Row Denormaliser

The Row denormaliser step can be used to try to improve performance either by adding redundant copies of data values or by grouping data. You can also use this step to convert data types.

You must normalize the data before you can denormalize it.

For example, consider the following set of normalized data:

Produce Category	Delivery Time	Produce Type
fruit	2018/01/05 08:00:00	apples
fruit	2018/01/05 08:10:00	oranges
fruit	2018/01/05 08:20:00	apples
fruit	2018/01/05 08:30:00	oranges

The data can be denormalized by grouping on the type of fruit:

Produce Category	Apples	Oranges
fruit	2018/01/05 08:00:00	2018/01/05 08:10:00
fruit	2018/01/05 08:20:00	2018/01/05 08:30:00

General tab

Enter the following information in the transformation step fields:

Field	Description
Step name	Specify the unique name of the transformation step on the canvas. The default is Row denormaliser.
Key field	Define the key of the output row. The available fields come from the incoming PDI data stream. Another step with output data must be connected to the Row denormaliser step for options to appear in this field.

The fields that make up the grouping

Group field table in Row denormaliser

Use the The fields that make up the grouping table to specify which fields from the incoming PDI data stream define the grouping for the denormalization process.

You can click Get Fields to automatically fill the table with all fields from the incoming PDI data stream.

For more information, see Understanding PDI data types and field metadata.

Target fields

Target fields table in Row denormaliser

Use the Target fields table to select the fields to denormalize by specifying the string value for the Key field.

Options are provided to convert data types. Strings are most common as key-value pairs, so you often convert to Integer, Number, or Date.

If you get key-value pair collisions (the key is not unique for the specified group), specify the aggregation method to use.

You can click Get lookup fields to retrieve fields from the PDI data stream.

Examples

The data-integration/samples/transformations directory contains the following example KTRs that show how to use this step:

Example KTR file	Description
Denormaliser – Simple Example.ktr	Converts a ‘month’ row in a record to multiple ‘month’ columns.
Denormaliser – Data Conversion and Aggregation.ktr	Converts data values by ‘month’.
Denormaliser – 2 Series of Key-Value Pairs.ktr	Converts ‘name’ and ‘ID’ rows to multiple ‘name’ and ‘ID’ columns.
Denormaliser – Capturing Last State Timestamp.ktr	Aggregates data values by ‘state’.

Metadata injection support

All fields of this step support metadata injection. You can use this step with ETL metadata injection to pass metadata to your transformation at runtime.