Using a Slicer to filter concatenated Column

[Jenni3hugs]

Hello! I am trying to use a slicer to filter a concatenated column and the measures I have found/edited from this fourm are not working. 

Esentially my data looks something like this:

Month	Item	Scope	Topic
Feb	Report A	Employees | Students | Programs	Finances | Human Resources | Enrollment
Jul	Report B	Employees	Human Resources
Sep	Report C	Students	Enrollment
Sep	Report D	Programs	Outcomes
Oct	Report E	Students | Programs	Enrollment
Oct	Report F	Programs | Employees	Finances

I would like to be able to use the slicer to view if the scope contains Students, programs or employees. 

I created a new table with just the scope values (Students, Programs, Employees) to use for the Slicer values. (named Scope)

I tried creating the following measures, (individually) but when I use the filter it doesn't show any of the concatenated rows. 

(In example: if I selected "students" on the slicer, only Report C would show. )

Measure1 = 

= IF(CONTAINSstring(Distinctcount(Table1[Scope]), "employees"), If(CONTAINSstring(Distinctcount(Table1[Scope]), "Students"), If(CONTAINSstring(Distinctcount(Table1[Scope]), "Programs"), 1, 0) ))

Measure 2  = 

 Calculate (Countrows ('Table1'), FILter ('Table1', SUMX ('Scope', Search('Scope'[Scope], 'Table1'[Scope],1,0))>0))

Any help on how to edit these measures (or create a new measure) to enable the slicer to work as I would like would be greatly appreciated. 

[v-cgao-msft · Accepted solution]

Hi @Jenni3hugs ,

We still need a new table for the filter. And don't create relationship with Table1.

Then create a measure and use it on the table visual filter, set the filter to show items when value is 1.

Measure = IF(CONTAINSSTRING(MAX('Table1'[Scope]),MAX('Scope'[Scope])),1,0)

The result should be like this.

Attach the PBIX file for reference. Hope it helps.

If this doesn't work for you, please consider sharing more details about it. And it would be great if there was a sample file without any sensitive information here.

It makes it easier to give you a solution.

1. Sample (dummy dataset) data as text, use the table tool in the editing bar.
2. Expected output from sample data.
3. Explanation in words of how to get from 1. to 2.

Best Regards,
Community Support Team_Gao

If there is any post helps, then please consider Accept it as the solution to help the other members find it more quickly. If I misunderstand your needs or you still have problems on it, please feel free to let us know. Thanks a lot!

How to get your questions answered quickly -- How to provide sample data

[johnt75]

I think your best bet would to use Power Query to manipulate the data so that you have one row per combination, so Report A would have rows for Employees / Finances, Employees / Human Resources etc.

Splitting the columns by delimiter and then turning the new columns into rows by pivoting or unpivoting ( I can never remember which is which ) should do the trick.

[gemcityzach]

Apologies to necro-threadjack - but I struggle with this option all of the time. I have datasets where numerous columns might be concatenated (label1;label2;label3) and I often need to develop analysis on the individual members of a concanentation and even as a funciton of another concatenated column (tag1;tag2;tag3). I.e., the unique combinations of of two columns. What I've been doing is as you suggested, creating a reference query off the main and splitting the reference by all of the fields I need to unroll label x tag as an example. And then I use those to display static calculations, percentages, totals etc on a summary page.

But I always struggle with - how do I give the user the ability to filter on one of those into the NON concatenated data source. I.e., how can I pass a TAG from the unconcatenated/expanded table into the original concatenated table so I can filter that table?

[Jenni3hugs]

Hi! Thanks for your reply, but unfortunately with the actual Dataset that would create way too many rows for the final product. (that would be 9 rows for report A alone) Additionally, the format it is in is the format my department wants to use for the final dashboard. That's why I'm hoping to use a measure.