In an aggregation pipeline, what does the $group stage do?

The $group stage in an aggregation pipeline is specifically designed to group documents by a specified criterion. This stage allows you to organize documents into subsets based on shared attributes, enabling the application of aggregation operations like counting the number of occurrences, summing values, or calculating averages within those groups.

For instance, if you were to analyze a dataset of sales transactions, you might use the $group stage to aggregate sales data by customer or by product category. This capability greatly enhances data analysis by allowing developers to gain insights from grouped data rather than individual documents, thus facilitating more complex queries and report generation.

The other options involve different functions that do not describe what the $group stage does. Sorting documents pertains to the $sort stage, calculating averages is often handled within the $group stage but is not its defining purpose, and filtering documents is accomplished using stages like $match rather than grouping.