Hi Sean,
Thanks for your reply! The method that you suggested is good for an
Orthogonal Multidimensional Array implementation. But we're dealing with a
large volume of profile dta from the World Ocean Database. The number of
depths across the millions of casts is very highly skewed and so
implementing an orthogonal array approach is highly impractical. That's the
reason we've implemented the Contiguous Ragged Array approach as described
in section H.3.4
-Ajay
On Fri, Mar 25, 2016 at 1:33 PM, Sean Arms <sarms@xxxxxxxx> wrote:
> Greetings Ajay,
>
> Unfortunately, I do not have any advice for aggregations or
> FeatureCollections on TDS 4.2.x - things have changed so much since then,
> especially for serving point data via the TDS. However, the use of netcdf
> files for point data within the TDS heavily relies on the data following
> the CF 1.6 discreet sampling geometry standard. The example files you
> pointed to state they follow the Profile DSG:
>
>
> http://cfconventions.org/Data/cf-conventions/cf-conventions-1.6/build/cf-conventions.html#idp8350112
>
> but they do not quite follow that standard. For example, the dimension
> casts is essentially the profile identifier (the variable wod_unique_cast
> is the profile id). In this case, all data variables should have the
> dimension casts as their first dimension. At that point, you would be able
> to aggregate the files on the casts dimension and things should work. NcML
> cannot be used to "reshape" data variables, so unfortunately this is a job
> that would require a rewrite of the data.
>
> Sorry I don't have better news at the moment - maybe someone else has
> dealt with this and would like to chime in?
>
> Sean
>
>
> On Thu, Mar 24, 2016 at 2:49 PM, Ajay Krishnan - NOAA Affiliate <
> ajay.krishnan@xxxxxxxx> wrote:
>
>> Hi TDS Users,
>>
>> I have monthly data from different CTD casts bundled into a single netCDF
>> file. I am looking at creating a virtual aggregation of such monthly files
>> spanning several decades. What's the best way to implement this using TDS?
>>
>> Can I do this using Feature Collections? I have not been successful in
>> doing a time aggregation using NcML.
>>
>> I have some sample data at:
>> http://data.nodc.noaa.gov/thredds/catalog/testdata/wod_test/catalog.html
>>
>> Running TDS 4.2.10
>>
>> Thanks,
>> Ajay
>>
>> _______________________________________________
>> thredds mailing list
>> thredds@xxxxxxxxxxxxxxxx
>> For list information or to unsubscribe, visit:
>> http://www.unidata.ucar.edu/mailing_lists/
>>
>
>
