Sort Data Transform

The Sort Data transform orders rows by one or more columns.

Basic Usage

To sort rows:

1. Select the Sort Data transform from the transform menu.
2. Choose a field under Target Columns.
3. Set its Order to ascending or descending.
4. To add another sort column, click +, then choose its Target Columns field and Order. Repeat as needed.
5. Apply the transformation.

Configuration Options

Basic Options

• Target Columns: Select one column in each sort row. Click + to add a row. The order of these rows sets the sort priority.
• Order: For each selected column, choose either:
  • Ascending (A to Z, 0 to 9)
  • Descending (Z to A, 9 to 0)

note

The first column is the primary sort. Each later column breaks ties left by the columns before it.

tip

The workflow UI puts null/NaN values last for both sort directions. Direct backend configurations can set na_position to first or last, but this setting is not available in the UI.

Examples

Example: Sorting a Sales Dataset

Input Dataset:

Date	Product	Sales	Region
2023-05-15	B	1500	North
2023-05-14	A	1000	South
2023-05-16	C	2000	East
2023-05-15	A	1200	West
2023-05-14	B	800	North

Configuration:

• Column 1: Date (Ascending)
• Column 2: Product (Descending)

Result:

Date	Product	Sales	Region
2023-05-14	B	800	North
2023-05-14	A	1000	South
2023-05-15	B	1500	North
2023-05-15	A	1200	West
2023-05-16	C	2000	East

The rows are sorted by Date in ascending order, then by Product in descending order within each date.

Best Practices

1. Put the most important sort column first.
2. Check that each sort column has a consistent data type.
3. Account for null/NaN values appearing last.
4. Add a row-number column first if you may need to recover the original order.

Troubleshooting

• The order looks wrong: Check the sort column's data type and look for mixed values.
• Dates sort incorrectly: Use one consistent format and confirm that the column is recognized as a date.
• Numbers sort incorrectly: Check for text values or hidden characters.