Currently, aggregation and resampling methods are part of the data cube view. This has the following disadvantages:
Methods cannot be specified per band.
Reusing data cube views for different datasets might be limited if both datasets require different methods.
It would be much cleaner to define aggregation and resampling methods as parameters to the image_collection_cube class. The data cube view definition should be completely independent from particular datasets.
Currently, aggregation and resampling methods are part of the data cube view. This has the following disadvantages:
It would be much cleaner to define aggregation and resampling methods as parameters to the
image_collection_cube
class. The data cube view definition should be completely independent from particular datasets.