drop target_dimension in aggregate_spatial?

[soxofaan]

While working on a vector cube implementation with aggregate_spatial I started wondering about the point of the target_dimension parameter: what is that for actually and how should that work?

aggregate_spatial starts from a raster data cube (e.g. x-y-t-bands) and aggregates spatially to a vector cube e.g. with dimensions geometry-t-bands.
So you start from a raster data cube and generate a completely new vector cube. You are not updating or appending to the original raster data cube, which this target_dimension seems to suggest in a way.

I don't think we need target_dimension. Unless that is about naming the "geometry" dimension (or whatever we are going to standardize on, I can't find the related issue at the moment)?

[m-mohr]

The new definition in #382 should clarify what target_dimension is used for in the future. Happy to receive feedback about it.

[soxofaan]

ok thanks. With #382 being the way forwards, I think we can close this one