Re: 20020724: Problem putting data to variables

>To: address@hidden
>From: James Adams <address@hidden>
>Subject: Re: Problem putting data to variables
>Organization: ?
>Keywords: 200207240011.g6O0B2917202 netCDF

Hi James,

> As it turns out the "Time" coordinate variable info is
> being modified by my writing of normal variable data. 
> If I don't put any variable data to the file then the
> "Time" coordinate variable stays at zero and no values
> are shown for this coordinate.  It turns out that the
> nc_put_vara_float() that I am using to populate the
> variables is what incremented the "Time" count.  The
> nc_put_var_float() call to populate the "Time"
> coordinate variable with values just doesn't work,
> even though it doesn't produce an exception or show
> any other signs of failure.  This approach works for
> populating the other three coordinate variables but
> not with "Time" which has a NC_UNLIMITED dimension. 
> Still no clue on why this is a special case and not
> working like the other calls using coordinate
> variables with normal (limited) dimensions
> 
> Is there any document which explains or gives an
> example of how to correctly define and then populate
> coordinate variables ?  This really shouldn't be so
> difficult and could probably be explained with a few
> lines of valid code.

In the netCDF User's Guide for Fortran 90 (and soon to be added to
the other netCDF User's Guides as well) there is the following
explanation of why you shouldn't use the simplest netCDF interface to
write arrays that use the NC_UNLIMITED (or record) dimension:

  Take care when using the simplest forms of this interface with record
  variables when you don't specify how many records are to be
  written. If you try to write all the values of a record variable into
  a netCDF file that has no record data yet (hence has 0 records),
  nothing will be written. Similarly, if you try to write all of a
  record variable but there are more records in the file than you
  assume, more data may be written to the file than you supply, which
  may result in a segmentation violation.

Here are a couple of previous support responses dealing with this issue

 http://www.unidata.ucar.edu/cgi-bin/mfs/70/4293
 http://www.unidata.ucar.edu/cgi-bin/mfs/70/4679

This apparently should be documented more clearly, and I should also
add it to the FAQ list of frequently asked questions, since several
users have indicated that it's a problem.

> Are there any NetCDF user mailing lists which I could
> post this sort of question ?  It appears that ncdigest
> isn't really appropriate for user questions such as
> this, but I am at a loss to find any other sources of
> information/help (which is why I am pestering this
> address with so many questions).

This (address@hidden) is the right address to use for such
questions.  And we don't mind being pestered, especially when we
forget or neglect to answer a well-posed question such as yours ...

--Russ

_____________________________________________________________________

Russ Rew                                         UCAR Unidata Program
address@hidden                     http://www.unidata.ucar.edu