This winter, staff from the Center for GIS and RESI took on a data visualization project for the Johns Hopkins University and the associated anchor institution strategy entitled Homewood Community Partners Initiative (HCPI). The partnership is built around stakeholder engagement with the vision of a “vibrant urban center,” a “livable community,” and ongoing community projects in ten neighborhoods and one commercial district surrounding the Johns Hopkins Homewood campus.

HCPI3The TU team selected Tableau business intelligence software for the project. Tableau is focused on “making databases and spreadsheets understandable” and allows for the creation of charts, tables, and maps that are dynamic and interactive. For this effort, HCPI wanted to create themed dashboards that reflect the conditions, trends, and progress toward goals in several focus areas: public education, commercial development, the housing market, quality of life, and JHU community influence. HCPI plans to use these dashboards to share information with their partners and as an aid in planning and decision-making.

To build these dashboards, our team collected and processed data from a variety of sources such as the American Community Survey (Census), Baltimore City Department of Planning, Open Baltimore, the Maryland State Department of Education, and several HCPI partner organizations. Although the dashboards feature far more charts than maps, GIS played an integral role in the project. In many cases, we obtained tabular data for Baltimore City or Maryland and then geocoded the records to determine which fell inside our specific interest area, or to make calculations by specific neighborhood. Using GIS capabilities in conjunction with Tableau allowed a great degree of flexibility and customization in the data presentation.

A few reflections on using Tableau for this project:

  • The visualizations are dynamic in the sense that as more years of data are added to the underlying spreadsheets, data points and totals will change, and the axes will expand.
  • We can maximize use of space by building charts that contain selectors that allow the user to switch between years, or other aspects of the data, or both.
  • While Tableau’s mapping capabilities are somewhat limited as compared to other mapping-specific software, we were able to build focused, interactive maps into the dashboards. These allow users to zoom in and out and identify on a data layer.
  • Every visualization—maps included—has built-in tooltips so users can hover over areas and see numbers and information from the spreadsheets.
  • Creating the dashboards has been an iterative process of exploring the data and experimenting to determine which techniques best illustrate trends in data, or best fit a particular metric.
  • While experimenting with color schemes and switching among the template chart types is quick in Tableau, adding final touches to produce a finished look can be time-consuming.

While the HCPI dashboard is intended to be internal, we are excited to share a few screenshots of the work we’ve done to help the partnership track and evaluate progress toward their goals.