Add Attributes Step

Adds new attributes to your data flow using ONE expressions.

This step is available for all transformation types: standalone plans, embedded plans, transformation rules, and transformation catalog items.

Overview

The Add Attributes step creates new columns in your data flow by evaluating expressions. Use this step when you need to derive new values from existing data, add calculated fields, or create constants.

Use this step to:

Combine multiple attributes into a single value.
Calculate derived values (totals, averages, percentages).
Add constant values or flags to your data.
Format or transform existing data into new attributes.
Create lookup keys by combining multiple fields.

Configuration

Add an attribute

In the step configuration, select Add expression.
Configure the new attribute:
- Attribute name: Enter a name for the new attribute. Use clear, descriptive names that indicate the attribute’s purpose.
- Data type: Select the appropriate data type (String, Integer, Long, Double, Boolean, Date, Datetime).
- Expression: Define the value using a ONE expression.
Select Add expression again to add additional attributes as needed.

Expression examples

Concatenate text values

trim(first_name + ' ' + last_name)

Creates a full name by combining first and last name, trimming whitespace.

Calculate a percentage

iif(total_tasks = 0, null, round((toFloat(completed_tasks) / toFloat(total_tasks)) * 100, 2))

Calculates completion percentage with type conversion and division-by-zero protection.

Conditional value

iif(amount > 1000, 'High', 'Standard')

Assigns a category based on a condition.

Extract year from date

datePart(order_date, 'YEAR')

Extracts the year component from a date field.

Create a composite key

region_code + '-' + toString(customer_id)

Combines multiple fields into a single key value.

Add a constant

'PROCESSED'

Adds a static value to all records.

Best practices

Use descriptive attribute names: Choose names that clearly indicate what the attribute contains. Avoid generic names like new_column or calc1.
Match data types to expressions: Ensure the selected data type matches what your expression returns. A mismatch can cause errors or unexpected results.
Handle null values: Consider what should happen when input values are null. Use functions like coalesce() or isnull() to provide default values.
Keep expressions focused: Each attribute should serve a single purpose. Complex logic is easier to maintain when split across multiple attributes.
Test expressions: Use the expression testing feature to validate your expressions before running the full transformation.

Comparison with Transform Data step

Aspect	Add Attributes	Transform Data
Purpose	Creates new attributes	Modifies existing attribute values
Original data	Preserved (new columns added)	Overwritten (values replaced)
Use case	Derive new values while keeping originals	Clean or standardize existing values

Aspect

Add Attributes

Transform Data

Purpose

Creates new attributes

Modifies existing attribute values

Original data

Preserved (new columns added)

Overwritten (values replaced)

Use case

Derive new values while keeping originals

Clean or standardize existing values