Page Level Filter for Value Range

[wmorris20]

We have a report that displays a collection of visualizations and one of the measures is Year to Date Dollars.  The problem we face is that many of the measure values are near $0, which makes the visualization quite busy.  I'd like to be able to add a visualization to the report that essentially allows the user to filter out the outliers and focus in on the data that really matters.

I know this can be done using filter at visualization, page, and report level filtering.  However, I feel this is quite unituitive for the end user to have to click on the visualization, open the filter toolbar, and select the correct measure to filter.  Since I already know what they will likely want to filter, it would be excellent if I could present a filter that allows them to select a range of values.  I imagine I would need to turn off interactions to other visualizations, because context probably matters in this case.

[Phil_Seamark · Accepted solution]

Hi @wmorris20

Have you tried the new Numeric Range Filter that was released in the March update?

https://powerbi.microsoft.com/en-us/documentation/powerbi-desktop-slicer-numeric-range/

You will have to turn it on in your options and I don't think it is live in the service yet (won't be far away)

[wmorris20]

Thanks for pointing me in the right direction.  The update says that it doesn't yet work with Measures, but that functionality is on the way.

https://powerbi.microsoft.com/en-us/blog/power-bi-desktop-march-feature-summary/

[Phil_Seamark]

Hi @wmorris20

While it won't work on measures, if you drop a field onto it that covers the number range you want it might work?

eg, what field do you use on the axis of your visual?

[wmorris20]

That's a good idea.  Although it technically works, it doesn't work in a practical sense.  In this case, we would need to filter Sales Dollars which is then aggregated to YTD Dollars, which is want we want to display.  The problem with filtering Sales Dollars is that one segment might have 1 record of $10 but another segment might have 200,000 records of $10.  Filtering it above $10 will significantly limit the use of displaying the YTD.

If we filter at the Measure level, we can achieve removing the noise.

[Phil_Seamark]

Some other ideas (although again, not ideal) are

• hard code the start/finish axis to fixed values
• use a Log scale to pull back the outliers.