[Thread Prev][Thread Next][Index]

Re: [ferret_users] mc files don't open anymore in v6.84



Yes, that's it! Thank you so much for this easy solution!

I am surprised that after years of .mc tradition - at least in my working environment - this should be something of the past now. After all, it saved some directory listing space.

So the change is possibly a side effect of the Enhancement to descriptor file definition ../../documentation/release-notes/version-6-8-4-release-notes#descript

Thanks again and best wishes,
Hella



On 29/05/13 21:07, Andrew Wittenberg - NOAA Federal wrote:
Hella,

Could it just be the file name of your descriptor?  Renaming your 1D.mc
to 1D.des makes it work for me.

I see the same results for your file in v6.72 as I do in v6.85, so I'm not
sure anything along these lines has actually changed in Ferret.

Andrew

On Wed, May 29, 2013 at 1:50 PM, Hella Riede <hella.riede@xxxxxxx> wrote:

Dear ferreters,

I cannot open multi-netCDF files (mc files) in ferret anymore. They are
created with nc2mc (http://www.pa.op.dlr.de/~**PatrickJoeckel/nc2mc/index.
**html <http://www.pa.op.dlr.de/~PatrickJoeckel/nc2mc/index.html>) on
Linux. It works fine in v6.72, but in v6.84, the following error message is
issued:

**netCDF error NetCDF: Unknown file format (OPeNDAP/netCDF Error code -51)
Data set: <whatever mc file>

Please find attached a very simple example I made by hand containing only
a time dimension and 6 time points. Both single files are recognized
normally, but the mc file causes the above error in v6.84.

So it must be a change in ferret between v6.72 and v6.84. From the
changelogs, I gathered that between these versions:

1) the number of dimensions was increased from 4 to 6

2) all coordinate axes are now treated as double precision.
../../documentation/release-notes/version-6-8-release-notes


3) Ferret adheres more closely with the CF conventions for axis directions
../../documentation/release-notes/version-6-8-2-release-notes>

4) Ferret v6.8.2 is linked with NetCDF 4.2
../../documentation/release-notes/version-6-8-2-release-notes>

To 3) I first guessed that mc files don't work anymore if the contained
single files include an axis that is decreasing-ordered (which is the case
for our large 3-D files), but my simple example without decreasingly
ordered axes seems to prove that theory wrong.

Many thanks in advance for your help.

Best wishes,
Hella


P.S: Ferret documentation for descriptor files
../../documentation/users-guide/converting-to-netcdf/CREATING-A-MULTI-FILE-NETCDF-DATA-SET>


[Thread Prev][Thread Next][Index]