[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[IDV #GOG-123988]: Point datasource changes

Hi Dave,

> Some of our users that work with point data often would like to see
> the point data chooser/field selector (datasource) work a little bit
> differently.  I see that when you select a point data source the field
> selector presents "Gridded Fields" with parameters that can be
> selected.  It would be nice if the "Point Data" would also list the
> parameters available, as well as derived parameters--similar to the
> Catalogs chooser displays gridded model data.  The values of the
> chosen parameter would be plotted without the need to select/edit a
> station layout model (I'm sure a station model would have to be
> created on the fly, but many users just want to see their data plotted
> without delving into the editor).
> 
> I have looked at AddePointDataSource, but I imagine this is something
> that the Unidata community would be interested in as well.  Would you
> be interested in developing this functionality, or would you like us
> to look further into it?
> 

This is something we have thought about in the past but haven't had the time to 
do.
If you would like to take a crack at it I'd suggest you modify 
data.point.PointDataSource
so this functionality is applicable for all point data. As a first pass at it I 
would create a set of data choices, one for each numeric parameter, like we do 
for the gridded fields. The data choice can have a property, e.g., POINT_PARAM, 
that is the String name of the parameter. This is all done in doMakeDataChoices.

Then in getDataInner just recognize that the data choice has this property and 
after we do:
            retField = PointObFactory.mergeData(datas);
write some code (probably in PointObFactory) that does a parameterSubset on the 
field. e.g.:
            retField = PointObFactory.parameterSubset(retField, param);
parameterSubset just needs to make new PointObs from the old PointObs but with 
a single RealTuple

After this first pass perhaps the sub-classes could also recognize that only 
one (or a set) of parameters
is needed and only read that data. But, this should also work at this top-level 
class for those datasources that don't do the parameter subset.

As to the layout model - a simple layout model can be created showing the 
value. The parameter for the value symbol can just be "*". This will match the 
first value in the observation tuple.


-Jeff


Ticket Details
===================
Ticket ID: GOG-123988
Department: Support IDV
Priority: Normal
Status: Closed