Learn · Tutorial

3. Pipes establish order

Concept

Read a pipe from top to bottom: each stage receives the previous result, and the authored order becomes the syntax of the analytical statement. Run the shared setup first.

Tiny example

daily_anomaly = (
    pipe(cube)
    | v.anomaly(dim="time")
    | v.mean(dim=("y", "x"), keep_dim=False)
).unwrap()

Explanation

Subtract each cell's mean over the observed period, then average those departures across space. Parentheses let the pipe span lines; .unwrap() returns the underlying result. CubeDynamics calls stages once, from left to right, and does not rearrange them. The pipe does not choose a baseline or units automatically.

unwrap() is a local statement boundary, not a compute or certification step. It returns the ordinary xarray value and leaves any Dask-backed work deferred.

Try it / worked example

daily_anomaly.plot()
plt.axhline(0, color="black", linewidth=0.8)
plt.show()

Negative values mean colder than this January extract's mean, not necessarily colder than a multi-decade climate normal.

What to learn next

4. Order can change meaning · Pipe · anomaly · Grammar vignette