Skip to content

DEV: FT.AGGREGATE expression precedence #1589

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 7 commits into
base: main
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
@@ -0,0 +1,193 @@
---
categories:
- docs
- develop
- stack
- oss
- rs
- rc
- oss
- kubernetes
- clients
description: Order of operations for the FT.AGGREGATE command
linkTitle: Aggregation syntax
title: FT.AGGREGATE order of operations
weight: 2
---

## Overview

[`FT.AGGREGATE`]({{< relref "/commands/ft.aggregate" >}}) is a powerful Redis Query Engine (RQE) command for performing advanced data aggregation, filtering, sorting, and transformations on indexed hash or JSON documents. This reference page provides a structured breakdown of syntax, ordering rules, and best practices.

The [main aggregations page]({{< relref "/develop/interact/search-and-query/advanced-concepts/aggregations" >}}) has a simple diagram showing how `FT.AGGREGATE` pipelines are constructed, but it doesn't tell the whole story. For example, you can create more complex aggregation pipelines by applying multiple `REDUCE` functions under a single `GROUPBY` clause, or you can chain groupings and mix in additional mapping steps:

`GROUPBY` ... `REDUCE` ... `APPLY` ... `GROUPBY` ... `REDUCE`

{{< note >}}
The examples on this page are based on a hypothetical "products" data set, which you can [download here](./data/products.txt).
{{< /note >}}

## Syntax and expression ordering

The `FT.AGGREGATE` command processes multiple expressions in a pipeline. Below is the recommended order:

1. `index` – the name of your index, which must be the first argument.
1. `query` – your query, which must be the second argument.
1. `FILTER` – filters raw documents before transformations or aggregation.
1. `LOAD` – loads additional document attributes.
1. `APPLY` – applies transformations on fields.
1. `GROUPBY` – groups results by specific fields.
1. `REDUCE` – performs aggregations. For example, `SUM`, `COUNT`, and `AVG`.
1. `SORTBY` – orders the results based on specified fields.
1. `LIMIT` – restricts the number of results returned.
1. `DIALECT 2` - provides for more comprehensive syntax, for example using parameters in `FILTER` expressions.

Other keywords will be discussed toward the end of this page.

## When to use `@`

You must add `@` at the start of a field name in the following circumstances:

- When referencing fields loaded from documents. In the following example, `price` is a document field and must be prefixed with `@`.

```sh
FT.AGGREGATE products "*"
LOAD 1 @price
APPLY "@price * 1.1" AS adjusted_price
SORTBY 2 @adjusted_price DESC
LIMIT 0 10
```

- When referencing fields inside a `FILTER` clause that were loaded from documents.

```sh
FT.AGGREGATE products "*"
LOAD 1 @rating
FILTER "@rating >= 4.5"
LIMIT 0 10
```

- When referencing fields inside `GROUPBY` or `REDUCE` clauses.

```sh
FT.AGGREGATE products "*"
GROUPBY 1 @category
REDUCE SUM 1 @price AS total_price
LIMIT 0 10
```

- When referencing fields created by `REDUCE` in an `APPLY` or `FILTER` clauses.

```sh
FT.AGGREGATE products "*"
GROUPBY 1 @category
REDUCE SUM 1 @price AS total_price
APPLY "@total_price * 1.2" AS boosted_price
FILTER "@total_price > 1000"
LIMIT 0 10
```

- When referencing fields created by `APPLY` in another `APPLY` or `FILTER` clause.

```sh
FT.AGGREGATE products "*"
LOAD 2 @price @discount
APPLY "@price - @discount" AS net_price
APPLY "@net_price * 1.1" AS marked_up
FILTER "@net_price > 200"
LIMIT 0 10
```

- When referencing fields created by `APPLY` in a `SORTBY` clause.

```sh
FT.AGGREGATE products "*"
LOAD 2 @price @discount
APPLY "@price - @discount" AS net_price
SORTBY 2 @net_price DESC
LIMIT 0 10
```

## GROUPBY with multiple REDUCE operations

`GROUPBY` can be followed by multiple `REDUCE` operations for different aggregations.

```sh
FT.AGGREGATE products "*"
GROUPBY 1 @category
REDUCE COUNT 0 AS product_count
REDUCE SUM 1 @price AS total_price
REDUCE AVG 1 @rating AS avg_rating
SORTBY 2 @total_price DESC
LIMIT 0 10
```

## Multiple APPLY operations followed by GROUPBY and REDUCE

You can use `APPLY` in various ways before and after `GROUPBY` and `REDUCE`.

```sh
FT.AGGREGATE products "*"
LOAD 3 @price @discount @quantity
APPLY "@price - @discount" AS final_price
APPLY "@final_price * @quantity" AS total_revenue
GROUPBY 1 @category
REDUCE SUM 1 @total_revenue AS total_category_revenue
SORTBY 2 @total_category_revenue DESC
LIMIT 0 10
```

## FILTER and PARAMS

Use `FILTER` to remove unwanted records, and `PARAMS` to pass values to parameterized queries.

```sh
FT.AGGREGATE products "*"
LOAD 3 @price @rating @quantity
FILTER "@price >= 500"
FILTER "@rating >= 4.0"
APPLY "@price * @quantity" AS total_value
SORTBY 2 @total_value DESC
LIMIT 0 10
DIALECT 2
```

## Placement of FILTER before and after GROUPBY/APPLY

- **Before GROUPBY:** Removes records before aggregation.
- **After GROUPBY:** Filters based on aggregated results.
- **Before APPLY:** Ensures calculations are applied only to certain records.
- **After APPLY:** Filters computed values.

## LOAD after GROUPBY/REDUCE

This is not allowed and you'll get a syntax error.

## Placement rules for specific parameters

| Parameter | Placement |
|----- |----- |
| `TIMEOUT` | Can be placed anywhere. |
| `LIMIT` | Must be at the end, before `DIALECT`. |
| `WITHCURSOR` | Must be at the end, before `DIALECT`. |
| `SCORER` | Can be placed anywhere. |
| `ADDSCORES` | Must be before sorting. |
| `DIALECT` | Must be at the end. |

## LIMIT and WITHCURSOR used together

While you wouldn't ordinarily use `LIMIT` and `WITHCURSOR` together in the same query, you can use them advantageously if doing so fits your workflow.
`LIMIT` returns immediate results, while `WITHCURSOR` retrieves results incrementally using the [cursor API]({{< relref "/develop/interact/search-and-query/advanced-concepts/aggregations/#cursor-api" >}}).

```sh
FT.AGGREGATE products "*"
GROUPBY 1 @category
REDUCE COUNT 0 AS product_count
LIMIT 0 100
WITHCURSOR COUNT 3
```

See the following resources for more information:

- [`FT.AGGREGATE` command page](https://redis.io/docs/latest/commands/ft.aggregate/)
- [RQE source code](https://github.com/RediSearch/RediSearch/tree/master/src/aggregate)
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@ categories:
description: Learn how to use the autocomplete feature of Redis for efficient prefix-based suggestion retrieval.
linkTitle: Autocomplete
title: Autocomplete with Redis
weight: 1
weight: 2
---

## Overview
Expand Down
Loading