Forum Discussion

cheryl0316's avatar
cheryl0316
Helper II
1 year ago
Solved

Create pie chart based on measure

This is a sample report regarding employee performance.   If an employee doesn't meet the sales target, he will receive a notice. If an employee exceeds the maximum allowed refund amount, he will...
Need Help
  • lbendlin's avatar
    lbendlin
    1 year ago

    Power BI  (unlike Tableau) does not support dynamic bucketing.  You need to bring your own buckets, in a disconnected table - like you did in the "Notice"  table. You can then base your horrible pie chart on that table.

     

     

lbendlin's avatar
lbendlin
Super User
1 year ago

You can change your buckets to represent notice types. (Sales or Refund)  Then adjust your measure accordingly

 

Are you sure a pie chart is the right chart for the data story you are trying to tell?

cheryl0316's avatar
cheryl0316
Helper II
1 year ago

Yes I know that's a bit weird

I want to create a pie chart in PowerBI to show the distribution of each notice type. This is an example created in Excel

 

I want to calculate the number of each notice type based on each country, city, employee, and product category.

The number of highlighted cells is same as the values shown for each notice type in the pie chart above

 

My current thought:

1. Create a new table with a column called "Notice Type," which will be used as the Legend in the pie chart.

2. Create a measure

NoticeTypeCount =
VAR Sales_Notice = IF([Sum_Sales Amount] < 100, 1, 0)
VAR Refund_Notice = IF([Sum_Refund Amount] > 10, 1, 0)
VAR a = SUMMARIZECOLUMNS(dimLocation[Country],dimLocation[City],dimEmployee[Employee Name],dimProduct[Product Category])
VAR b = ADDCOLUMNS(a, "Sales notice", Sales_Notice, "Refund notice", Refund_Notice, "Revenue amount", [Sum_Sales Amount])
return countrows(filter(b,[Revenue amount]>0))
 
I think var b is wrong because "Sales notice" and "Refund notice" should be grouped under a column (?)
I’m a bit confused when it comes to summarizing tables or columns. Any advice would be appreciated. Thank you for your help!