Power BI scatter chart - tooltip showing incorrect values...

[Matty]

Hi Team,

I am seeing some really odd behaviour in Power BI whereby a tooltip within a scatter chart is showing incorrect data for certain data points compare with the same data displayed via a matrix.

Is there a known issue with this visual?  Or does Power BI interpret data differently depending on the visual used?  

Thanks,

Matty

[v-huizhn-msft]

Hi @Matty,

You need to understand how the scatter chart and Matrix display the date. Please check the calculation in different visuals.

For my sample table below, create measure Count = COUNTA(Table1[physicians]). Then I create two visuals deperately.

 
In martix, the rows and column level shows the row and column context in resource table, and the value shows what your expected number value categoried by rows level and column level. For example, 2 means there are 2 physicians(who statue is in) in category DRA.



In scatter chart visual, the X-axis and Y-axis show the number value(count of different categories), if the field is string type, it shows count of them automatically. Then legend shows the category. Please hover over the point, you will see the details, please check your date and visual carefully.

scatter chart  
Best Regards,
Angelia

[Matty]

Hi Angelia,

Thanks for taking the time to respond, but I don't think that's it.

Oddly enough, if I choose to exclude a single data point (any data point) from the chart, the chart recalculates and displays correctly.  But as soon as I choose to put the data I've removed back in (meaning all filtered data points are shown), it displays incorrectly again.

Give that the same data in an adjacent matrix remains correct throughout, it looks to me like a bug in the scatter chart visual.

Cheers,

Matty

[v-huizhn-msft]

Hi @Matty,

I am unable to reproduce your scenario, could you please share a screenshot for further analysis?

Best Regards,
Angelia

[andrewbrick]

I have this issue, too.  When adding tooltips to a scatter chart, some filtering doesn't seem to work correctly when data is not being summarized and both the X Axis and Y Axis buckets are populated with numeric values.

Unfortunately, this is difficult to describe, so I'll use a series of pictures of a simplified case.  Here's the table relationships:

t1 c1, t1 c2, t2 c1, and t2 c2 all contain numerical data.  Here's table1:

and table2:

Here's the visualization setup (emphasis on the scatter chart):

Note the "tt" field from table1 being used as the tooltip.  All fields in this scatter chart come from table1.  Here's what happens when I use the "Display/Don't Display" filter:

The scatter chart did not change, with the exception of the tooltip field not being displayed in the tooltip window for data points that should not be displayed (note the "tt" field missing in the popup below):

When the "tt" field is removed from the tooltip bucket of the scatter chart visualization, the following happens:

Now, only the properly-filtered values are shown in the scatter chart.

I've only been able to reproduce this when using unsummarized numeric values in both the X Axis and Y Axis buckets.  I also can't reproduce this behavior with other visual types.

It's possible I'm either not understanding some complexity in even this simplified model, or I'm getting this behavior because it's what the scatter chart is supposed to do.  However...

...if I change the Y Axis bucket to some sort of summarization, the issue disappears: the "Display/Don't Display" filter produces expected results.  It's only when both axes are not summarized that I see the buggy behavior.

[Femke]

Hi,

I have the excact same problem. Did you solve this one or have a workaround?

Thanks,

Femke