High Performance Python, 3rd Edition

Chapter 4. Dictionaries and Sets

Questions You’ll Be Able to Answer After This Chapter

• What are dictionaries and sets good for?

• How are dictionaries and sets the same?

• What is the overhead when using a dictionary?

• How can I optimize the performance of a dictionary?

• How does Python use dictionaries to keep track of namespaces?

Sets and dictionaries are ideal data structures to be used when your data has no intrinsic order (except for insertion order) but does have a unique object that can be used to reference it (the reference object is normally a string, but it can be any hashable type). This reference object is called the key, while the data is the value. Dictionaries and sets are almost identical, except that sets do not actually contain values: a set is simply a collection of unique keys. As the name implies, sets are very useful for doing set operations (such as union, intersection, and difference).

Note

A hashable type is one that implements both the __hash__ magic function and either __eq__ or __cmp__. All native types in Python already implement these, and any user classes have default values. See “Hash Functions and Entropy” for more details.

We saw in the previous chapter that for lists with no intrinsic order we are limited to O(n) lookup time. Dictionaries and sets give us O(1) lookups based on the arbitrary index. In addition, like lists/tuples, dictionaries and sets have O(1) insertion time.¹ As we will see in “How Do Dictionaries and Sets Work?”, this ...