Plotting methods

What is your issue?

From @TomNicholas

It would be nice to have plotting methods that allow for easily comparing variables in different parts of a tree.

One simple way would be to have a dt.plot(variable, **kwargs) method which uses a legend to distinguish between variables, i.e.

def plot(self, variable, **kwargs):
    fig, ax = plt.subplots()

    for node in self.subtree:
        da = node[variable]
        da.plot(ax=ax, label=node.name, **kwargs)

    ax.legend()
    ax.set_title(variable)
    plt.show()

The use cases for this would be multi-resolution datasets, or multi-model ensembles, where the user wants to see how a single variable varies across different datasets.

Care would have to be taken to ensure that incompatible dimensions or coordinates were handled smoothly.

Start from the proposed dt.plot entry point and inspect how node[variable] and da.plot are expected to work across the tree. Define completion as comparing one variable across nodes with a legend and title while handling incompatible dimensions or coordinates smoothly; no file or test is identified in the issue.