This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
>From: address@hidden (Leslie Hartten) >Organization: NOAA >Keywords: 199501301713.AA10564 netCDF use Leslie, > I am interested in netCDF because so much of the TOGA COARE data seems >to be in netCDF format. I have downloaded the netcdf source code and >documentation, printed out the documentation, and read through portions >of the Users Guide. I am unsure of how to proceed. > > I do most of my data analysis by writing FORTRAN code to manipulate >the data; I occasionally port ascii files (original or derived) to a >Macintosh for use in a spreadsheet or plotting program. I am trying to >understand how netCDF data files are meant to be used. Is data put in >that format only because it is then easy to ship from machine to machine? Primarily, but the netCDF is actually an API for reading writing data in network transparent fashion. Many people like this API for development. The other thing that the netCDF allows for is storage of meta-data, i.e. data about the data that you are storing. Meta-data allows you to document the data you have in the netCDF format. >Should I leave the files (e.g. the surface observations at the 6 ISS sites >during COARE) in .cdf format and write FORTRAN code which includes >netCDF FORTRAN calls to open the files and load in the data, then process >that data inside the rest of the FORTRAN program? That is typically what people do. I would venture to guess that this is what they do regardless of the format-of/access-to the data set (i.e. you read in data sets and then process them). >Should I use netcdf >utilities to creade .cdl files, then transform these ascii files into my >own format (with my own headers) and proceed with my processing? The method forcreation of a cdl is the choice of the user. I have created CDLs from existing netCDF files (using ncdump) and then modified them for my own use. I have also created CDLs with a text editor and then used netCDF utilities (ncgen) to create template applications that I then expanded upon. The choice is yours. Your decision will be become easier as you become more familiar with the netCDF package and the generic applications, ncdump and ncgen, that come with the package. >Is there a way to "read" just the headers to see what's inside a file? The generic netCDF application ncdump is used for doing this and more. >Thank you for any assistance, >Dr. Leslie M. Hartten >CIRES, Univ. of Colorado email: address@hidden >Campus Box 216 ^^^^^^^^^^^ new address!! >Boulder CO 80309-0216 phone: (303)497-7052 fax: 497-5373 Tom Yoksas