> ## Documentation Index
> Fetch the complete documentation index at: https://parabola.io/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# How to check and filter a column based on the uniqueness of a cell?

> There are many ways to work with unique values in Parabola. This article focusses on how to check if a cell value is unique or not. The second section outlines how other steps interact with uniqueness.

## Checking if a cell value in unique

In the **Filter rows** step and the **Add if/else column** step, the "is unique" and "is not unique" operations can be used to check for uniqueness.

**Is unique**: returns true for any row where the cell value of the column being checked is unique across the dataset.

**Is not unique**: returns true for any row where the cell value of the column being checked is not unique across the dataset.

### Example usage

* A step set to "keep rows where column A is not unique" will return any rows, and their duplicates, that are not considered unique. This is a good way to find any rows with duplicate values, and work with them. The inverse will also work: "remove rows where column A is unique".
* A step set to "Keep rows where column A is unique" will return any rows that have a unique value in column A. This is useful for ensuring that a dataset only has unique values in a specific row, like an ID field. The inverse will also work: "remove rows where column A is not unique".

## Other steps that take uniqueness into account

* Remove duplicates: Keeps a specific number of duplicates as defined by the values of a specific column. Setting this to "1" will result in a unique dataset.
* Merge duplicates: Merges specific columns across any rows that share the same value in another column.
* Count by group: Count's the number of instances of each unique value in a specified column. Other "by group" steps like Sum by group, Average by group, Find min by group, Find max by group also use grouping based on uniqueness.
* Add rows numbers: Some settings allow rows to be numbered by grouping unique values in a specific column. Non-unique rows will have a number higher than 1 assigned to them.