local\\temp\\tmprftlmv.nc' ====================================================================== ERROR: testing variable slicing ---------------------------------------------------------------------- Traceback (most recent call last): File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_slicing.py", line 17, in setU p f = Dataset(file_name,'w') File "netCDF4.pyx", line 1289, in netCDF4.Dataset.__init__ (netCDF4.c:6640) RuntimeError: Can't add HDF5 file metadata ---------------------------------------------------------------------- Ran 39 tests in 0.839s FAILED (errors=5) D:\Dev\Compile\NetCDF\netcdf4-svn\examples>python bench_compress.py reading and writing a 30 by 15 by 73 by 144 random array .. (average of 10 trials) testing compression ... zlib=False,shuffle=False writing took 0.329739393199 seconds reading took 0.0708450962895 seconds size of test.nc = 37849750 testing compression ... zlib=True,shuffle=False writing took 9.04099844541 seconds reading took 0.632003957132 seconds size of test.nc = 12816711 testing compression ... zlib=True,shuffle=True writing took 3.21693664287 seconds reading took 0.6279646234 seconds size of test.nc = 9085764

  Thanks. You were right. It works now, after reloading setup.py from svn. Test/run_all.py returns the following 5 errors in both 32- and 64-bit versions. In bench_compress.py writing seems slow (I don't have anything to compare to): D:\Dev\Compile\NetCDF\netcdf4-svn\test>python26-x64 run_all.py ...EE...........E....EE................ ====================================================================== ERROR: testing zlib and shuffle compression filters ---------------------------------------------------------------------- Traceback (most recent call last): File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_compression.py", line 41, in tearDown os.remove(file) WindowsError: [Error 32] The process cannot access the file because it is being used by another process: 'c:\\users\\gohlke\\appdata\\local\\temp\\tmp5r0ztl.nc' ====================================================================== ERROR: testing access of data over http using opendap ---------------------------------------------------------------------- Traceback (most recent call last): File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_dap.py", line 28, in runTest file = netCDF4.Dataset(URL) File "netCDF4.pyx", line 1289, in netCDF4.Dataset.__init__ (netCDF4.c:6640) RuntimeError: Invalid argument ====================================================================== ERROR: test_select_nc (tst_netcdftime.TestDate2index) ---------------------------------------------------------------------- Traceback (most recent call last): File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_netcdftime.py", line 214, in tearDown os.remove(self.file) WindowsError: [Error 32] The process cannot access the file because it is being used by another process: 'c:\\users\\gohlke\\appdata\\local\\temp\\tmprz0vdd.nc' ====================================================================== ERROR: test_1d (tst_slicing.VariablesTestCase) ---------------------------------------------------------------------- Traceback (most recent call last): File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_slicing.py", line 39, in tear Down os.remove(self.file) WindowsError: [Error 32] The process cannot access the file because it is being used by another process: 'c:\\users\\gohlke\\appdata\\local\\temp\\tmprftlmv.nc' ====================================================================== ERROR: testing variable slicing ---------------------------------------------------------------------- Traceback (most recent call last): File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_slicing.py", line 17, in setU p f = Dataset(file_name,'w') File "netCDF4.pyx", line 1289, in netCDF4.Dataset.__init__ (netCDF4.c:6640) RuntimeError: Can't add HDF5 file metadata ---------------------------------------------------------------------- Ran 39 tests in 0.839s FAILED (errors=5) D:\Dev\Compile\NetCDF\netcdf4-svn\examples>python bench_compress.py reading and writing a 30 by 15 by 73 by 144 random array .. (average of 10 trials) testing compression ... zlib=False,shuffle=False writing took 0.329739393199 seconds reading took 0.0708450962895 seconds size of test.nc = 37849750 testing compression ... zlib=True,shuffle=False writing took 9.04099844541 seconds reading took 0.632003957132 seconds size of test.nc = 12816711 testing compression ... zlib=True,shuffle=True writing took 3.21693664287 seconds reading took 0.6279646234 seconds size of test.nc = 9085764
• Nov 21, 2009

  issue 29 (netCDF4.c(230) : error C2026: string too big, trailing chara...) commented on by whitaker.jeffrey   -   The compilation of ordereddict.c must have failed. Could you try compiling again and see what the error was?

  The compilation of ordereddict.c must have failed. Could you try compiling again and see what the error was?
• Nov 21, 2009

  issue 29 (netCDF4.c(230) : error C2026: string too big, trailing chara...) commented on by cjgohlke   -   I am using HDF5 1.8.4 and netCDF 4.1-rc1, compiled from sources using Visual Studio 2008. Just noticed that netCDF4-python 0.9svn fails the run_all.py tests with the following error. Version 0.8.2 passes most tests. Traceback (most recent call last): File "run_all.py", line 17, in <module> m = __import__(os.path.splitext(f)[0]) File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_atts.py", line 8, in <module> import netCDF4 File "netCDF4.pyx", line 793, in netCDF4 (netCDF4.c:32198) ImportError: No module named _ordereddict Christoph

  I am using HDF5 1.8.4 and netCDF 4.1-rc1, compiled from sources using Visual Studio 2008. Just noticed that netCDF4-python 0.9svn fails the run_all.py tests with the following error. Version 0.8.2 passes most tests. Traceback (most recent call last): File "run_all.py", line 17, in <module> m = __import__(os.path.splitext(f)[0]) File "D:\Dev\Compile\NetCDF\netcdf4-svn\test\tst_atts.py", line 8, in <module> import netCDF4 File "netCDF4.pyx", line 793, in netCDF4 (netCDF4.c:32198) ImportError: No module named _ordereddict Christoph
• Nov 21, 2009

  issue 29 (netCDF4.c(230) : error C2026: string too big, trailing chara...) commented on by whitaker.jeffrey   -   Thanks Christoph! What versions of HDF5 and netcdf-4 did you link to?

  Thanks Christoph! What versions of HDF5 and netcdf-4 did you link to?
• Nov 21, 2009

  issue 29 (netCDF4.c(230) : error C2026: string too big, trailing chara...) reported by cjgohlke   -   When building netCDF4-python, version 0.9 from svn, with Microsoft Visual Studio 2008, the compiler throws error C2026 at line 230 of netCDF4.c. After breaking up the string into smaller ones as follows, the build succeeds: static char __pyx_mdoc[] = "\n" "Introduction\n" "============\n" "\n" ... "OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR\n" "PERFORMANCE OF THIS SOFTWARE."; 32- and 64-bit binaries for Python 2.6 for Windows are available at <http://www.lfd.uci.edu/~gohlke/pythonlibs/#netcdf4> Christoph

  When building netCDF4-python, version 0.9 from svn, with Microsoft Visual Studio 2008, the compiler throws error C2026 at line 230 of netCDF4.c. After breaking up the string into smaller ones as follows, the build succeeds: static char __pyx_mdoc[] = "\n" "Introduction\n" "============\n" "\n" ... "OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR\n" "PERFORMANCE OF THIS SOFTWARE."; 32- and 64-bit binaries for Python 2.6 for Windows are available at <http://www.lfd.uci.edu/~gohlke/pythonlibs/#netcdf4> Christoph
• Nov 21, 2009

  r805 (preparing for 0.9 release ) committed by whitaker.jeffrey   -   preparing for 0.9 release

  preparing for 0.9 release
• Nov 19, 2009

  r804 (update to use C version of ordered dict if built-in not avai...) committed by whitaker.jeffrey   -   update to use C version of ordered dict if built-in not available.

  update to use C version of ordered dict if built-in not available.
• Nov 19, 2009

  r803 (try to use ordered dict included in python >= 2.7, otherwise...) committed by whitaker.jeffrey   -   try to use ordered dict included in python >= 2.7, otherwise fall back to C implementation from http://www.xs4all.nl/~anthon/Python/ordereddict.

  try to use ordered dict included in python >= 2.7, otherwise fall back to C implementation from http://www.xs4all.nl/~anthon/Python/ordereddict.
• Nov 17, 2009

  issue 28 (Dict might not be the right datatype for dimensions etc) commented on by freehu...@gmx.net   -   I tested it with my code (only netCDF3) and it works fine. Thanks a lot!

  I tested it with my code (only netCDF3) and it works fine. Thanks a lot!
• Nov 16, 2009

  r802 (remove some uneeded code from netCDF4_utils.py ) committed by whitaker.jeffrey   -   remove some uneeded code from netCDF4_utils.py

  remove some uneeded code from netCDF4_utils.py
• Nov 16, 2009

  issue 28 (Dict might not be the right datatype for dimensions etc) commented on by whitaker.jeffrey   -   BTW: Instead of for name, val in ncfile.variables.iteritems(): ncfile_new.createVariable(name, 'f8', val.dimensions) # ?? the type is a bit difficult to get out ?? you can do for name, val in ncfile.variables.iteritems(): ncfile_new.createVariable(name, val.dtype, val.dimensions) # !! the dtype attribute of a variable is a numpy dtype !! -Jeff

  BTW: Instead of for name, val in ncfile.variables.iteritems(): ncfile_new.createVariable(name, 'f8', val.dimensions) # ?? the type is a bit difficult to get out ?? you can do for name, val in ncfile.variables.iteritems(): ncfile_new.createVariable(name, val.dtype, val.dimensions) # !! the dtype attribute of a variable is a numpy dtype !! -Jeff
• Nov 16, 2009

  issue 28 (Dict might not be the right datatype for dimensions etc) commented on by whitaker.jeffrey   -   I've updated SVN to use the ordered dict from http://www.voidspace.org.uk/python/odict.html#introduction. This should fix your issue, hopefully without introducing any other new ones. Please let me know how it works for you. It should also preserve the creation order of attributes, if you access them from __dict__, but I haven't tested this yet. -Jeff

  I've updated SVN to use the ordered dict from http://www.voidspace.org.uk/python/odict.html#introduction. This should fix your issue, hopefully without introducing any other new ones. Please let me know how it works for you. It should also preserve the creation order of attributes, if you access them from __dict__, but I haven't tested this yet. -Jeff
• Nov 16, 2009

  r801 (used ordered dictionaries for variables, dimensions etc. ) committed by whitaker.jeffrey   -   used ordered dictionaries for variables, dimensions etc.

  used ordered dictionaries for variables, dimensions etc.
• Nov 16, 2009

  issue 28 (Dict might not be the right datatype for dimensions etc) commented on by freehu...@gmx.net   -   In Python 2.7 there will be an ordered dict type: http://docs.python.org/dev/whatsnew/2.7.html so it would probably make sense to wait for that. Furthermore, the same problem is (of course) encountered for attributes. However, I would quite like to have order in my attributes now. Can anyone suggest a quick'n'dirty way?

  In Python 2.7 there will be an ordered dict type: http://docs.python.org/dev/whatsnew/2.7.html so it would probably make sense to wait for that. Furthermore, the same problem is (of course) encountered for attributes. However, I would quite like to have order in my attributes now. Can anyone suggest a quick'n'dirty way?
• Nov 11, 2009

  issue 28 (Dict might not be the right datatype for dimensions etc) reported by freehu...@gmx.net   -   What steps will reproduce the problem? 1. Load a netCDF file. 2. move the data into an object of your own for processing etc 3. make a new netCDF file and fill it with above data See attached files: run dict_bug.py, it produces dict_bug_new.nc from dict_bug.nc. But order in dict_bug.nc and dict_bug_out.nc is not the same. What is the expected output? What do you see instead? The order of dimensions, variables and attributes are changed, due to the use of dict data structure. It would be nice to be able to keep the order or choose it. For example, ncview displays coordinates in the way they appear, thus having first X then Y would be good. What version of the product are you using? On what operating system? netCDF 0.7.3.1 on linux, Python 2.5 Please provide any additional information below. There are a number of ordered dict types floating around but not in the standard library, sadly.

  What steps will reproduce the problem? 1. Load a netCDF file. 2. move the data into an object of your own for processing etc 3. make a new netCDF file and fill it with above data See attached files: run dict_bug.py, it produces dict_bug_new.nc from dict_bug.nc. But order in dict_bug.nc and dict_bug_out.nc is not the same. What is the expected output? What do you see instead? The order of dimensions, variables and attributes are changed, due to the use of dict data structure. It would be nice to be able to keep the order or choose it. For example, ncview displays coordinates in the way they appear, thus having first X then Y would be good. What version of the product are you using? On what operating system? netCDF 0.7.3.1 on linux, Python 2.5 Please provide any additional information below. There are a number of ordered dict types floating around but not in the standard library, sadly.
• Oct 29, 2009

  r800 (change default maskandscale for netCDF3 to True. ) committed by whitaker.jeffrey   -   change default maskandscale for netCDF3 to True.

  change default maskandscale for netCDF3 to True.
• Oct 29, 2009

  r799 (maskandscale is now True by default ) committed by whitaker.jeffrey   -   maskandscale is now True by default

  maskandscale is now True by default
• Oct 29, 2009

  r798 (change default maskandscale to True ) committed by whitaker.jeffrey   -   change default maskandscale to True

  change default maskandscale to True
• Oct 27, 2009

  r797 (remove szip support ) committed by whitaker.jeffrey   -   remove szip support

  remove szip support
• Oct 27, 2009

  r796 (removed szip support ) committed by whitaker.jeffrey   -   removed szip support

  removed szip support
• Oct 26, 2009

  issue 27 ([request] win32-py2.6 binary release) commented on by whitaker.jeffrey   -   I don't have access to a Windows machine to make binary releases for that platform. The binary 2.5 releases on the download page were contributed by a user, and he is not ready to install 2.6 yet. So, the bottom line is that there will be no Python26 windows binary installer until someone steps up to contribute one. -Jeff

  I don't have access to a Windows machine to make binary releases for that platform. The binary 2.5 releases on the download page were contributed by a user, and he is not ready to install 2.6 yet. So, the bottom line is that there will be no Python26 windows binary installer until someone steps up to contribute one. -Jeff
• Oct 24, 2009

  issue 27 ([request] win32-py2.6 binary release) reported by csigaa   -   Please make a python 2.6 build for win32, it would be a great help - I don't have the time to build it myself but I should install netCDF on many workstations with activepython 2.6...

  Please make a python 2.6 build for win32, it would be a great help - I don't have the time to build it myself but I should install netCDF on many workstations with activepython 2.6...
• Oct 22, 2009

  netCDF4-0.8.2.win32-py2.5.exe (netCDF4 version 0.8.2 (32-bit MinGW, netcdf-4.1-beta2, hdf-1...) file uploaded by byeonguk.kim   -  

  Labels: Featured Type-Installer OpSys-Windows

  Labels: Featured Type-Installer OpSys-Windows
• Oct 21, 2009

  r795 (update docs with szip info. ) committed by whitaker.jeffrey   -   update docs with szip info.

  update docs with szip info.
• Oct 21, 2009

  r794 (improve error message when szip support not available ) committed by whitaker.jeffrey   -   improve error message when szip support not available

  improve error message when szip support not available