Introduction

For more than 20 years, I did a lot of groundwater flow and transport numerical modeling.  This work prepared me well for my current career in business analytics, but life sure would have been easier if I had Tableau Desktop back then!  

If you like this article and would like to see more of what I write, please subscribe to my blog by taking 5 seconds to enter your email address below. It is free and it motivates me to continue writing, so thanks!

A Look Back to My Earlier Career

I just came across a Powerpoint from 2003 (click here for models_at_mmr) that has some nice examples of visualizations and comparisons of measured and simulated data (Figure 1 and 2) from over 15 years ago. This file gave me a chance to look back at my Massachusetts Military Reservation (MMR) groundwater modeling work from the late 1990’s to early 2000’s.

What we used to call model calibration and verification is now called model training and testing in the world of statistical business analytics. Instead of running finite-element and finite-difference models, I now run regression models and perform visual and statistical analytics on data from a wide variety of businesses.

Figure 1 – Simulated vs Observed Contaminant Mass Removal

Figure 2 – Simulated vs Measured Contaminant Mass Removal

Although the simulations and software tools are different and the computers are now much more capable, the modeling  goals remain the same.  In any type of modeling, we attempt to simulate natural or human-impacted processes so that we can predict future system behavior.

In many ways, running deterministic or stochastic groundwater models in four dimensions (x,y,z, and time) is not much different than trying to predict which marketing campaigns to be conducted in a retail business will lead to better sales. What I have come to understand and appreciate is that (1) all models are approximations of reality, (2) data is how we measure responses and quantify the processes, and (3) all predictions are intrinsically inaccurate.

Our goal in simulating a system is to minimize the inaccuracies in our predictions. When we can do this successfully, there is a lot of fun to be had. I now have this understanding because I have had a lot of time to learn about data, how to use it, manage it, analyze it, and to incorporate it into models that predict future system behaviors.

Tableau, What Took So Long?

One of my biggest regrets is that I had to work the first 20 years of my career without Tableau desktop software.  Looking back, I can see how so many computer codes I designed and wrote, or software teams that I directed, were necessary because we didn’t have a tool like Tableau to help us perform our analysis.

There were capable tools we used like TecpPlot and Autocad for visualizations, but much of the quantitative analysis had to be programmed on a case-by-case basis.  If I had Tableau throughout my career, things would have been much easier, more insights would have been possible, and better models could have been built.

Now that Tableau is here, I’d like to take another crack at analyzing some of my earlier work. I think that this is one of the reasons why I appreciate Tableau to the degree that I do.  When you try to do something yourself in a computer program and you realize the complexity of the task, having a tremendously capable tool like Tableau really opens up your world to unlimited possibilities.  For this, I thank the founders of Tableau on a daily basis.