Forum Discussion

Anonymous's avatar
Anonymous
Not applicable
7 years ago
Solved

Query Return Different Result in Desktop than when Published to Workspace

I have a query to calculate sales score with the sales%(image 1), it is calculating correctly in desktop (image 2), but when I published the report to the workspace and check right away, the sales score does not follow the query logic(image 3). For example,  13.12%*30*5 =19.67, which is correct on the desktop(image 2) calculation, but returned 40.72 when published to the workspace. I download the report to check what's wrong after it gets published, the calculation is fine again. Can anyone let me know what is causing the issue? 

 

 

  • Anonymous's avatar
    Anonymous
    7 years ago

    Hi Qiuyun, thanks for the update. I will update my Power BI Desktop. But I find out the problem was actually with the a.sales % calculation, I used "/"  to do the division, so the calculation didn't work out when published to the service. And when using the Divide function to do the division, the issue resolved. Thank you for your help! 

7 Replies

  • v-qiuyu-msft's avatar
    v-qiuyu-msft
    Community Support

    Hi Anonymous,

     

    Is it possible to share pbix file if it's import mode? If it is, you can upload the pbix file to your Onedrive and send the share link to me. 

     

    Best Regards,
    Qiuyun Yu 

  • v-qiuyu-msft's avatar
    v-qiuyu-msft
    Community Support

    Hi Anonymous,

     

    I have tested with your pbix file but not reproduce the issue. If I open your pbix file in Power BI desktop, there is no selection in any slicer, publishing the report to Power BI service, the report display the same result as in Power BI desktop. 

     

    In your scenario, please ensure these the filter condition both in desktop and service are the same then compare the results. Also please browser cache then republish the report to see if the same issue occurs. 

     

    By the way, please cancel the OneDirve share if the report contains sensitive data. You can send share link to me via private message next time if you don't want to expose your data. 

     

    Best Regards,
    Qiuyun Yu 

  • Anonymous's avatar
    Anonymous
    Not applicable

    Hi Qiuyun,

     

    Thank you for looking into this. But I want to confirm that you actually compare the a.rank column when published to the Service, and compare with the marketplaces filter selected in the desktop version and the service version? The rank is ok when no filter is selected in both versions. But in the service version, when the marketplace filter is selected, the rank is missing the 1 and 2, and it has two 4s when the a.salesscore is different. I have other people tried to publish it and they have the same issue. I really appreciate your help! 

     

    Power BI Desktop Marketplaces filter Selected

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

    Power BI Service with Marketplace Filter Selected

     

     

  • v-qiuyu-msft's avatar
    v-qiuyu-msft
    Community Support

    Hi Anonymous,

     

    If you update the Power BI desktop to the latest version 2.65.5313.1381 64-bit (December 2018), then open the report, you will see the same result as on Power BI service. 

     

    After some research, the issue is related to measure a.salesscore, you might need to modify the rank measure. 

     

    Best Regards,
    Qiuyun Yu

  • Anonymous's avatar
    Anonymous
    Not applicable

    Hi Qiuyun, thanks for the update. I will update my Power BI Desktop. But I find out the problem was actually with the a.sales % calculation, I used "/"  to do the division, so the calculation didn't work out when published to the service. And when using the Divide function to do the division, the issue resolved. Thank you for your help! 

  • v-qiuyu-msft's avatar
    v-qiuyu-msft
    Community Support

    Hi Anonymous,

     

    Thanks for your update. As the issue is resolved. I will close this thread. :smileyhappy:

     

    Best Regards,
    Qiuyun Yu