How to cluster data over new coords?

[loganchen39]

I have the initial data SST (Sea Surface Temperature) over a fine lat/lon resolution of 0.02 degree. So lat: -89.99, -89.97, ..., 89.97, 89.99, 9000 in total; lon: -179.99, -179.97, ..., 179.97, 179.99, 18000 in total.
I want to map (with nearest neighbot) the SST data to a new coarser new_lat/new_lon grid with 0.25 degree, so new_lat: -89.875, -89.625, ..., 89.625, 89.875; and new_lon: -179.875, -179.625, ..., 179.625, 179.875. So for one certain location (new_lat, new_lon), there might be several corresponding initial SST data, and I want to take the average of them to be the result data on that certain location.
Does xarray provide such functions which allows us to operate on the arrays directly to get the result? Or do we have to loop over all the grid points to calculate manually? I can do the latter, but the time efficiency is not good. Thanks,

[raybellwaves]

Without understanding fully what you are trying to do there is http://xarray.pydata.org/en/stable/user-guide/interpolation.html and https://pangeo-xesmf.readthedocs.io/en/latest/