# Aggregations

{% hint style="success" %}
This is the official documentation of the `@forestadmin/agent` Node.js agent.
{% endhint %}

An aggregation represents a query to a collection that aggregates on records.

They are simple 3 keys objects:

* An `operation` that specifies how the data should be aggregated (`Count`, `Sum`, `Avg`, `Max`, `Min`)
* A `field`, that specifies the data that should be aggregated
* `Groups`, which may be rounded when they are Dates

Supported group rounding operations are `Year`, `Month`, `Week`, `Day`, and `null` (let the field as it is).

#### Count records

The simplest possible query is to count records from a collection.

```json
{ "operation": "Count", "field": null, "groups": [] }
```

Equivalent in SQL: `SELECT COUNT(*) FROM books`

Equivalent in Mongo: `{ $count: 'value' }`

#### Average rating

```json
{ "operation": "Average", "field": "rating", "groups": [] }
```

Equivalent in SQL: `SELECT AVG(rating) FROM books`

Equivalent in Mongo: `{ $group: { _id: null, rating: { $avg: "$rating" } } }`

#### Average rating by author

```json
{ "operation": "Average", "field": "rating", "groups": [{ "field": "author:name" }] }
```

Equivalent in SQL: `SELECT authorName, AVG(rating) FROM books GROUP BY 1`

Equivalent in Mongo: `{ $group: { _id: "$authorName", rating: { $avg: "$rating" } } }`

#### Average rating by author and year

```json
{
  "operation": "Average",
  "field": "rating",
  "groups": [
    { "field": "authorName" },
    { "field": "createdAt", "operation": "Year" }
  ]
}
```

Equivalent in SQL: `SELECT authorName, TO_YEAR(createdAt), AVG(rating) FROM books GROUP BY 1, 2`

Equivalent in Mongo: `{ $group: { _id: { author: "$authorName", created: { $year: "$createdAt" }, rating: { $avg: "$rating" } } }`