On 8/9/2011 3:57 PM, Roland Schweitzer wrote:
Hi,
I'm starting to create some netCDF files using the new discrete sampling
geometries in the CF standard. My first example is a trajectory. The
header is appended below. However, when I try to read the fie in the
latest toolsUI (4.2.26) under the "Trajectory" tab I get a error message
that say "Failed to find DataType factory for datatype = TRAJECTORY. I'm
looking for clues about my clueless-ness that is leading to this error.
I'd be grateful if others who have created example files of the various
discrete sampling geometries would share them with me. Bob Simons, I'm
looking at you. ;-)
My ears are burning.
1) Which version of the Discrete Sampling Geometries (DSG) document to
you have? Mine is from Steve Hankin. I don't know where else to get
it(!). It says "May 10, 2011" at the top. I believe it is the latest
version and the ratified version.
Please, someone tell us where the current version can be found publicly.
It isn't at
http://cf-pcmdi.llnl.gov/documents/cf-conventions/1.5/cf-conventions-multi.html
It isn't at
https://cf-pcmdi.llnl.gov/trac/wiki/PointObservationConventions
2) Which type of trajectory file are you trying to create?
The documentation for "A9.4.3 Contiguous ragged array representation of
trajectories" says there will be a variable with an attribute called
"sample_dimension". I don't see that in your file.
The documentation for "A9.4.4 Indexed ragged representation of
trajectories" says there will be a variable with an attribute called
"instance_dimension". I don't see that in your file.
I do see trajectory_counts:trajectory_dimension in your file -- perhaps
that is from a different version of the DSG document.
Your file clearly isn't a "multidimensional array" or a "single
trajectory". So I'm guessing we're looking at different versions of DSG.
3) As John Caron points out, the problem is also probably that
netcdf-java isn't up-to-date with the latest DSG version.
4) Your file's time_coverage_start value is after the time_coverage_end
value. Both are probably nonsense values. Perhaps it's better just to
omit them?
5) ERDDAP can make "Contiguous ragged array representations" for all of
the feature types in the DSG.
In general:
* Find relevant datasets by picking a cdm_data_type here
http://coastwatch.pfeg.noaa.gov/erddap/categorize/cdm_data_type/index.html
Note that some datasets are in odd categories (e.g., point) because
of issues with the data.
* Click on the "data" link of the desired dataset.
* For a good test, make a data selection that includes multiple
instances of the feature (e.g., >1 trajectory). This may require a
little experimenting.
* Change "File type" to ".ncCF"
* Click on Submit.
Examples which directly download DSG .nc files are:
Notes:
* If the whole URL isn't underlined, copy and paste the whole URL into
your browser.
* Although the extension of the request is .ncCF, a .nc file will be
created.
* Or change the file extension of the URL from ".ncCF" to ".html", to
see the form to request data.
point
http://coastwatch.pfeg.noaa.gov/erddap/tabledap/nwioosCoral.ncCF?&time>=2005-01-01
timeSeries
http://coastwatch.pfeg.noaa.gov/erddap/tabledap/cwwcNDBCMet.ncCF?&longitude>=165&time>=2011-08-01T00:00:00Z&time<=2011-08-02T00:00:00Z
trajectory
http://coastwatch.pfeg.noaa.gov/erddap/tabledap/erdGlobecMoc1.ncCF?&time<=2000-09-08T05:12:00
profile
http://upwell.pfeg.noaa.gov/erddap/tabledap/erdGtsppBest.ncCF?&time>=2011-06-21T00:00&time<=2011-06-21T00:10
timeSeriesProfile
http://coastwatch.pfeg.noaa.gov/erddap/tabledap/erdCalcofiSub.ncCF?&time>=1998-01-26T00:00:00Z
trajectoryProfile
http://coastwatch.pfeg.noaa.gov/erddap/tabledap/erdCalcofiSub.ncCF?&time>=1998-01-26T00:00:00Z
This is all untested in the sense that there are currently (to my
knowledge) no clients which can automatically interpret these files as
specific DSG types. If you see problems or have suggestions for
improvement, please let me know.
6) I would be grateful if other people would also make sample files
(particularly of different types) available.
I hope that is all correct and answers the questions.
If not, please let me know.
Good luck.
Thanks,
Roland
--
netcdf trajectories {
dimensions:
index = 15991 ;
dim_one = 1 ;
trdim = 2 ;
cruise_ID_width = 64 ;
subsample_width = 21 ;
trajectory = 330 ;
trajectory_width = 64 ;
variables:
double trdim(trdim) ;
trdim:units = "hours since 1970-01-01 00:00:00" ;
trdim:time_origin = "01-Jan-1970 00:00:00" ;
double trange(trdim) ;
trange:units = "hours" ;
float NUMPROFS(dim_one) ;
NUMPROFS:long_name = "Number of Profiles" ;
NUMPROFS:units = "unitless" ;
float NUMOBS(dim_one) ;
NUMOBS:long_name = "Number of Observations" ;
NUMOBS:units = "unitless" ;
char CRUISE_ID(index, cruise_ID_width) ;
CRUISE_ID:long_name = "CRUISE ID" ;
CRUISE_ID:units = "unitless" ;
CRUISE_ID:missing_value = "" ;
CRUISE_ID:database_table = "data_6hourly" ;
char subsample(index, subsample_width) ;
subsample:long_name = "" ;
subsample:units = "" ;
subsample:missing_value = "" ;
subsample:database_table = "data_6hourly" ;
float latitude(index) ;
latitude:missing = -1.e+34 ;
latitude:units = "degrees_north" ;
latitude:long_name = "Latitude" ;
latitude:database_table = "data_6hourly" ;
latitude:_CoordinateAxisType = "Lat" ;
float longitude(index) ;
longitude:missing = -1.e+34f ;
longitude:units = "degrees_east" ;
longitude:long_name = "Longitude" ;
longitude:database_table = "data_6hourly" ;
longitude:_CoordinateAxisType = "Lon" ;
double sal(index) ;
sal:long_name = "Salinity" ;
sal:units = "PSU" ;
sal:missing_value = -1.e+34 ;
sal:database_table = "" ;
double PROF_ID(index) ;
PROF_ID:long_name = "Profile ID" ;
PROF_ID:units = "unitless" ;
PROF_ID:missing_value = -999. ;
char trajectories(trajectory, trajectory_width) ;
trajectories:cf_role = "trajectory_id" ;
int trajectory_counts(trajectory) ;
trajectory_counts:trajectory_dimension = "index" ;
trajectory_counts:long_name = "Number of observations in this
trajectory." ;
// global attributes:
:featureType = "trajectory" ;
:Conventions = "CF-1.6" ;
:geospatial_lat_min = -69.338996887207 ;
:geospatial_lat_max = 76.3619995117188 ;
:geospatial_lon_min = -179.880996704102 ;
:geospatial_lon_max = 179.998001098633 ;
:time_coverage_start = "9.99999999E8 hours since 1970-01-01 00:00:00" ;
:time_coverage_end = "-9.99999999E8 hours since 1970-01-01 00:00:00" ;
}
_______________________________________________
netcdf-java mailing list
netcdf-java@xxxxxxxxxxxxxxxx
For list information or to unsubscribe, visit:
http://www.unidata.ucar.edu/mailing_lists/
--
Sincerely,
Bob Simons
IT Specialist
Environmental Research Division
NOAA Southwest Fisheries Science Center
1352 Lighthouse Ave
Pacific Grove, CA 93950-2079
Phone: (831)658-3205
Fax: (831)648-8440
Email: bob.simons@xxxxxxxx
The contents of this message are mine personally and
do not necessarily reflect any position of the
Government or the National Oceanic and Atmospheric
Administration.
<>< <>< <>< <>< <>< <>< <>< <>< <><
