Example in Chapter 9 Aggregate Methods use NumPy backend without informing the reader.

[sjgtheunissen]

In 9.2 there is an example to calculate the percentage of cars with mileage greater than 20. Both these examples use NumPy backend, while the data was loaded in, in Chapter 7 using PyArrow backend (and in the erata using NumPy backend).

city_mpg.gt(20).sum() works correctly
city_mpg.gt(20).mul(100).mean() does not work with PyArrow backend TypeError: operation 'mul' not supported for dtype 'bool[pyarrow]' with object of type <class 'int'>