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

# Expression builder

> Expression Builder

Analyzing and manipulating data with Spark functions often requires building complex expressions that combine multiple functions. Prophecy makes this easier with the **Expression Builder**, which streamlines the creation of complex expressions, saving time and effort. It also helps you better understand the relationships between functions and their arguments.

## Spark Expression Builder

To navigate to the Spark Expression Builder:

1. Navigate to the column you want to edit, and open the expanded editor.
2. Click on the **Expression Builder** button.

Now, you can search for, and insert functions, columns or configurations in your canvas.

### Search

To search for a function in Spark Expression Builder:

1. Click the **fx** button.
2. In the search bar at the top of the screen, type the function name.

A list of matching functions appears. When you click on a function, you see information about its syntax, arguments, and an example on how to use it.

### Insert

You can insert a function, column, or configuration into your expression.

To insert a function:

1. Navigate to, or search for the function you want to use.
2. Click on the **Insert Function** button.

To specify an argument for the function, click on the function and fill in the required fields.

To insert a column:

1. Click the **Columns** tab.
2. Click on the column you want to insert, and click **Insert Column**.

To insert a configuration:

1. Click the **Configurations** tab.
2. Click on the column you want to insert, and click **Insert Configuration**.

### Run and Verify the output

You can attach to a cluster and run your pipeline for the current gem on the same screen.

To run your code:

1. Click the **Run** button.
2. Click the **Data** button to see your results.

Verify the results to ensure that your data analysis tasks are accurate and reliable.

<Note>This data is same as what you see in [interims](/data-engineering/development/runs/data-sampling) view.</Note>