Hi Dan:
I assume you've seen these docs:
http://motherlode.ucar.edu:8080/thredds/docs/NcML.htm
http://www.unidata.ucar.edu/software/netcdf/ncml/v2.2/Aggregation.html
??
Anyway, you dont need scan, you can just list all the datasets
<netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2";>
(1)<dimension name="time" length="0"/>
<variable name="time" type="int" shape="time">
<attribute name="units" value="months since 2000-6-16 6:00"/>
<attribute name="_CoordinateAxisType" value="Time" />
</variable>
(2)<aggregation dimName="time" type="joinNew">
(3) <variableAgg name="T"/>
(4) <netcdf location="file:test/data/ncml/nc/time0.nc" coordValue="0"/>
<netcdf location="file:test/data/ncml/nc/time1.nc" coordValue="1"/>
<netcdf location="file:test/data/ncml/nc/time2.nc" coordValue="2"/>
</aggregation>
</netcdf>
see
http://www.unidata.ucar.edu/software/netcdf/ncml/v2.2/Aggregation.html#joinNew
lemme know if that helps...
Daniel Holloway wrote:
Hi John,
I'm setting up a TDS to replace our older Aggregation Server. I've
reviewed the documentation you have, at least all the links that I
could find/follow. My question is does the new aggregation
configuration require the use of 'scan', or can I setup the
configuration file similarly to the old way? If so, would you have any
examples I could look at for type 'join-new', that is where I'm
creating a new dimension. The typical new 'time' dimension, hopefully
expressible using the old 'coord=' attribute.
Thanks,
Dan