4/11/2014
Dynamic Geo-Optimization in Tableau Using Integration with R - a work in progress
The video below demonstrates a work in progress. It's a dynamic geo-optimization in Tableau using integration with R. There is a post on r-bloggers here that outlines how this works as well as sample code that I modified and built on top off for this project. The top map in this visualization plots a quantity by location. Example, number of widgets in each factory. The algorithm uses the quantity as a weight in the calculation and calculates the optimal center location based on those quantities and their location. Therefore, larger circles (larger quantity) on the map will have more influence on where the optimal location will be.
There are a number of real-world applications for this technique. For example, finding the ideal location of a distribution center or identifying the location for a new store based on customer location. One recent example was a final project in our data visualization class at the University of Cincinnati. The team gathered data for a local fitness club that was moving locations. This is another great application for this technique, plotting the customers of the fitness club by address to determine where the ideal location in the city should be. Weights could be set based on customer spend or the number of trips each customer makes to the fitness club per month. These are just a few examples of the many application of this.
Here is a short video to demonstrate this functionality. The entire data set is plotted and the geo-optimized center is calculated and placed on the second map. When points are selected on the top map it recalculated in R for a new center point based on the selection. By using Tableau as the user interface it allows for this dynamic analysis, recalculating real-time as the user selects the points to consider in the analysis. This visualizaiton is still a work in progress and I normally wouldn't post something that isn't finished, but I thought I would share this as a video to spark other ideas on using Tableau and R integration.
I hope you find this information useful. If you have any questions feel free to email me at Jeff@DataPlusScience.com
