Stack data based on a selection of nested header names.
Add Unnest to any node in your Dataflow. Select any node and use the Transform (+) button to open the list of options to select Unnest. You could work with the node in the Dataflow tab or use the Properties panel.
The Unnest node is broken up into two sections:
Use the first dropdown to select which columns you would like to Unnest. Headers that match the same nested group will be stacked on top of each other.
In our example, multiple columns contain the word "lineup" in the header (i.e., lineup.player_id, lineup.player_name). Each column containing a matching header is stacked, producing a more workable dataset.
Select the columns you would like to remain in your data from the Keep Columns dropdown. In the example below, the column "team_name" remains, and its values are stacked for each unnested record.
You could click the gear icon to choose how to select columns. There are two options:
Your data may not have nested columns. In other words, each column name is most likely a unique name.