SAP Visual Intelligence at SAP TechEd

I had the opportunity once again to attend SAP TechEd in Last Vegas. As an SAP Mentor, I also volunteered to be an “SAP InnoJam Guru” for the InnoJam event last weekend. I’ve competed before, and may compete again. But this year, I wanted to take it easy and help those that wanted to use the SAP BusinessObjects tools in their projects.

During the event, laptops all over the room are loaded up by SAP with the latest and greatest versions of products that all connect up to SAP HANA, SAP BusinessObjects, and SAP Data Services, to name a few. To my delight, I had about 36 hours that I could sit down and play with some new technology. Namely, SAP Visual Intelligence (SAP Visi). I say…cool.

My first experience with SAP Visual Intelligence is that it is so far, an intuitive tool. I didn’t read a book. I didn’t read a manual. I didn’t watch a webinar. I did spend enough time with it that I asked a friend (thanks Josh) a few questions, but it was intuitive enough that I went from raw data to data visualization within an hour.

The Data

I did a brief search for free data sets and stumbled upon explore.data.gov. There are easy filters to help you get free and public data sets and play with data. The first thing to probably mention here is you probably aren’t going to find gigantic data sets here. I was more interested in capability of visualization vs. sucking in 3 million rows of data or something. I settled initially on a data set that displays the location, timing, and magnitude of earthquakes in the last 7 days in the US. With that, in that short hour, I became a #datageek.

blog011201

At the time, my first inclination was “WHAT ON EARTH AM I DOING ON THE WEST COAST IT IS ABOUT TO FALL INTO THE PACIFIC OCEAN?”

After a few failed attempts at lame data sets, I found this earthquake data, set a measure to count the occurrences of earthquakes, and defined the geography of city/state as my geographic dimension, and I was ready to go. I barely sanitized some capitalization issues in the data, but made node of some really nifty data cleansing capabilities including a Find/Replace, the ability to split columns in concatenated fields, the ability to trim characters, and more. In addition, you can replicate columns of data and apply some limited formulas to the data. It would have been nice to see a formula editor similar to Webi, but I’m not going to get hung up here.

Moving Beyond InnoJam

Back home with my new SAP TechEd Collateral thumb drive in hand, I installed SAP Visi on a Windows 7 64-bit VM on my Mac. I did bump it to 2 virtual CPUs and 4 GB of ram so I could have some runway to play. So far, even with a data set of a few megabytes, it sings. With my own VM and installation of SAP Visi in hand, it was time to explore a data set that was a bit more rich. Returning to explore.data.gov, I kept the disaster-theme going and grabbed year over year details on FEMA Disaster Declarations. Yikes.

Getting the data acquired into the tool is a no-brainer. Import the .xlsx file, select the proper worksheet, grumble that it did not recognize the headings correct, but move on. Next up, I created time hierarchies on my time dimensions by simply right-clicking and choosing Create a Time Hierarchy. Nifty.

blog011202

In addition to that, you can imply hierarchies in a similar manner for geography. I wished the grain of this data would have actually gone to a city level, but instead was at an aggregate State level. That’s ok. We can work with this too.

blog011203

The only thing remaining for me here again was to create a measure. I chose the high road for this demonstration and again, right-clicked on an attribute and asked SAP Visi to Create a Measure. In seconds, I have a “sum” of Disaster Numbers. This isn’t particularly helpful, so changing it to a distinct count of disasters I can better display my data.

blog011204

Seriously, looking at the escalation of the number of FEMA declared disasters by time and disaster type attribute makes it look like we like fire or something.

I didn’t seek to educate you on SAP Visi here. What I did try to do is show you that in minutes you can have data pulled in, sanitized as appropriate, and looking kind of awesome.

A few notes before I tidy this up:

  • I know 1.0.6 just came out and includes some predictive capability. This blog was written with 1.0.5 that was included on the SAP TechEd Collateral media.
  • I want to emphasize that I think SAP needs to rethink the approach to mobilizing this tool. It’d be a disservice if a customer had to own Explorer in order to show this stuff off on an iPad when done.
  • I’d love to see the user interface tightened up. I saw a few occurrences during my play time on the machine at InnoJam in which the formula box didn’t play well with the layout when it was done “Doing it”.
  • The Manipulation Tools are great and I hope this library gets beefier.
  • I’ve heard there is about a 3 to 3.5 million row limit on the desktop version. I’d love to hear from SAP what the practical limit is based on memory size because after all, it is a 64-bit app, right?
  • UPDATE: I would really love to see some public SAP HANA data sets on Amazon AWS to let developers give SAP Visi a run for its money. Thanks to Nic Smith for pointing to a page on the SCN that shows some public data sets ready to go!
  • UPDATE: Word on the street is 32-bit SAP Visi is on its way. | Lastly, will there be any 32-bit love in the future for SAP Visi? I hope so. Enterprises are still just a bit behind in their 64-bit desktop strategy I think.

2 thoughts on “SAP Visual Intelligence at SAP TechEd

  1. Hi Eric,
    Great suggestions; I know some of them are already circulating internally!
    On SCN, we have created an Idea Place for SAP Visual Intelligence where people can submit product/feature ideas and vote on them
    https://cw.sdn.sap.com/cw/community/ideas/businessanalytics/sap\_visual\_intelligence.

    For the current version 1.0.6, the max number of cells is 15 millions and you could see this limit moving way up in upcoming versions. Cheers!

Leave a Reply