How to create a LandMask from arbitrary data 
--------------------------------------------

Instantiating masks directly can be done as follows.

.. plot::
    :include-source:
    :context: close-figs

    for proportion in [0.1, 0.5, 0.9]:
        anydata = np.zeros((500, 1000), dtype=bool)

        whr = np.random.randint(0, anydata.size, size=int(anydata.size*proportion))
        anydata.flat[whr] = True

        lm = dm.mask.LandMask.from_array(
            anydata)
        
        fig, ax =  plt.subplots()
        lm.show(ax=ax)
        plt.show()