On 3/9/2012 4:13 AM, Kristian Sebastián wrote:
Hi,
We have configured a Forescat Model Run Collection from Wmed output
(http://thredds.socib.es/thredds/catalog/operational_models/oceanographical/hydrodynamics/model_run_aggregation/wmed/catalog.html).
Some files that generates the feature collection, such as the "Best
time series" file and the "Constant Forecast Date" files, contains
data with NaN values. I have check it through godiva.
For example, all the data with date 2012-03-06T00:00:00Z
<http://thredds.socib.es/thredds/godiva2/godiva2.html?server=http://thredds.socib.es/thredds/wms/operational_models/oceanographical/hydrodynamics/model_run_aggregation/wmed/forecast/wmed_ConstantForecast_2012-03-06T00:00:00Z> has
NaN values, in the "Best time series" files too. I have checked the
original files and the data seems to be right with the same date
2012-03-06T00:00:00Z
<http://thredds.socib.es/thredds/godiva2/godiva2.html?server=http://thredds.socib.es/thredds/wms/operational_models/oceanographical/hydrodynamics/model_run_aggregation/wmed/files/roms_wmed_20120306_0000.nc>.
We have installed Thredds v.4.2.20 and in my PC I have installed the
latest version, with same results. The featureCollectionScan.log
and catalogInit.log doesn't show errors.
hi kristian:
if you have run it on version 4.2.9 and still see the problem, then i
will need to reproduce it. can you make a minimal aggregation with the
problem, then send me the files and ncml?
thanks,
john
