Forum Discussion
slicers
If I understand this right, you need will need a categories table which has a unique list of categories and another mapping table which lists the categories that a provider belongs. Basically 1:M relationship between Categories and Mapping table and 1:M relationship between Providers and Mapping table. This should solve your problem.
Thanks
Check out Chicagoland Power BI User Group
OK, I will check that out. I already have the 1:M between Providers and Mappings, and now I will create a M:1 between Mappings and Cohort Names and see how the slicer works etc. Thank you...will report back!
Tom
Hello again,
I changed things around as I understood the suggestion to create an intermediate mapping file.
Here's the mapping file--with 17 records--7 for the Tot IP Beds categories (like Tot IP Beds 500-750)--7 for the Tot Routine Bed categories, and 3 for the Hosp Type categories. Each provider is in each category SO...each Provider has 3 records.
The net result seems to be that when I put a slicer out there with the 17 Name_Group categories and select multiple Name_Groups--I get multiple provider records. So, that's a bummer.
RelationshipsThe Match File
I've highlighted the relationship fields, show the match file, and finally the results for boston--selecting GenAcute and a few sizes--you see providers in multiple columns--one for each category they're in.
Any ideas?
I think I'll either want to create new columns in the main provider file--one for each cohort. So eventually I'll add 100 columns to 5000 records. OR perhaps I could add 100 tables and relate them to the main provider file--one for each cohort and each provider will only have one record in each. What do you think?
Either way, the slicers will be AND slicers so the results will be one line for each provider that matches all of them. Sound right?
Thanks up front for bearing with me on this.
Tom
Which of these two approaches seems best?
