Re: [thredds] Union of time-aggregated datasets

• To: Kyle Wilcox <KWilcox@xxxxxxxxxxxxxx>, "Thomas LOUBRIEU (Thomas.Loubrieu@xxxxxxxxxx)" <Thomas.Loubrieu@xxxxxxxxxx>, "thredds@xxxxxxxxxxxxxxxx" <thredds@xxxxxxxxxxxxxxxx>
• Subject: Re: [thredds] Union of time-aggregated datasets
• From: Jon Blower <j.d.blower@xxxxxxxxxxxxx>
• Date: Tue, 6 Aug 2013 15:00:38 +0000

Thanks Kyle that worked perfectly!

(Thomas- thanks for your suggestion too!)

Jon

-----Original Message-----
From: Kyle Wilcox [mailto:KWilcox@xxxxxxxxxxxxxx] 
Sent: 06 August 2013 15:12
To: Jon Blower; thredds@xxxxxxxxxxxxxxxx
Subject: RE: Union of time-aggregated datasets

Hi Jon -

There may be some more efficient ways to do this.  A scan will open every 
single file before it is cached.  If your individual files have different time 
units, you will need the "timeUnitsChange" attribute on each aggregation scan.  
If not, you can remove it.
 
<dataset name="blah" ID="blah" urlPath="blah.nc">
  ...
  <netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2";>
    <aggregation type="union">
      <netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2";>
        <aggregation dimName="time" type="joinExisting" timeUnitsChange="true">
          <scan location="/foo/bar/" suffix="*temp.nc" />
        </aggregation>
      </netcdf>
      <netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2";>
        <aggregation dimName="time" type="joinExisting" timeUnitsChange="true">
          <scan location="/foo/bar/" suffix="*salt.nc" />
        </aggregation>
      </netcdf>
      <netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2";>
        <aggregation dimName="time" type="joinExisting" timeUnitsChange="true">
          <scan location="/foo/bar/" suffix="*do.nc" />
        </aggregation>
      </netcdf>
      <netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2";>
        <aggregation dimName="time" type="joinExisting" timeUnitsChange="true">
          <scan location="/foo/bar/" suffix="*elevation.nc" />
        </aggregation>
      </netcdf>
    </aggregation>
  </netcdf>
</dataset>

Let me know if it works!

> -----Original Message-----
> From: thredds-bounces@xxxxxxxxxxxxxxxx [mailto:thredds- 
> bounces@xxxxxxxxxxxxxxxx] On Behalf Of Jon Blower
> Sent: Tuesday, August 06, 2013 9:49 AM
> To: thredds@xxxxxxxxxxxxxxxx
> Subject: [thredds] Union of time-aggregated datasets
> 
> Hi,
> 
> I have a dataset consisting of 4 variables, each recorded on around 
> 1000 timesteps.  Unfortunately each variable and each timestep is 
> provided as a separate NetCDF file, so I have about 4000 files.  
> Ideally I'd like to create a single dataset aggregating all of these 
> variables and timesteps. How can I structure my THREDDS catalog to do this?
> 
> (It's not a forecast model so there are no overlapping times, so I'd 
> like to avoid the FMRC aggregation type if possible.)
> 
> Cheers,
> Jon
> 
> --
> Dr Jon Blower
> Technical Director, Reading e-Science Centre School of Mathematical 
> and Physical Sciences University of Reading, UK
> Tel: +44 (0)118 378 5213
> Mob: +44 (0)7919 112687
> http://www.resc.reading.ac.uk
> 
> 
> _______________________________________________
> thredds mailing list
> thredds@xxxxxxxxxxxxxxxx
> For list information or to unsubscribe,  visit:
> http://www.unidata.ucar.edu/mailing_lists/

• References:
  • [thredds] Union of time-aggregated datasets
    • From: Jon Blower
  • Re: [thredds] Union of time-aggregated datasets
    • From: Kyle Wilcox