# INDEXBY

INDEXBY( `column`, `integer`, `partition column`, `[sort column]` , `[ascending]`)

Indexes a given column when grouped by unique values within a `partition column`

{% hint style="info" %}
INDEXBY() is a [**window function**](https://docs.cascade.io/cascade/building-expressions/guide-to-window-functions), which performs like INDEX() except partitioned by the unique values of a different column.
{% endhint %}

### Arguments

| Argument             | Description                                                                                          |
| -------------------- | ---------------------------------------------------------------------------------------------------- |
| `column`             | Column to index                                                                                      |
| `integer`            | Index of indexed `column` to ouput                                                                   |
| `partition` `column` | Column containing unique values                                                                      |
| `[sort column]`      | Optional. Column with which to sort partitioned rows by. Defaults to index.                          |
| `[ascending]`        | Optional. Enter `True` to sort in ascending order (default), or `False` to sort in descending order. |

### Output

Outputs a given index from each partitioned group within the indexed `column`

### Examples

Let's say that we want to find the first state in each region in a table. We have a table that lists each state and its corresponding region -- by using `INDEXBY({State}, 1, {Region})` , we can select the first state for whatever region that state is in.

{% embed url="<https://datawrapper.dwcdn.net/LzfAE/1/>" %}