issue_comments

@fmaussion done, let's see what CI has to say about my patches ;)

I remember reading a thread somewhere on xarray github repo discussing whether xarray should include the rasterio backend or not.

I understand that bridges between two libraries are always hard to maintain, because you need to know both products (we actually have the same kind of problem with OTB and QGis), but from a user standpoint, they need to exist somewhere. I would probably never have turned to xarray if someone with the required knowledge had not implemented the rasterio backend.

Then of course the user community should take care of maintaining those backends (this is what I am doing right now).

Bridging xarray with rasterio opens xarray to the remote sensing imagery community. And behind rasterio there is gdal, which is an awesome library with so many great capabilities (like this on-the-fly reprojection during reading I mentioned).

@fmaussion done.

That's a tough question ;) In the current dataset I have 950 unique labels, but in my use cases it can be be a lot more (e.g. agricultaral crops) or a lot less (adminstrative boundaries or regions).

Many thanks for your answers @shoyer and @rabernat .

I am relatively new to xarray and dask, I am trying to determine if it can fit our need for analysis of large stacks of Sentinel data on our cluster.

I will give a try to dask.array.histogram ass @rabernat suggested.

I also had the following idea. Given that: * I know exactly beforehand which labels (or groups) I want to analyse, * .where(label=xxx).mean('variable') does the job perfectly for one label,

I do not actually need the discovery of unique labels that groupby() performs, what I really need is an efficient way to perform multiple where() aggregate operations at once, to avoid traversing the data multiple time.

Maybe there is already something like that in xarray, or maybe this is something I can derive from the implementation of where() ?