Re: [thredds] Ncml aggregation, scan, prototype dataset

• To: thredds@xxxxxxxxxxxxxxxx
• Subject: Re: [thredds] Ncml aggregation, scan, prototype dataset
• From: John Caron <caron@xxxxxxxxxxxxxxxx>
• Date: Tue, 08 Jan 2013 15:27:58 -0700

Hi Heiko:


On 1/8/2013 3:57 AM, Heiko Klein wrote:

Hi,

I'm using a simple time-aggregated product in my thredds setup, like:

    <netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2";>
     <aggregation dimName="time" type="joinExisting" recheckEvery="15min">
      <variableAgg name="temperature" />
      <variableAgg name="salinity" />
<scan location="/my/path/" suffix=".nc" />
     </aggregation>
...


Which file will be used as prototype dataset, i.e. which dataset defines
the attributes if they change? In a featurecollection, I am able to
define the prototype like: protoDataset choice="Penultimate" but I
couldn't find that here.

  <Aggregation>
    <typicalDataset>latest</typicalDataset>
  </Aggregation>

see:

in CDM outside of TDS default is to choose a random dataset.

You can set globally with Aggregation.setTypicalDatasetMode().

And, since the variables all have a time-dimension: is the variableAgg
realy required, or does the aggregation automatically aggregate all
variables having a time-dimension?

variableAgg is only used for joinNew

see:

http://www.unidata.ucar.edu/software/netcdf/ncml/v2.2/AnnotatedSchema4.html

Best regards,

Heiko

• References:
  • [thredds] Ncml aggregation, scan, prototype dataset
    • From: Heiko Klein