Video Transcript (English)

Hi folks! In this video, we’ll introduce and explore the MongoDB aggregation framework that's used to build multi-stage queries. By the end of the lesson, you will be able to identify the components of an aggregation pipeline.

In the context of databases, aggregation is the analysis and summary of data.

An aggregation stage is an aggregation operation that’s performed on the data and does not permanently alter the source data.

MongoDB took this concept a step further by creating an aggregation pipeline, in which the developer specifies the aggregation operations one at a time, in order.

We can run a single aggregation stage to get the information that we need. But what makes aggregation in MongoDB unique is that we can string together aggregation stages to create a pipeline. A pipeline consists of multiple stages, where data can be filtered, sorted, grouped, and transformed. Documents that are output from one stage become input into the next stage

In MongoDB Atlas, we can use a visual editor to create and test aggregation pipelines by adding stages, one-by-one. We can see the results from each stage. This is a great way to start with aggregation, and you may want to use this tool to explore the different stages.

However, for the purposes of this course, we will be showing aggregation done using the MongoDB CLI, so that you can see the whole pipeline more easily in the video.

There is also support for aggregation in MongoDB language drivers, so take a look at the documentation for the language of your choice.

Let's break down the pieces of an aggregation pipeline using dot-notation on the collection - db.collection_name and the aggregate function, which takes an array of aggregation stages to form the pipeline.

First, let’s learn about the stages of aggregation. Each stage is a single operation on the data.

Some of the commonly used stages are: $match, which filters for data that matches certain criteria, $group, which groups documents based on criteria, and $sort, which puts the documents in a specified order.

These operations are useful both individually and together in a pipeline for analyzing the data.

Each stage includes syntax to carry out the operation. The syntax for each stage is dependent on the stage - if you are unsure about the syntax, refer to documentation for that particular stage.

The syntax for some stages may include expressions. For example, the $match stage expects the same syntax as the MongoDB find method.

In most other aggregation stages, like this example with $project, you can expect to use expression operators.

When looking at aggregation syntax, you may notice that sometimes field names are prefixed with a dollar sign. This is a field path. It allows us to refer to the value in that field. In this example, we’re setting a new field called default Username to a concatenated string composed of the values from the first_name field and the last_name field.

In this video, we discussed what aggregation is and how to set up an aggregation pipeline. We also explained different aggregation stages, operators and expressions, and field references.

These basics of aggregation will take you far! The next step is to explore each aggregation stage and learn more about what they can do.