Have you tried xray.open_mfdataset? All of the examples should work the same with a dataset loaded with that function as from a single file.

In your case, something like the following should work:

``` python

load data

ds = xray.open_mfdataset('path/to/my/files/*.nc')

calculate anomalies

clim = ds.groupby('time.month').mean('time') anom = ds.groupby('time.month') - clim

plot anomalies over time

(in practice, would probably want to use .sel here to do

labeled lookups)

anom.temperature.isel(x=0, y=0).to_pandas().plot()

plot anomalies over space

plt.imshow(anom.temperature.isel(time=0).values) ```

Plotting is currently not so easy as it should be with xray (hence why you see me exporting everything to pandas), but that's something we plan to start work on very soon.