Visualizations failing to show

[Erik_L]

I have a SQL query that generates a table. I created a visualization from that table with the UI. I then have a widget that updates a value used in the query and re-runs the SQL, but then the visualization shows nothing, that there is "1 row," but if I select the table it shows all the rows and if I click "edit visualization" it shows the correct visualization. What's wrong with visualizations?

[Kaniz]

Hi @Erik_L , It seems like you’re encountering an issue with your visualization in Databricks.

Let’s troubleshoot this!

Here are a few common reasons why visualizations might not display as expected:

1. Data Issues:

   • Ensure that your SQL query is correctly generating the table data. Check for any missing or incomplete data, formatting issues, or incorrect data relationships within your dataset.
   • Verify that the widget’s updated value is being correctly incorporated into the query.

2. Filters and Slicers:

   • Visual-level filters and slicers can impact which data is displayed in your visuals. If they are not set correctly, they may inadvertently filter out all data, resulting in empty visuals.
   • Consider checking the interactions between your slicer and other visuals. By default, if nothing is selected in a slicer, all records are shown in visuals where the interactions are set as filters. You can customize this behavior.

3. Visual-Level Filters:

   • Sometimes, visual-level filters within Power BI visuals can affect their display. Make sure you haven’t applied filters that are too restrictive.
   • To address this, you can create a measure that checks if there is a filter on the desired column. For example:

     SlicerCheck = IF(
         CALCULATE(DISTINCTCOUNT([SlicerColumn]), ALLSELECTED([SlicerColumn])) = 1,
         "Y",
         "N"
     )
     

     Then place this measure in a table visual and set the filter condition accordingly¹.

4. Index Column:

   • Try adding an index column to your data and include it in your visual. This can help avoid issues related to duplicate data.

Remember to thoroughly review your query, filters, and interactions to identify any discrepancies. 

[Erik_L]

To be clear: the data is correct in the table. It's also correctly displayed if I click "edit visualization." It's only the re-run visualization in the notebook / dashboard that is empty when I update a widget to use a different value.

The code looks something like:

location = dbutils.widgets.get("locations")
ids = spark.read.table("production.gold.devices").select("id").where(f"location = '{location}'").distinct().toPandas()['id']
dbutils.widgets.dropdown('device_ids', values=ids, default='', label='Device IDs')

# ... Another cell

device_id = dbutils.widgets.get('device_ids')
spark.read.table("production.gold.data").where(f"location = '{location}' AND device_id = '{device_id}' AND create_time > now() - INTERVAL 30 DAYS")

# PowerBI Linear graph visualization