3 Comments

  • Anonymous's avatar
    Anonymous
    Not applicable

    Hi Greg_Deckler ,

     

    We reproduced the issue and need to confirm if it's a bug or a functional change, I'll update here once there's any progress!

     

    Best regards.
    Community Support Team_Caitlyn

  • Experiencing the same issue. I do not have a legend field populated. Both my x and y axes are populated with numerical values. 

    Version: 2.126.1261.0 64-bit (February 2024)
    User ID: 1620cf17-aa53-4bdc-b132-c46079f0ef06
    Session ID: 31a0d3cb-8faf-4925-b220-7ffb488b4e37

     

  • Going to reply to my own comment 😅

    I found a work around by building my scatter plot as a table - the 'automatically detect clusters' feature was still available in that visual. 

    My values are being pulled from a calculated table which houses all my linear regressions calculated columns and measures within scope. 

    LR Table = SUMMARIZE(
        'vw_PropertyDesignations',
        'vw_PropertyDesignations'[PropertyKey],
        "Yvalue", [ServiceRequests_AVG_Days_To_Complete_This_Month],
        "XValue", AVERAGE('vw_PropertyDesignations'[UnitCount])
    )

    Interestingly, after re-avergaing my "XValue" field, the cluster option worked, and I was able to convert back to a Scatter Plot. Not necessarily ideal, and I'm still relatively new to Power BI (10 Months) so I'm probably doing something egregious as well .... ðŸ˜¬

    Nonetheless thought I would chime in.