Visual level filter behavior differs

[dslatteryHCA]

A visual level filter is behaving differently depending on how it is introduced into the visual. If a numeric column is added to a table visual and the column is then filtered, the visible report data matching the filter is affected. This data is at an aggregate level. If the same column is dragged into the visual level filters and a value is specified, this value seems to be applied to the low level data detail data used in the aggregation.

Another difference is that in the first scenario, the only filtering option is advanced. In the second, a Filter Type drop down is exposed allowing Advanced, Basic or Top N filtering. This applies to both the desktop and service.

Is this the intended behavior? Can someone point to any documentation describing how these filtering methods differ?

Thanks.

[dslatteryHCA]

Thanks for the reply Cherry,. I will try to clarify.

When you add a column to a visual it automatically appears in the Filters section. You can also manually drag a column (the same or another) into the filters. What I am seeing is the behavior of these two differ. In the first case, the filtering is done at the aggregated level (sum, count, avg, etc.). If I could use a SQL reference, this is like using "Having". In the second case, the filtering is done at the underlying data level. A SQL "Where". Also the filtering options are different between the two as described in my previous post.

I assume that this is the intended behavior, but would like some verification.

Thanks again.

Dennis.

[v-piga-msft]

Hi @dslatteryHCA

I'm a little confused about your first scenario 'If a numeric column is added to a table visual and the column is then filtered ' does the column filtered by slicer or visual level filter? Could you describe this scenario in more details?

Normally, when you add a visual level filter, Power BI filters on the aggregated results. The default aggregation is Sum, but you can change the aggregation type.

Best  Regards,

Cherry