by Mark Kromer
It does seem these days that anything BI related coming out of Redmond is called Pivot something. Keeping with that trend, Microsoft Live Labs has made available their Silverlight 4 based control “PivotViewer”. PivotViewer is generalized control meant to be used in Web applications that allows users to consume very large data sets of heterogeneous types & sources. Very similar to the concepts used in BI dashboards, PivotViewer enables slicing & dicing and does so in a very visually compelling and interesting way. And that is exactly why we are going to talk a bit about Pivot in this post as part of our series on creating exciting Microsoft BI dashboards.
The best place to get started with Pivot is here at the Microsoft Live Labs site for Pivot. You will need Silverlight 4 installed and then download the PivotViewer control here. What I’ll discuss in this brief blog are the concepts to get you started. Keep in mind that PivotViewer is going to generate visualizations by taking images and combining those with metadata.
Collections are at the heart of Pivot and the PivotViewer will use this as the data source. These can be built from many sources including command-line tools, application code and Excel. Just to show you a simple mechanism to load up a Pivot collection, I’m using the add-in for Pivot in Excel. Below is a screenshot of a very simple collection that I made to show you what a collection looks like. For a BI dashboard, you will want to connect to a data source much larger than this sample data set. I am using photos from Bing to ensure that I am not stealing copyrighted pics. To try to put this in context of a BI business app, I added a region field to the collection so that you can think of this collection as a way to sort, say, store locations or offices or buildings, etc. and then have facts associated with those dimensions.
Now that I have my collection published in CXML format, I can consume it in an application such as a Silverlight app. The Silverlight app will not need anything added other than the PivotViewer control to test your collection out. Just include a reference to the assemblies for System.Windows.Pivot. You will also need the application to have access to your image files. But when you’ve point PivotViewer to your source of collection data, you’ll get an interactive, working Silverlight control that will look something like this Microsoft demo screenshot that has a lot more data for a much cooler demo than my simple Excel-based Pivot walk-thru. If you watch the demo from http://www.microsoft.com/silverlight/pivotviewer/, you’ll see that this control will work great for allowing users to interact with vast amounts of data to filter and refine interactively.
Next steps are to download PivotViewer and start playing with the collections. In our next couple blogs, we’ll talk about strategies to incorporate this control as well as others in a coordinated dashboards and how to add mobile displays for Microsoft BI as an option for your users.