Wannabe Awesome Me
  • Viz 'Em
    • Tiny Tableau Tips
  • Mentor Match
    • Mentor Match Blog
  • A Pop of Color

Making the MS Viz

4/24/2014

0 Comments

 
I wrote a non-viz blog post (I know, I'm still shocked!) about why MS matters to me and why I thought it would be a good idea to do a MS viz. Way easier said than done. But I stumbled upon the Atlas of MS database and I thought 'This is great! I'll get all of the data for 2008 and 2013, it will come down in excel, and I'll create a Viz of the Day worthy viz.' Oh and while I'm at it, a genie will appear and grant me three wishes (including unlimited wishes). No genie appeared. Major frownie face. So I had to resort to trying to do this myself. After I downloaded some data and tried to work with it and then banging my head against my chair, I looked to my network, keeping in mind what Jonathan Drummey once wrote; we're not experts in every area. So instead of hanging my head in shame, I asked Matt Francis to help organize it in a manner that would be tableau friendly and then help me understand how I could have done it myself. Some data makes me happy. Other times, it's a four letter word that makes me channel my inner Ralphie (who says the mother of all curse words, the F--- word). Matt and I discussed the data side of things, which you can listen to here (look at me being all multimedia).

So you heard (or spoiler alert), the data was horrid...even for a seasoned tableau stud like Matt (thanks for that little gem, Nelson). So, lesson learned for me: I need to plan to spend 50% of my time getting up close and personal with the data (coffee or wine may facilitate this). Towards the end, Matt used Alteryx to blend the data together to make it useful. Here's a look at what he did.
Picture
Data magic via Alteryx
Now we get to the part of the viz that makes me happy; the visual. I think it's important to understand my original objective. I wanted this viz to help raise awareness about MS and as part of that, I thought it would be valuable to have a visualization that could be printed out that shows the comparison for the two years (2008 vs 2013). It was also important for me to use the MS colors; orange being the primary color. This led me to use gray and orange theme. For some reason, I didn't want to go with a black background, I felt gray worked better. While there was so much data, I had a tendency to want to include as much as I could (more information is better, right??? uh, maybe not). While I used Matt for data help, he also provided my first level of review (a concept he's talked about in his blog post about rubber ducks). I also sought out feedback from three other folks; Kelly Martin, Anya A'Hearn, and Ramon Martinez. I felt uncomfortably comfortable asking them to review the viz. All three provided excellent constructive criticism. For example, I underlined some text, totally spacing on the fact that it would then make people think it was actionable (if not printed out). Anya dished out some tough love that made me re-think what I presented. That was super difficult to hear, but after taking the afternoon to reflect, I applied a saying I love, shown below. I looked at it as a valuable learning experience.
Picture

Ramon provided some feedback on showing the data on the map as well as the bar graph to show the change from the two years. As a result of the feedback received, I reassessed what I wanted to show in the viz and tweaked how I wanted to show it. Essentially I scrapped the concept of being able to print out the visualization. When it came down to it, I thought the relevant point was still being shown and if folks wanted to print it out, they could. I reframed the viz so that I had three major sections; How prevalent is MS?; When are people diagnosed?; and What prescriptions are they using to manage their symptoms? I scaled back what I was showing, applying the less is more concept. I used Ramon's suggestion of the showing the difference on the map, for the bar chart to the entire view next to the map to show the range of values, showing the mean age on onset for 2008 and 2013, and then ranking the positions of prescriptions for 2008 and 2013 using a slope graph.

Learn About Tableau
I think the end result is much better and while I've had moments of wishing I could have done this all on my own, I remind myself that there are very few people that are experts in everything and (this is probably the hardest part for me), I should not expect myself to be as good as people who have years of practice or studied visualization. However, since I can count on one hand the visualizations I've done, it's not too shabby. And more importantly, I learned something from the experience (least that's what I keep telling myself) and if one person views the viz and learns something new or supports the cause to provide support for MS, then I'll be happy.  

Though I mentioned them in the post, I would be remiss if I did not extend a big thanks to Matt, Kelly, Anya, and Ramon for taking their time to review and provide support me in developing a visualization that is extremely significant to me.

0 Comments

Analytic Tools for the Data Warrior Princess

4/23/2014

0 Comments

 

I kind of remember Alteryx from TCC13, but to me, they were just another vendor (sorry, y'all). It wasn't until this year, when either I got smarter, they're doing a heck of a job with marketing, or a combination of both (which is probably the case), that I kind of understood what this Alteryx was about.  So when I saw that they had an event nearby called Analytic Tools for the Data Warrior, I knew I had to attend.  I rounded up an analyst at work who likes this kind of thing, and we headed over to Bethesda.  First, the Tableau rep, Pat, nailed it when he said that events in DC are tough, that it seems like people can't escape their offices.  Which is pretty much why I was sitting front and center; a result of arriving five minutes before the event was supposed to begin (but in fact, had just started.  Side note, they must prescribe to the Tom Coughlin rule of meetings...If you're five minutes early, you're late!). Anyway, the Alteryx dude gave a brief background on Alteryx, which has been around for over 10 years, with a name change (and mission change) in 2010.  I thought the demo was pretty interesting, as was the happy hour (data=happy). It was quite clear that they have a strong background in spatial analysis.  I think Alteryx is similar to tableau in the sense that you can see what you're doing as you go along.  I think the preview functionality is brilliant! Now, is the software completely intuitive and so easy that someone off of the street could pick it up and start going?  No, and that's not a bad thing. I think for data management tool, I think it's probably as intuitive as it can possibly be.   There are a few things that I saw from the demo that I thought were pretty cool.  

  • Previewing results (let's me make sure I'm on the right path).
  • Cool spatial analytic tools. Like drive-time!! I actually thought of a use case at work for this.
  • No scripting for R. I know very little about R, except that you can do some awesome predictive analytics with it. Now, why is this bullet point worthy? Because you don't actually have to know how to code in R...they have wrappers to help people like me (or data management people, at least). 
  • I also like that it can create an extract that will feed tableau (yay! less work for me!) and
  • Drag and drop workflow

After Alteryx did their portion of the demo, it was time for tableau to show their stuff. Based on raised hands in the crowd, about 2/3rd of attendees had used Tableau.  That's pretty good!  The Tableau representative took the extract and re-built the sample dashboard and stopped there, which kind of annoyed me, especially as there was time remaining in his portion of the event.  The question being answered was 'Who Do We Send A Second Mailing To?' While his primary objective was to rebuild a dashboard with four views on it, since he had time, he could have made the dashboard more meaningful by doing a few tweaks to it, like putting a title on it and changing the color palettes (because I'm pretty sure he used orange twice). Thinking back, I'm not sure if the R stuff done in Alteryx was explained clearly when it came to showing it in Tableau. Maybe it was and I just have a bad memory.  Overall, I felt like there was so much more than could have been done to really show off Tableau, that I was just itching to go up there and do some tweaks.  But I didn't. Though I did quietly say something about reusing orange and the other analyst from work just chuckled at me. He knows that I don't like re-using colors, especially on views in the same dashboard.  I guess that means that I was interested in the tools being shown, since I wanted to fill the time with information about both.  

Next up was the customer story by Bernardo Roschke of Discovery CommunicatIons about how they use Alteryx and Tableau to track operations and how productive their equipment and people are. This was a great example of how both tools worked well together. The highlight for me was that Bernardo was able to take seven separate data sources and use the whiz bangery of alteryx to solve their data problem.

At the happy hour (which was really an extension of the presentations), I was able to ask my burning questions. How much does it cost and what about these duplicative offerings between the two tools? Alteryx tiers their licenses and natch, a server license costs more. I think it might be a matter of perspective. While it seems expensive (to me, though I haven't researched the costs of data prep tools), it certainly costs less than waiting for the work to be done by others (like the cost of one FTE). If the data/report/dashboard is not timely, then it's not relevant. Next up...what's the deal with redundancy. When is it appropriate to use the visualization and R features in the respective tools? From what I gathered, visualization in alteryx seems like a side dish to the main meal. As for R, the benefit of using alteryx for this is that you don't need to know how to code in R, as you would in tableau. Non R coders Rjoice!

Overall, this was a nice little event and I was glad I was able to get out of DC to go. I expanded my network and learned more about a data prep tool, which gets me thinking about how we might be able to use it at work. We even got a free copy of their booklet Predictive Analytics for Dummies. The wheels are a-turnin', so I'm counting this as a productive outing.

Picture
0 Comments
    Follow my blog with Bloglovin

    RSS Feed

    What's This About?

    After using Tableau in my last position, I'm now a data viz hobbyist and this is my little slice of world dedicated to data visualization...typically Tableau related. 

    Vizzed and Confused? 
    Here are some great resources!

    Anya A'Hearn 
    Andy Cotgreave 
    Kelly Martin
    Andy Kriebel
    Matt Francis
    Jonathan Drummey
    Jewel Loree
    Peter Gilks
    Paul Banoub
    Dan Murray
    Ramon Martinez
    Ben Jones
    Allan Walker

    Tweets about "#dataviz"
    Tweets about "@tableau"

    Archives

    May 2018
    December 2017
    November 2017
    October 2017
    June 2017
    May 2017
    January 2017
    November 2016
    October 2016
    August 2016
    July 2016
    June 2016
    April 2016
    March 2016
    February 2016
    January 2016
    December 2015
    October 2015
    September 2015
    August 2015
    July 2015
    June 2015
    May 2015
    April 2015
    March 2015
    February 2015
    January 2015
    December 2014
    November 2014
    October 2014
    September 2014
    August 2014
    July 2014
    June 2014
    May 2014
    April 2014
    March 2014
    February 2014
    January 2014
    December 2013
    October 2013
    September 2013

Powered by Create your own unique website with customizable templates.