Currently, we assume that DELTA_PLUS_VAR and DELTA_MINUS_VAR are attached to "data" variables. This is specified in the ISTP compliance page:
https://spdf.gsfc.nasa.gov/istp_guide/vattributes.html
So right now if these uncertainty variables are attached to a support_data, it throws the error message in _verify_dimension_sizes(created_data_vars, created_coord_vars), when it has the delta variable in the create_data_variables, and then throws an error saying that the dimensions are all wrong because it is expecting that an "epoch" dimension is present.
So we either need to verify that indeed uncertainty can only be attached to data rather than support_data, in which case we should make the error message more obvious, or we should set the uncertainty variable to be the same data type as the parent data type, and make sure doing that doesn't break anything.
Currently, we assume that DELTA_PLUS_VAR and DELTA_MINUS_VAR are attached to "data" variables. This is specified in the ISTP compliance page:
https://spdf.gsfc.nasa.gov/istp_guide/vattributes.html
So right now if these uncertainty variables are attached to a support_data, it throws the error message in _verify_dimension_sizes(created_data_vars, created_coord_vars), when it has the delta variable in the create_data_variables, and then throws an error saying that the dimensions are all wrong because it is expecting that an "epoch" dimension is present.
So we either need to verify that indeed uncertainty can only be attached to data rather than support_data, in which case we should make the error message more obvious, or we should set the uncertainty variable to be the same data type as the parent data type, and make sure doing that doesn't break anything.