Do not input private or sensitive data. View Qlik Privacy & Cookie Policy.
Skip to main content

Announcements
Learn how to migrate to Qlik Cloud Analytics™: On-Demand Briefing!
cancel
Showing results for 
Search instead for 
Did you mean: 
RobO
Contributor
Contributor

Sankey Diagram Visual

I'm looking to build a sankey diagram that has 3 or more dimensions.  Each of these dimensions will reduce the amount count of records as they go along.  I want to be able to remove any NULL values for the 3rd dimension but still keep the total record count for the first and second dimensions the same.  The screenshot is from Sankey in Qlik Sense.  The large red bar is all NULL so I don't want them showing.

RobO_1-1600280056066.png

 

The first dimension has 5 categories.  The second dimension has 6 categories.  The third dimension will have 4 categories.

The first dimension will flow into each of the 6 categories in the second dimensions but only 1 of the categories in the second dimension will flow into the 4 categories of the third dimension.  Below is a screenshot (from Power BI) of what I want to do in Qlik Sense Cloud.

Is there a way to do this?

RobO_0-1600279700380.png

 

1 Reply
William11
Contributor
Contributor

Yes, you can definitely replicate that type of multi-level flow in Qlik Sense Cloud. To build a similar structure, you’ll need to use a Sankey diagram extension (like the one available in the Visualization Bundle or a custom extension from Qlik’s Developer Hub).

Here’s the basic approach:

  1. Prepare your data model so that it clearly defines the three dimensions and their relationships (e.g., Category1 → Category2 → Category3).
  2. Load the data with proper key fields that represent each connection between dimensions.
  3. Use the Sankey chart and assign:
    • Source: your first and second dimensions
    • Target: your second and third dimensions
    • Measure: a numeric field representing the flow value or count

Since only one of your second-dimension categories needs to connect to the third dimension, you can manage this easily by using a conditional expression or filtering the data before loading it into the Sankey.

To get a visual idea of how this flow should look, you can check this interactive example:
👉 https://app.powerbi.com/view?r=eyJrIjoiOTEzZTdjZGUtNzE5MS00ODk2LWJkOTAtNTI3ZjMwNWI1MTdhIiwidCI6IjYy…

That example shows a similar multi-layer flow setup that matches what you’re describing.

Hope this helps you get the structure right in Qlik Sense!

— Rob