[esip-semanticweb] some sample rules for ToolMatch rules

Tue Dec 4 11:58:45 EST 2012

On Dec 3, 2012, at 6:04 PM, Eric Rozell <rozele at rpi.edu> wrote:

> Also, we should start discussing what sort of metadata should be captured for tools and datasets.
> 
> I.e., for tools, should there be a link to documentation, source download, etc.?

This is a tricky question, as there seems to be no widespread common structure for tool metadata.  Also, some tools are online, no download (e.g., Giovanni).

One possibility is just to have a landing page URL, sort of the URL's home page.  Thus, we would have, to take two rich, but very different examples:
  Panoply -> http://www.giss.nasa.gov/tools/panoply
  Giovanni -> http://giovanni.gsfc.nasa.gov/
etc.

Another possibility is to point to an entry in an existing catalog of tools, such as GCMD SERFs.

> 
> For datasets, what things should we capture besides the basics required to classify (i.e., what else besides the attributes listed in the rules below)?

For datasets, I would definitely advocate we stick to the catalogs:  capture only a dataset identifier that can be resolved to an existing dataset catalog, like GCMD, ECHO, GCI, and/or data.gov, etc. We can then pull whatever attributes we need from that catalog.  Of course, we would also need to capture which catalog an identifier comes from, e.g.,:

DatasetCatalog: GlobalChangeMasterDirectory
DatasetCatalogIdentifier:  GES_DISC_AIRX3STD_V005

> 
> On Dec 3, 2012, at 8:59 PM, Eric Rozell wrote:
> 
>> Hi Chris,
>> 
>> I'm trying to convert these to DL rules, here are some clarification questions:
>> 
>> Can you elaborate on the following statement?
>>> o  is on a regular lat/lon grid OR contains auxiliary coordinates for a lat/lon grid
>> What is a "regular lat/lon grid" and what are "auxiliary coordinates for lat/lon grid"?
>> 
>> Is there something more specific than CF-1 that could be used?  It seems there are at least 6 versions of CF, 1.0 through 1.5.  Is there a specific subset of these that you refer to?
>> 
>> Cheers,
>> -Eric
>> 
>> On Oct 30, 2012, at 4:29 PM, Lynnes, Christopher S. (GSFC-6102) wrote:
>> 
>>> Here are some sample rules, in pseudo-English, that describe some of the relationships.  Note that there are almost always some exceptions, usually due to bugs in tool or server code, or edge cases of data structures that technically meet conventions, but stretch them to the breaking point.
>>> 
>>> BTW, here is the wiki page for ToolMatch:  http://wiki.esipfed.org/index.php/ToolMatch
>>> 
>>> Feel free to suggest your own rules!
>>> 
>>> If a data product 
>>> o  is netCDF OR is available via OPeNDAP
>>> o  follows CF-1 conventions for coordinates
>>> o  is on a regular lat/lon grid OR contains auxiliary coordinates for a lat/lon grid
>>> Then the following tools can visualize it on a map:
>>> o  Panoply
>>> o  IDV
>>> o  McIDAS-V
>>> 
>>> If a data product 
>>> o  is netCDF OR is available via OPeNDAP
>>> o  follows CF-1 conventions for coordinates
>>> o  is on a regular lat/lon grid
>>> Then the following tools can visualize it on a map:
>>> o  GrADS
>>> o  Ferret
>>> 
>>> If a data product 
>>> o  is netCDF OR is available via OPeNDAP
>>> o  follows CF-1 conventions for coordinates
>>> o  is on a regular lat/lon grid
>>> Then the following tools can visualize it on a map:
>>> o  GrADS
>>> o  Ferret
>>> 
>>> If a data product 
>>> o  is netCDF OR is available via OPeNDAP
>>> o  follows CF-1 conventions for coordinates
>>> o  contains auxiliary coordinates for a lat/lon grid
>>> Then the Ferret can visualize it as a grid.
>>> 
>>> If a data product is offered through:
>>> o  Hyrax
>>> o  THREDDS Data Server
>>> o  GrADS Data Server
>>> o  erddap
>>> Then it is available through OPeNDAP
>>> 
>>> There are exceptions to these rules of course:  data products that contain variables over 15 characters, or more than one vertical dimension, are problematic for GrADS, requiring significant extra work (control/XDF files vs. use of the sdfopen command).  Also, the rules for HDF4 and HDF5 are more complicated, but would look similar to these.
>>> 
>>> Is this enough to start with?
>>> 
>>> BTW, are we still on for tomorrow afternoon? I have another telecon earlier in the day.  Assuming I lose power, my cell phone may or may not last through the second one...
>>> --
>>> Chris Lynnes
>>> _______________________________________________
>>> esip-semanticweb mailing list
>>> esip-semanticweb at lists.esipfed.org
>>> http://www.lists.esipfed.org/mailman/listinfo/esip-semanticweb
>>> 
>> 
>> _______________________________________________
>> esip-semanticweb mailing list
>> esip-semanticweb at lists.esipfed.org
>> http://www.lists.esipfed.org/mailman/listinfo/esip-semanticweb
>> 
> 
> _______________________________________________
> esip-semanticweb mailing list
> esip-semanticweb at lists.esipfed.org
> http://www.lists.esipfed.org/mailman/listinfo/esip-semanticweb

--
Dr. Christopher Lynnes, NASA/GSFC, ph: 301-614-5185