NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.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.
A problem with size 0 dimensions is when you want to know how much data you have got. When all dimensions are greater than or equal to 1, you can simply multiply all the dimension lengths together in order to determine how much memory etc that will need to be allocated. With size 0 dimensions it will be necessary to have an if statement to check if the dimension length is 0 or not. Existing code would likely break if 0 length dimensions were introduced. What applications could benefit from 0 length dimensions? Tim Hume Bureau of Meteorology Research Centre -----Original Message----- From: owner-netcdfgroup@xxxxxxxxxxxxxxxx [mailto:owner-netcdfgroup@xxxxxxxxxxxxxxxx] On Behalf Of Harvey.Davies@xxxxxxxx Sent: Tuesday, 16 August 2005 2:48 p.m. To: netcdfgroup@xxxxxxxxxxxxxxxx Subject: Dimensions with size 0 Netcdf-3 does not allow dimensions (apart from unlimited dimension) to have a size of 0 (although I recall a decision at a netcdf-3 design meeting in 1996 to allow size 0). It is probably not worthwhile fixing this in netcdf-3, given the immanent change to netcdf-4. Does netcdf-4 allow dimensions with size 0? If not, then I urge the developers to allow them. Things like empty strings and empty sets are useful. Zero is one of man's great inventions! Harvey Davies CSIRO Marine and Atmospheric Research
netcdfgroup
archives: