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

Re: 20030325: Integrated Data Viewer (IDV) problems (cont.)

Hans-Juergen,

That is good work by you! I am very glad you found how to 
make the wind vectors from this important data set.

This morning before coming in I realized I might have told 
you the wrong way to use the Formula editor, and then you found
the right way to make it work.

Yes, the parameter alias approach, and the automatic
"Derived parameters" will not work in ths case, and for a single reason.

The automatic Derived parameters are offered when the
required data to compute them are available, juding by parameter
names or their aliases.  But all needed parameters must
come from one data source, in this case, from one single
netCDF file.  I did not realize on reading your first email
that you had separate files of u and v wind components.
In this case the parameter names and aliases will not
create a sub category of Derived wind vectors in either file.

I will improve the User Guide about this entire issue
as you suggest.  The IDV is in early stages of use
and experiences such as yours are very valuable in helping
us make it work properly and helping us explain how
to use it.

The CDC web site providing access to subsets of NCEP Reanalysis
data is new to me, and this is clearly very important 
data the IDV can use.  I will add material to the
IDV User Guide, based on your instructions, so that
others can learn about this NCEP Reanalysis data access and subset service.

I am a little puzzled that the formula approach using the
formula expression "makeVector" did not work. Since you have
found a solution this question is purely academic for you, but
I will explain.  The formulas can use either simple math or can 
call methods in the IDV code - what you did - or call a
method in the IDV Jyhton library. The IDV Jython library is
a list of methods to simplify access to IDV code by the
IDV user, and eventually will include, we hope, common
complex computations also.  Look at the IDV picture attached.
A formula is made with expression makeVector(u, v). 
This
works since makeVector(a,b) is a defined method in the
IDV Jython libraries. You can see the Jyhton library supplied
with your IDV with the menu selection Edit->Jython library,
and in the window that pops up, click on the "System Library 1" tab.
On my IDV the method makeVector(a,b) is provided. It should be
in your version also. I suspect I made a typing mistake
such as "makeVectors" when telling you how to write the formula.
What you did was remove the Jython middle man and go directly
to the underlying code by calling DerivedGridFactory.createWindVectors (u, v).

You can write your own Jython methods.
If you write some of widespread interest we will be glad to
include them in future releases of the IDV.

We are always interested in how the IDV is being used. If you like
please send word what project uses the IDV.

Thanks again for letting us know about this important
data and access mechanism.  We are pleased you find the IDV
useful and are sharing it with others.

Stuart Wier

> 
> I believe I found a solution.
> As I told you, I loaded also gridded data from
> 
> http://motherlode.ucar.edu/dods/catalog.xml
> 
> for example one time of the NCEP Eta dataset.
> 
> If you do this, if you select either the category 2D or 3D, the category
> "Derived" appears.
> You cant't open the formulas but just moving the mouse cursor to one button
> (e.g. a windvector button) will show what's in the formula.
> For windvectors the correct syntax is:
> 
> DerivedGridFactory.createWindVectors (u, v)
> 
> (see also:
> http://www.unidata.ucar.edu/community/committees/metapps/docs/ucar/unidata/data/
> grid/DerivedGridFactory.html)
> 
> I changed my vector formula accordingly, and, voila, I could display the
> vectors for my data.
> 
> But it is the only way to do it. With the "parameter aliases" solely I wasn't
> successful.
> 
> The description of "Derived parameters" in the online Users Guide is confusing
> to me.
> As I understood the Guide, the category "Derived" will be created by the IDV
> automatically, each time when appropriate datasets are loaded.
> 
> This category  appears if data from
> http://motherlode.ucar.edu/dods/catalog.xml
> are loaded,
> but it doesn't if I load data from
> http://www.cdc.noaa.gov/THREDDS/catalog.xml
> or if I use the NCEP reanalysis that I described in my first mail this 
> morning.
> 
> But however, I demonstrated the general capabilities of IDV to some 
> colleagues.
> Their reaction was enthusiastic!
> 
> Hans-Juergen
> > > >
> > ****************************************************************************
> > > > **
> > > > > *                                        \
> > > > *
> > > > > *               III  M   M  K   K         \  Dr. Hans-Juergen Panitz
> > > > *
> > > > > *                I   MM MM  K  K           \
> > > > *
> > > > > *                I   M M M  KKK             \  E-mail:
> > > > *
> > > > > *                I   M   M  K  K             \  address@hidden
> > > > *
> > > > > *               III  M   M  K   K             \
> > > > *
> > > > > * postal address:                              \
> > > > *
> > > > > * Forschungszentrum Karlsruhe                   \   Phone:
> > > > *
> > > > > * Institut fuer Meteorologie und Klimaforschung  \  +49-7247-82-2802
> > > > *
> > > > > * Hermann-von-Helmholtz-Platz 1                   \ Fax:
> > > > *
> > > > > * D-76344 Eggenstein-Leopoldshafen                 \  +49-7247-82-4377
> > > > *
> > > > > * Germany                                           \
> > > >

****************************************************************************
Unidata User Support                                    UCAR Unidata Program
303 497 8643                                                  P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata WWW Service                        http://www.unidata.ucar.edu/     
****************************************************************************

GIF image