Forum Discussion
Break out Multi-Select field into Slicer
Hello GabeFig,
Thanks for sharing the sample data.
I have reproduced your scenario using the sample data you shared and implemented the required filtering logic based on multi-valued Categories and Features columns (with delimiters like ;).
Output:- When “Application” is selected from the Categories slicer:
You will see only rows where “Application” is one of the multiple categories. - When “AI-Enabled” is selected from the Features slicer:
You will see only rows where “AI-Enabled” is one of the feature values.
For your reference, I’m attaching the .pbix file.
Best regards,
Ganesh Singamshetty- When “Application” is selected from the Categories slicer:
Thanks Ganesh for the more detailed instruction BUT that seems to be what I did.
I could not tell there was anything different even after the original slicer was removed and recreated w/ Category field that was split (by rows).
Below are some screenshots of what was done with the existing MS BI.I wonder that this needs to be done completely from scratch to work.
Both Catefory & Feature were done. The split was done by Row.
I also checked to see if there was a hidden CR or LF or CRLF but there were none.
Community SupportHello GabeFig,
Thanks for sharing your detailed screenshots and you’ve done a great job applying the split correctly. I now suspect the issue is not with the split step itself, but with how Power BI handles row-level filtering from slicers in a 1-to-many relationship scenario.
To make the slicers work properly without affecting or duplicating your main data table:
- Duplicate your main table in Power Query call it CategoryMapping and Keep only ID and Categories columns.
- Apply Split Column by Delimiter Into Rows on Categories and Rename the result column to Category. Repeat same steps to create a FeatureMapping table.
- Create Relationships between CategoryMapping[ID] >> MainTable[ID] and FeatureMapping[ID]>> MainTable[ID]
- Then Use CategoryMapping[Category] in the Category slicer and FeatureMapping[Feature] in the Feature slicer
If the issue persists, try to share your .pbix file or sample data so I can better understand the scenario and give you the necessary solution.
Thank you.
- v-ssriganesh
Hi GabeFig,
Thanks again for your detailed follow-up and for sharing the screenshot.What you're seeing with the "Business" category behaving differently is due to how Power BI handles split-to-rows transformations on delimited text. When a category like "Application; Business" is split, Power BI creates separate rows one with "Application", and another with "Business". However, "Business" also appears as a standalone value in other rows, which may result in multiple "Business" values in your slicer or inconsistent filtering behavior.
To make sure the "Business" category filters all relevant rows whether it appears alone or with others. We suggest the following steps:
Duplicate your main table and name it CategoryMapping.
- Keep only ID and Categories columns and Split Column by Delimiter >> choose ; and split into rows.
- Trim whitespace from the split column to clean up any extra spaces and rename the new column to Category then Remove duplicates so that "Business" appears only once per ID.
- Create a relationship between CategoryMapping[ID] to MainTable[ID] and Use CategoryMapping[Category] in your slicer.
If the issue persists try to share your sample data: How to provide sample data in the Power BI Forum - Microsoft Fabric Community
Best Regards,
Ganesh singamshetty.The CategoryMapping & FeatureMapping look great but again if 'application' is selected it will show only those that have just 'application'.
If the item has 'application; cloud' then that will not show.
If 'application' & 'cloud' are chosen then only those that have either 'application' or 'cloud' but not those that have 'application; cloud'.
The problem seems to be the Manage Relationship' I've tried several if not all variations.
It seems best to use the Categories field cause if the Records ID# is used it only returns (1) as w/ the Feature