From Data Sets to Data Visualization: Tanzania Dashboards

September 19, 2016
Diego Andres Dimunzio, Mark Irura
Open Data, Tech Stack

In an earlier post, we shared the context and intended purpose of the Tanzania Sectoral Dashboards; in this post, we dive into the technology behind the data visualizations.

Seeking to put the Principles for Digital Development into practice, we built the Tanzania Dashboards using open source tools and frameworks, which reduces the total cost of ownership. The dashboards pull data dynamically from the Tanzania Open Data portal APIs, and uses a powerful combination of Angular, ReactJs, CSS, and HTML for the user interface. For mapping, they use Leaflet and Open Street Maps. The resultant source code (Water, Education and Health) is also open for download and further extension.

<p>Key libraries and technology used to build the dashboards</p>

Figure 1: Key libraries and technology used to build the dashboards.

The technology stack was chosen based on needs and functionality. In the case of education, since there was already an application in place that what we had to enhance, the whole development was done in AngularJS; for the other two dashboards we had greater flexibility to choose the stack.

From a user experience perspective, we had determined to include both georeferenced visualizations (maps) and charts as part of the dashboard. In maps, clusters are a useful way to show data points, especially in maps with high datapoint density — the Water Dashboard’s waterpoints are a strong example of this. After some research, we found Prune Cluster, useful because it (i) permits a category to be specified for the markers, after which (ii) a small object representing the number of markers for each category is attached to the clusters. In this way, you can create cluster icons adapted to their content — in our case, it helped us to show water point status for a group of the points at-a-glance.

<p>Waterpoint Map with Prune Clusters</p>

Figure 2: Waterpoint Map with Prune Clusters.

When developing charts, our goal was to create powerful tools for data analysis — we also needed to use a specific library, based on the government’s requirements. For these reasons, we selected the Highcharts library. Both the charts and maps have been optimized for low-bandwidth environments.

<p>Tanzania Health Dashboard charts</p>

Figure 3: Tanzania Health Dashboard charts.

The dashboards are built on open data released by the Government of Tanzania, and read the raw data published on the Tanzanian Government Open Data Portal (ODP). The data are displayed on the map at the most disaggregated level possible — some datasets from ODP only contain data that is aggregated at the region, district, or ward level.

Data sources for each dashboard are as follows:

In some cases, the dashboards use data that has been cleaned or merged with other data to create new datasets — but even in these cases, the data itself draws exclusively from open data released by the Government of Tanzania. All data is available for download in machine-readable format, with a license that encourages re-use.

However, it is important to note that the dashboards are as good as the raw data sources which drive them. Consequently, an important consideration for all the dashboards should be the provision of timely and complete data by the sector ministries. We are hopeful that planned trainings and sensitizations carried so far will increase the uptake. Naturally, timely updates and relevance to the latest approved datasets will stimulate demand through increased interest and use by end users.

In summary, we hope these dashboards help government and the public quickly sift through the government’s open data, and make it easier to understand, analyze, and use this information. But technology is just the first step in a data-driven decision-making chain: it is now for individuals to use their judgement and leadership instincts to act upon this information to progress toward national goals. We look forward to championing the Government of Tanzania’s successes, and working with our partners DataVision International and the World Bank again in future.

Share This Post

Related from our library

An aerial view of Bogotá, Colombia on the cover of the Data Value Project’s white paper. Photo credit- Bergslay:Pixabay.
Challenging Pessimists—and Optimists—to Reimagine Data and Power

Josh Powell and Jenna Slotin reflect on the Data Values Project and building a movement for change in data for development.

May 10, 2022 Global Data Policy
April 7, 2022 Agriculture
AD3ventures in Mapping

Administrative data – data from civil registration and vital statistics systems, and sectoral (health, education, judicial, social protection) management information systems – is rich with insights on citizen needs and priorities. Though timely and disaggregated administrative data can complement official statistics, it comes with its own data quality, availability, and use challenges.

March 16, 2022 Agriculture