Bitwise operators are conflated with intersection, union, symmetric difference operators in NB 03.01

[deburg0]

In the Section "Index as ordered Set", Bitwise operators are mistakenly equated with Set Operators: intersection, union, symmetric difference.

indA = pd.Index([1, 3, 5, 7, 9])
indB = pd.Index([2, 3, 5, 7, 11])

indA.intersection(indB)
Index([3, 5, 7], dtype='int64')

indA & indB # intersection (actually binary AND)
Index([0, 3, 5, 7, 9], dtype='int64')

Not a big deal, but confusing.

Thank you for your book.

[tearsaalainah26]

The confusion between the & operator and the intersection method in the context of pd.Index is indeed noteworthy. The & operator performs a bitwise AND, which can lead to unexpected results if users are not aware of this distinction. It might be helpful to clarify this in the documentation by providing a specific example showing both methods side by side, highlighting the difference in their outputs. Have you considered adding a section that explicitly explains the difference to prevent this misunderstanding?

[aleanacamry26]

It might be useful to add a note in the documentation for pd.Index about the difference between indA.intersection(indB) and indA & indB. Including a brief example that shows the distinct behavior of both methods could prevent confusion for users unfamiliar with bitwise operations. Additionally, could we mention that while the & operator is meant for bitwise operations, users should rely on the intersection method for set-like behavior? This could enhance understanding without requiring users to dig deeper into the documentation.