Forum Discussion

chonchar's avatar
chonchar
Helper V
3 years ago

Extract/Create New Table

Morning, All,

 

I have a sinlge table in my model called "Supplier Interests." One of the column names is "Interest Type." In this intrest type column, there are about 60 different values. Some are regions, some are lines of business, etc. 

 

The first column in my table is Supplier ID. There are 98k+ total rows and 11k distinct values for supplier ID. 

 

I want to create a new table that can be filtered in the visualization with filter context for "interest type." But I need to extract all of the regions and all of the interest types from this table. The goal is to create a filter that can identify Supplier ID -> Region-> Line of Business. But region and line of business are in the same column. I have been trying to do it with DAX, but I am not sure if it is possible or if we have to go to the backend (SQL Server, not PQE). 

 

Any insight is appreciated. 

 

Thank you, 

 

Cam

7 Replies

  • Anonymous's avatar
    Anonymous
    Not applicable

    Is there a distinction between Region and Line of Business?

    • chonchar's avatar
      chonchar
      Helper V

      Maybe this makes more sense: the three distinct entries in "interest type" should all be their own individual tables in the model with Supplier ID as the PK

    • chonchar's avatar
      chonchar
      Helper V

      In the Interest Type Column there are 3 distinct data points. Indusrty, Project, and Region. I want to separate all 3 of those into their own tables with Supplier ID being the Primary Key. Then add all three of those tables, filter context created, so be able to visualize: Supplier ID->Region-> and then the line of business as the final part. 

       

      I have tried creating a hierarchy, adding new tables, (may have the wrong DAX), and other filter context creating DAX. 

    • chonchar's avatar
      chonchar
      Helper V

      I have tried Calculatetable and group by. I really just want to create three new tables from the three unique data point in a single column, such that they have filter context. 

  • foodd's avatar
    foodd
    Community Champion

    Roche’s Maxim of Data Transformation states: Data should be transformed as far upstream as possible, and as far downstream as necessary.     It would be super helpful by adding a work-in-progress Power BI Desktop File (anonymized of course).