The Sandbox
An Organized Platform for Uganda Malaria Analytics
High quality, robust analytics for malaria in Uganda have the potential to leverage information and transform malaria control. Here, we describe the concept of a shared sandbox, an organized platform that enables development and maintenance of malaria analytics for malaria control programs.
In the overview, we described some guiding principles: malaria analytics should be timely, replicable, and robust. Malaria analysts should be held accountable for the advice they give. To make this work, we have organized a sandbox for ops and dev.
In Uganda, the HMIS is maintained by the Department of Health Information (DHI). Surveillance data, including the facility data, are transmitted electronically and entered into a DHIS2 database. There is also a large set of research data describing malaria. To do some kinds of analysis, we need access to other data assets: gridded human population surfaces; environmental covariates; and other data. The sandbox concept was set up to organize these data so they were available for users, and some of these data assets are used in prod.
ramptools
An R package that holds some important shared data assets and functions to perform common tasks
ramptools is an R package on github: https://github.com/dd-harp/ramptools. The R package was developed to make it easy to use some shared data assets and perform some common tasks:
loc_table– a location table to navigate spatial hierarchies: health facilities are in sub-counties, which are in districts in regions in Ugandaindicator_table– meta-data for the malaria indicators in theshapefiles – a set of shape files has been included that were aligned with the location table with sub-county, district, and regional boundaries. In addition, the shapefiles include water bodies to make it easy to visualize outputs.
github repositories
Algorithms for data processing and analytics
Some of the prod algorithms are stored in github repositories (see github (overview)).
Data Warehouse
The warehouse provides clean, up-to-date, access controlled data assets for malaria analytics
Uganda’s malaria data warehouse is a defined set of data assets developed and maintained DHI and NMED. As the data assets are updated in prod, authorized members of NMED, MoH, or supporting teams can access the data assets through a web server. Each one of the data assets (see the sidebar) describes what it contains and how it can be accessed.
The server details are currently being worked out through discussions with Paul Mbaka (DHI), Rek John (NMED), and the RAMP team.
Warehouse or Lakehouse?
Also, see The Data Warehouse
Websites
We have developed a set of websites to store and disseminate information. The two main websites are:
Adaptive Malaria Control in Uganda (this website)
Malaria in Uganda
Adaptive Malaria Control in Uganda
This website holds information about the sandbox.
Malaria in Uganda
Malaria in Uganda shares visualizations and routine analytics with NMED.
MoH / DHIS2 browser
The data browser that Meddy built