mmaelicke
commented
3 years ago Additionally, the ListExporter needs an option to apply a unified (temporal) scaling by cutting extent and aggregate resolution.
For upscaling: remove (data) from the merged dataset if the resolution does not match the needed target scaling.
For all variables, a hardcoded mapping is defined, which may be overwritten by Entry.details.get('metacatalog').get('aggregation function')`:
Sum: <7> daily rainfall sum, <19> evaptranspiration, <20> drainage
Max: <13> precision
Mean: everything else.
Implement this only optionally, as .harmonize_scale and a property to yield the excluded Entry due to scaling
The export extension can export a single ResultSet. A new class is needed to wrap many ResultSets and drop everything that is duplicated, based on their md5 sums. This list can then iteratively export all contents into a tarball. Alternatively, some export endpoints might support merging.