Скачать или смотреть Using Yield for Memory-Efficient Data Handling

The 'yield' statement is central to the functionality of a generator. Unlike 'return', which exits a function, 'yield' provides a value and pauses the function's execution, allowing it to resume later from where it left off. This behavior makes generators exceptionally memory efficient because they only generate data as needed. This is crucial when dealing with large datasets or streams because the entire dataset is not loaded into memory at once. Instead, data is processed one item at a time. Use 'yield' when you want to iterate over a sequence without the overhead of creating and storing the entire sequence in memory. For example, generating Fibonacci numbers can be efficiently done with a generator, where each number is produced only when requested.