Among the many reasons Tableau has remained a leader within Gartner’s Magic Quadrant for Business Intelligence and Analytics Platforms is that it continues to empower analysts without expert knowledge of reporting tools to create a wide variety of meaningful and clear data visualizations. Tableau’s mapping capabilities in particular are a prime example of how a potentially complicated report (e.g. mapping the number of emergency rooms per ZIP code in Chicago) is now at the fingertips of analysts and report creators.
Versions of Tableau 8.2 and later contain a number of very useful built-in geographic roles, where Tableau will automatically recognize and display geographic data and associated metrics at different geographic granularities: cities, addresses, area codes, congressional districts, counties, states, countries, ZIP codes, and others.
Therefore, with just a few steps analysts can create colorful, dynamic geographic visualizations to highlight and explore geospatial patterns in their data that will help answer questions for stakeholders: Is the ROI on advertising campaigns better in certain counties in a metropolitan area, and across the country? Are there clusters of high rates of hospital readmission among different ZIP codes in a given metropolitan area?
Having these built-in administrative boundaries makes map creation easy and allows analysts to focus more on telling a story with their visualization. However, what if an analyst wants to create a map using boundaries not pre-built into Tableau’s mapping capabilities, for example, hospital catchment areas, electric power markets, or local administrative units such as Chicago wards? With the ability to map custom geographies such as these in Tableau, analysts would have much greater flexibility in visualizations and would be able to tailor their maps more specifically to the needs of the client.
With no native functionality in Tableau to work with geographic boundaries outside of standard geographic boundaries such as ZIP codes, counties, states, etc., users can turn to Open-source geographic information systems (GIS) to manage shapefiles (geographic data storage files) and manipulate them into a format reproducible in a Tableau visualization.
A shapefile is a very common geospatial vector data format used by GIS software packages. Storing information on attributes, topology, and projection, a shapefile contains geographic information about points, lines, and polygons and is used in a GIS to project data onto a map visualization and conduct spatial analyses. Shapefiles can store and display geographic data of any boundary type – for example, medically underserved areas, or national forest boundaries—thus opening the possibilities for mapping of any boundary in Tableau.
While clients may be able to offer their own proprietary shapefiles to analysts for use in creating maps, many shapefiles also are publicly available for free download, making it easy to procure a boundary file specific to your needs. Many cities and municipalities provide a wealth of shapefiles, for example, the City of Chicago Data Portal. Another popular source for a variety of administrative boundaries is the US Census TIGER/Line Shapefiles, offering shapefiles ranging from census tracts, to congressional districts, to metropolitan zones. These are just a few of the common places to freely download shapefiles, though there are many more.
Having procured the proper shapefile, analysts can then work with Open-source GIS programs to manipulate the shapefile in a way that ultimately becomes readable for Tableau. One recommended Open-source software is Quantum GIS (QGIS). QGIS is free for download and has become one of the more popular Open-source GIS software applications. QGIS is often a preferred GIS because of its customizability with many user-created plugins and because it is more user-friendly than some other GIS applications.
QGIS provides a wealth of user-written spatial data management and analysis functions, some of which can transform shapefiles to Tableau-ready data files. As a result, while Tableau cannot natively read the boundaries of Chicago wards, for example, it can, however, read a sequence of points that correspond to the nodes of each ward boundary as generated by QGIS through the ‘Extract Nodes’ tool.
After plugging in these points as the data source in Tableau, either as .xlsx or .csv, an analyst can then draw the ward boundaries that previously could not have been drawn in the software. The examples below show a simple visualization of Chicago wards without any accompanying metric (left) and a visualization of median home value per ward (right).
With the ability to draw a choropleth or filled map with wards, boundaries not natively recognized by Tableau, an analyst could now report more accurately on the geographic distribution of median home values, pointing to higher values on the north side and along Lake Michigan, with lower values on the far south side of the city.
While analysts can already create very powerful and telling geographic reports with Tableau, incorporating geographic shapefiles and open-source GIS software into the process can greatly enhance visualization of geographic boundaries in Tableau and help to uncover geospatial trends in your data. With the flexibility to visualize geographic boundaries of any type and granularity analysts can tell a more compelling story to the client and tailor the report to the geographic boundaries most relevant to each business.
To read a step by step tutorial visit my whitepaper on our website here.