文件列表:

IPIX\ipixazm.m (712, 2010-02-01)
IPIX\ipixinfo.m (4171, 2010-02-01)
IPIX\ipixload.m (4051, 2010-02-01)
IPIX-radar-data-for-matlab\IPIX数据读取方法\cdfhowto.files\home.gif (986, 2006-02-24)
IPIX-radar-data-for-matlab\IPIX数据读取方法\cdfhowto.htm (11741, 2006-02-24)
IPIX-radar-data-for-matlab\IPIX数据读取方法\datasets.mht (10579, 2010-06-17)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\dir-wcprops (94, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\empty-file (0, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\entries (10791, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\format (2, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\AUTHORS.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\ChangeLog.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\common.c.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\Contents.m.svn-base (4, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\COPYING.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\INSTALL.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\makefile.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\mexcdf.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\mexcdf53.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\mexcdf60.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\mexgateway.c.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\mexnc-doc.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\mexnc.h.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\mexnc_msvc71_st.bat.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\mexnc_msvc71_st_R11.bat.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_64bit_offset_mode.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_byte.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_char.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_clobber_mode.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_double.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_fill_mode.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_float.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_global.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_int.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_lock_mode.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_noclobber_mode.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_nofill_mode.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_nowrite_mode.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_share_mode.m.svn-base (81, 2006-09-20)
IPIX-radar-data-for-matlab\IPIX数据读取方法\mexnc\.svn\prop-base\nc_short.m.svn-base (81, 2006-09-20)
... ...

The Short And Sweet Of It... ============================ Add this directory to your MATLAB path, and you are now setup to use mexnc. Check the READMEs for SNCTOOLS for further instructions. Please note that mexnc is NOT the same as the "NetCDF Toolbox", which is something that I did not write and therefore cannot support. I strongly recommend that you not use the NetCDF Toolbox. Some Efficiency Considerations ============================== Mexnc can currently use two different backends for accessing your netcdf data, the old mexnc mex-file, and the new MATLAB netcdf package. Choosing between them may be determined by as simple a factor as what version of MATLAB you have, or by more complex criterion such as access to password-protected opendap datasets. If you have 2008a or earlier, you must use the mexnc mex-file. If you have 2008b or later, you have the choice of using the original mex-file (which may be opendap-enabled) or using MathWork's netcdf package (the netcdf package is the default). Mexnc determines all this by looking at your MATLAB version and a preference called USE_TMW. If you wish mexnc to >>NOT USE<< MATLAB's native netcdf package (this might make sense if you have netcdf-4 files) you may force the issue with >> setpref('MEXNC','USE_TMW',false); So mexnc is pretty flexible about things, but all this flexibility comes at a cost of performance. If If you know for certain that you want to ONLY one backend or the other, then you may wish to examine the source code for mexnc.m starting at line 546. Therein lies an explanation of a way to speed up the execution of mexnc IF YOU KNOW EXACTLY WHAT YOU ARE DOING. Specifically, you would be hardcoding the backend by which mexnc accesses your netcdf data. Anyway, just go look at line 546 and make your own decision. Just keep in mind that if you do this, you may not be able to easily share your code with someone else whose configuration differs from your configuration. Regarding Official Mathworks Support ======================================================================= Beginning with R2008b, the MathWorks is shipping MATLAB with official support for netCDF. The interface looks somewhat different than mexnc, but mexnc has been augmented to be able to use the native MathWorks support instead of the mexnc mex-file. Just think, all you windows users, no more fooling around with getting the netcdf.dll installed into the proper location! The first question that may arise here is "will my old code continue to work?" The answer is a qualified "yes". Just about everything works as before (so far as I know), but there are a couple of exceptions. 1. Writing attributes with NaN or Inf does not result in the same datatype as it did with the mexnc-2 interface. Since I cannot think of a good reason why one would want to do this in the first place, I haven't much sympathy for anyone who has a problem with this. 2. The memory-mapped I/O routines are not supported by the MathWorks native mex-file. For example, one cannot use mexnc('PUT_VARM_DOUBLE',...) anymore. I do not believe this to be much of an issue because the main thing that these routines was good for was to permute data, and MATLAB is much faster at this than the netCDF library anyway. R2008a and Earlier ======================================================================= Mexnc is only supported on versions R14 and later. R13 will not pass all of the mexnc test suite, but that is due more to the way the test suite is written than the code itself. If you want to use mexnc on R13 or below, you are welcome to do so, but you are on your own. If you have MATLAB versions R2008a or earlier, then you need to realize that you still need the mex-file "mexnc". This is not included in this download. You need to either retrieve it from another download at sourceforge or compile it yourself (the source code IS still included). You can find the source code in the src subdirectory. If you are using windows, then you also need to place the netcdf.dll file somewhere on your SYSTEM path (not your matlab path, but your system path). Something like either [matlabroot '/bin'] or [matlabroot '/bin/win32'] would be good choices. You might have to reboot. SUPPORT ======================================================================= Please note, mexnc is NOT supported by The MathWorks, so please do not ask them questions about it. If you encounter problems with mexnc, you should continue to use the same contact email address that is listed on the sourceforge web site. OPeNDAP ======================================================================= If you wish to use an opendap-enabled version of mexnc, then you have to compile your mex-file yourself. Take a look at the online notes at http://mexcdf.sourceforge.net/opendap/ for further details. You can also look at the various mexopts files in the src directory to try to get a sense out of what you need to grok in order to get this done. Depending upon which release you have, you should try something like the following. Suppose you have R2008a. In order to compile an opendap- enabled mex-file, you would have to invoke make R2008a If you have R2008b and higher and you want to use the native matlab functionality along with opendap, then you should first look at how SNCTOOLS manages this. It can use a java backend for all URLs while using the native netcdf package for local I/O. If you do not wish to use SNCTOOLS, then you need to do a careful two-step dance with a preference. When you are doing your opendap work, set this preference >> setpref('MEXNC','USE_TMW',false); This will cause ALL I/O to go thru the opendap-enabled mex-file (that you have to compile), so remember to turn the preference back on when you are done with it. Problems? ======================================================================== I don't have a lot of time for mexnc anymore, but you can still email me at johne.g.evans.ne@gmail.com

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：