A summary of the new features in the October 2018 update of Power BI Desktop
Part four of an eight-part series of blogs

Not an earth-shaking update, but lots of nice new features nonetheless to make your life easier, including a better DAX editor, the ability to search for text when filtering a visual and a clever new connector which guesses what you want to import from a website.

  1. Changes introduced in the October 2018 Power BI update
  2. Changes to the DAX editor
  3. Filters are now searchable
  4. Explaining why an increase or decrease has happened (this blog)
  5. Controlling exporting of data
  6. The Web by Example connector
  7. Other changes in the October 2018 update
  8. Features awaiting in preview as of October 2018

For a cumulative list of all of the updates to Power BI Desktop in the last few year or two, see this blog, or have a look at the Power BI courses that we run.

Posted by Andy Brown on 19 October 2018

You need a minimum screen resolution of about 700 pixels width to see our blogs. This is because they contain diagrams and tables which would not be viewable easily on a mobile phone or small laptop. Please use a larger tablet, notebook or desktop computer, or change your screen resolution settings.

Explaining why an increase or decrease has happened

So you're looking at the following chart of sales by year, and wondering to yourself: I wonder why sales are higher in 2015?

Sales peak

Sales are clearly higher in 2015 ... but why?

You can now ask Power BI to tell you:

Explain the increase

Right-click on the relevant data point and ask Power BI to explain the increase (or decrease, if there is one).

Here's what you get for the above example:

Charts showing increase

Power BI has prepared four charts explaining the difference.

You can click on the icons at the bottom to see a different chart:

Regression chart

I really like this one - it shows Imogen is an outlier in the data.

You'll get different visuals according to how you're aggregating the data. The example above is summing; had you been averaging you'd get something like this:

Averaging help

If you're averaging sales by year, this is what you get initially.

What I don't get is how Power BI decided which field to analyse by for the above examples.  Here's the relevant bit of the underlying relationship diagram:

The relationship diagram

The purchase table is linked to two parents - centres and products - so why did Power BI decide to analyse sales by product, not centre? Answers on an email postcard please.

I'm not a huge fan of these attempts to apply artificial intelligence to Power BI; it's surely better that you understand the data yourself, isn't it?  But maybe I'm just a Luddite, standing in the way of the inexorable rise to supremacy of the machines that we've created ...

This blog has 0 threads Add post