Closed lbesnard closed 9 years ago
I'd call this a bug. history
is not a data variable, so this check shouldn't apply. This needs to be fixed in the helper function that identifies data variables.
@mhidas based on the definition of data variable, history is not coordinate and ancillary variable. Do you think this is a special case for data variable?
@fxmzb123 Let's add to the definition of a data variable the condition that it needs to have numeric type. This would exclude the history
variable.
@mhidas fixed and see pull request #52
imos checker output:
ncdump
file http://thredds.aodn.org.au/thredds/catalog/IMOS/SOOP/SOOP-ASF/VNAA_Aurora-Australis/flux_product/2013/catalog.html?dataset=IMOS/SOOP/SOOP-ASF/VNAA_Aurora-Australis/flux_product/2013/IMOS_SOOP-ASF_FMT_20130315T000200Z_VNAA_FV02.nc
The history variable is required to have a XYZT dimension. I don't think this should be the case. Maybe a black list of variables could be created (in a text file) ?