Click here to close the sidebar. ↗
The visualization consists of three planes. The leftmost plane visualizes the number of trips between regions in the first week of February 2020 (2020/02/03 to 2020/02/09). The thickness of the lines between the regions indicates the relative share of trips the total inter-regional trips that were measured in that week and the size of the doughnuts visualize the share of inter-regional trips departing from a region. You will see that the doughnut in Greater Accra is the largest doughnut, indicating that of all inter-regional trips, most departed from Greater Accra. Similarly, the thickest line is between Easter Region and Greater Accra, meaning that of all the inter-regional trips in Ghana, most were between Easter Region and Greater Accra. To protect sensitive data, the aggregates of weekly inter-regional trips were first rounded before they were plotted and even these rounded number on which the line thickness and doughnut size were based are not included in the visualization.
Instead, the figures in the left-most plot show, per region, what percentage of inter-regional mobility departed from a certain region to another region. So, the percentage 5.9% from Greater Accra to Volta, means that of all the trips that departed from Greater Accra, 5.9% went to Volta. This figure is repeated in both the doughnut for Greater Accra and the line segment between Greater Accra and Volta. Similarly, 44.8% of inter-regional trips that departed from Volta went to Greater Accra. This also means that the first percentage ("Total Outflow") that is shown in pop-up menu in every doughnut is always 100%.
The middle and rightmost plane are interpreted differently. The "Total Outflow" percentage shown in the doughnuts is no longer fixed to a 100%, but now indicates the change in total inter-regional trips departing from the 16 regions. So, the -41.7% in the pop-up menu of Greater Accra in the center plane, indicates that total number of inter-regional departing from Greater Accra decreased by 41.7%. The other percentages in the pop-up menu still indicate the same data as in the left plane (namely what percentage of total inter-regional trips from a given region went to another region, these will also still sum to 100%). The percentage that are shown when clicking on the lines, now indicate the change in that specific mobility between two regions. So, the -34.1% that shows when clicking on the line segment from Greater Accra to Easter Region, means that in the first week of April 2020 (specifically, 2020/03/30 to 2020/04/05) the number of trips between Greater Accra and Easter Region decreased by 34.1% as compared to the first week of February 2020. The right plane (2021-02-01 to 2021-02-07) is interpreted similar to the center one, meaning that for both the reference period is the left plane.
This visualization was built by Laurent Smeets (GSS) as a showcase of the possibilities of CDR data and inspired by similar map by Statistics Netherland (CBS) and Institut national de la statistique et des études économiques (INSEE)[Link to Code and more information] . At this stage, this visualization should very much be treated as a proof of concept and experimental. Adjustments to this map can be expected.
The input data is based on aggregated and anonymized data from Vodafone Ghana though the ‘Data for Good’ project.
When interpreting this map, it is paramount to keep in mind all the usual limitations of CDR data and the fact that in this version numbers are not yet adjusted for potentially increasing or decreasing telephone users. Finally, for this visualization a Directed all pairs’ origin-destination matrix is used, meaning that intermediate stops from one region to another are also included. For a more detailed analysis of the results and limitations, please refer to the [third Mobility analysis report].
last updated: 22/04/2021