This work is led by the Institut National de Recherche Biomédicale (INRB) Kinshasa/One Health Institute for Africa (INOHA) Kinshasa (Dav Ebengo, Placide Mbala-Kingebeni and Tania Bishola), and the Institut National de Santé Publique (INSP) (Pierre Akilimali, Adelard Lofungola).
Collaborating institutions and agencies
- Institut National de Santé Publique (INSP)
- National Institute of Biomedical Research (INRB)
- Africa Centres for Disease Control and Prevention, Addis Ababa, Ethiopia
- World Health Organization, Geneva, Switzerland
- World Health Organization Country Office, Kinshasa, Democratic Republic of the Congo
- Northeastern University, United States
- University of Oxford, United Kingdom
Please note that the epidemiological data presented here is based on work in progress and should be considered preliminary. Our analyses are ongoing, and a publication communicating our findings is in preparation. Contextual data are publicly accessible; please refer to the original license when re-using these data. If you intend to use the epidemiological data prior to our publication, or have other enquiries, please contact Prof. Placide Mbala-Kingebeni (INRB, DRC), Prof. Dav Ebengo (INRB, DRC), and Pierre Akilimali (INSP).
Last successful build: 11 June 2026, 18:07:02 (UTC) — build/ on main at commit 37f84e5 (data snapshot 37f84e5, see build/manifest.json).
- DRC health zones: Humanitarian Data Exchange (MoH zones de santé shapefile)
- Epidemiological data (INSP): Institut National de Santé Publique (INSP) SitRep MVE PDF series (
data/insp_sitrep/, currently through SitRep 023) — daily case, death, and contact-tracing indicators by health zone manually transcribed from the sitreps - Processed Linelists: Following establishment of an epi data collection pipeline by INSP and INRB, aggregated linelist data will be housed in (
data/epi/) - Operational data (INSP): Institut National de Santé Publique (INSP) Contextual data on the public health response transcribed from SitRep MVE PDF series (
data/insp_sitrep/, currently for SitReps 020-023, but backfilling is in progress) (data/public_health_response) - Testing Capacity (AfricaCDC): data on PCR machine and reagent availability (
data/testing_capacity)
- Road travel times: OSRM public demo (
data/osrm/, matrix outputs) - Cross-border travel: Imperial College Report
- Internal relocations: International Organisation for Migrants (IOM)
- Mobile phone-based internal relocation estimates: Flowminder.org (
data/flowminder/anddata/flowminder_short_trips/)
- Conflicts and acts of violence: ACLED (
data/ACLED_conflict/) - Not currently included in build - Population counts: WorldPop (
data/worldpop/) - Health facilities (GRID3): GRID3 COD Health Facilities v8.0 (
data/grid3_healthsites/) - Health facilities (OSM / crowdsourced): Healthsites.io (
data/healthsites_io/) - Socioeconomic deprivation and inequality: Climate-Conflict Vulnerability Index (
data/ccvi/) - Degree of Urbanisation: FAO Land Classification System (
data/fao_lccs/) - GPD per capita: Kummu et. al (
data/gdp_pc/)
For the latest BDBV genomic data, please visit Pathoplexus.
We are tracking pending data sources over on the issues tab. If you want to request a specific publicly available dataset, raise an issue (although raising an issue does not guarantee that we will incorporate a dataset).
The current build is committed on main and refreshed automatically by CI on every merge that touches data/** — see Release internals. Run python -m tools.build_geojson locally only if you're working on a branch with un-merged data changes.
updating data with sitRep 27
Per-layer catalogue for the current build:
- Embedded in the GeoJSON — vector layers merged into
build/drc_health_zones.geojson - Matrix outputs — origin–destination tables (not in the GeoJSON)
Full tables live in data/README.md. Machine-readable index: build/manifest.json.
Not in build: ACLED_conflict — province-grain placeholder, no QA-passing output yet.
| Tag | Date | Summary | Download |
|---|---|---|---|
build-2026-06-11-37f84e5 |
2026-06-11 | updating data with sitRep 27 | release |
build-2026-06-11-e1e67f3 |
2026-06-11 | Removed aggregated linelist data | release |
build-2026-06-11-1499e80 |
2026-06-11 | Decomposed and summarized information from different pillars based on the previous version in order to make the pillar information more concise. | release |
build-2026-06-10-8645bde |
2026-06-10 | Adding New sitrep_026 | release |
build-2026-06-10-1b71e8e |
2026-06-10 | Aggregated confirmed positives from the INSP linelist are now included in data/aggregated_insp_linelist. At present these are aggregated to the province level. | release |
build-2026-06-10-10d11cb |
2026-06-10 | - Create a folder for archived weekly WHO reports (unprocessed) to serve as a digital record | release |
build-2026-06-10-690016b |
2026-06-10 | New sitrep Adding | release |
build-2026-06-09-6bc4479 |
2026-06-09 | Update to the public health context section, where province level actions are now stored separately. In general, we have adjusted the QA and build process to allow province level info, which just gets broadcast to each included healthzone during the build. This will be relevant for many datasets, but should obviously be used carefully (e.g. a numeric metric such as total hospitalised on the province level, broadcast to health zone level, may be misinterpreted as overreporting that metric for the healthzone). | release |
build-2026-06-09-60230f8 |
2026-06-09 | Integration of SITREP 24 data | release |
build-2026-06-08-979a344 |
2026-06-08 | Adding Pillars od sitrep 21 | release |
build-2026-06-07-df291a5 |
2026-06-07 | Update data of sitRep 23 from June 6, published on June 7 | release |
build-2026-06-06-39fc6f0 |
2026-06-06 | Addition of new data according to SitRep 22 | release |
build-2026-06-05-628c054 |
2026-06-05 | Sitrep 21 Adding | release |
build-2026-06-04-169614d |
2026-06-04 | Data sitrep020, updating for cumulative_confirme_cases, cumulative_confirme_deaths, Nation_cumulative_confirme_cases, Nation_cumulative_confirme_death, Nation_cumulative_isolation | release |
build-2026-06-03-8d24ff8 |
2026-06-03 | Revised numbers of national cumulative suspected cases given latest reporting (see data/insp_sitrep/reports/SitRep_MVE_019-2026.md). |
release |
build-2026-06-03-06ffe1a |
2026-06-03 | SitRep 19 data added and digitisation report updated. | release |
build-2026-06-03-ea78c16 |
2026-06-03 | Added public health pillar data | release |
build-2026-06-02-32e9ebd |
2026-06-02 | Sitrep 18 added, national counts updated and healthzone level counts where reported. | release |
build-2026-06-02-f3b3051 |
2026-06-02 | Flowminder short trip data is now formatted to be visualised in the dashboard. | release |
build-2026-06-02-125e4e0 |
2026-06-02 | Report added for Sitrep 17 describing digitisation process | release |
build-2026-06-02-d1ceb9c |
2026-06-02 | Sitrep 17 Added | release |
build-2026-06-01-b4cafc9 |
2026-06-01 | Updates to SitReps 15 and 16 | release |
build-2026-06-01-0a87d65 |
2026-06-01 | - National level tables now just take the nom 'DRC' | release |
build-2026-05-30-507a2a2 |
2026-05-30 | - Added sitrep 14 (in reality fix some issues with sitrep 14, but those issues aborted the earlier release so this is the first release with sitrep 14 anyway | release |
build-2026-05-30-e125835 |
2026-05-30 | Latest release fixing a number of minor processing issues in past sitreps. | release |
build-2026-05-30-5a34b18 |
2026-05-30 | An additional data folder flowminder_short_trips/ is created. This contains updated Flowminder data for short trips for April 2026 (see report in data/raw/ for details). QA tests show warnings in unrelated data/** folders. |
release |
build-2026-05-29-ff1e796 |
2026-05-29 | Sitrep 13 added. | release |
build-2026-05-28-bb8b7d5 |
2026-05-28 | Updated to allocate a new suspected case from Mabanga (not a healthzone) to the Mambasa healthzone. Team at INRB reviewed and decided this is the most accurate place to put it for now, but let's note that there may also be a place called Mabanga in Mangala | release |
build-2026-05-27-e40bc9e |
2026-05-27 | - Healthzone level wpi data up to 26th May from INSP (The public sitrep did not have this data) | release |
build-2026-05-27-059661a |
2026-05-27 | - Updated INSP Sitrep data with the new version of Sitrep 12 (Updated national suspected deaths) | release |
build-2026-05-27-af1f2b5 |
2026-05-27 | - Added the updated DRC totals from SitRep 12 to a new metric for that dataset with prefix national_* | release |
| build-2026-05-26-683a564 | 2026-05-26 | INSP Sitrep data through report 010 | release |
build-2026-05-22-12db0c2 |
2026-05-22 | 25 vector layers; INSP through SitRep 007 + GRID3 health facilities | release |
build-2026-05-22-9694d10 |
2026-05-22 | First GitHub release (11 vector layers; pre-INSP / pre-GRID3) | release |
data/
README.md index of all dataset folders
shapefiles/ source of truth for health-zone boundaries
aliases.csv observed_name -> canonical_nom mappings
<dataset>/ one folder per source
raw/ untouched source files
process.{py,R} script that produces files in processed/
processed/ standardized contract-conformant outputs
metadata.yaml source, citation, retrieved_on, license, contact, runtime
README.md optional human notes
tools/
lib/schema.py canonical Noms, alias resolver, filename contract
qa.py walks data/, validates, writes qa/qa_log.csv & qa/matrix_log.csv
build_geojson.py merges passing non-matrix outputs into build/drc_health_zones.geojson
requirements.txt pyshp, pyyaml, shapely
qa/
qa_log.csv per-artifact QA results (all statuses)
matrix_log.csv catalog of QA-passing matrices
reports/<dataset>.md per-folder human-readable report
build/
drc_health_zones.geojson shapefile + latest per-zone values
long/<dataset>__<metric>.csv full long-format copy of each vector file
manifest.json sources + build timestamp
Join key: the canonical Nom from data/shapefiles/DRC_Health_zones.shp. The two natural collisions (Bili, Lubunga) are disambiguated with a province suffix, e.g. Lubunga (Tshopo). Observed spellings that differ are listed in data/aliases.csv. Province roll-ups use nom = shapefile PROVINCE (aliases in data/province_aliases.csv); national roll-ups use nom = DRC.
Processed-file naming: <dataset>__<metric>__<resolution>.{csv|matrix.csv} - <dataset> and <metric> are lower_snake_case. - <resolution> ∈ {static, daily, weekly, monthly, yearly}. - Suffix is .matrix.csv for matrix outputs, .csv for vector (one-row-per-zone) outputs.
Vector files carry a nom column. Non-static resolutions also carry a date column (ISO 8601).
Matrix files (.matrix.csv): snapshot matrices have header nom, <dest_nom_1>, ...; time-series matrices have date, nom, <dest_nom_1>, .... Present cells must be non-negative numeric; missing values may be empty or NA (e.g. unroutable OSRM pairs).
Contributors add or update data. PRs touch data/** (and tests/** and unrelated docs only) — never build/, qa/, dist/, or README.md's build/release sections.
-
One-time setup (anyone cloning):
git lfs install python -m venv .venv && .venv/bin/pip install -r tools/requirements.txtLFS is required because binary raw blobs (
*.xlsx,*.zip,*.pdf,*.tif, etc.) underdata/*/raw/are stored via Git LFS — see.gitattributes. -
Create
data/<your_dataset>/withraw/,metadata.yaml, and (when you have outputs)process.{py,R}+processed/. -
Make sure your processed filenames match the contract above. Add zone aliases to
data/aliases.csvand province aliases todata/province_aliases.csvwhen needed. -
Sync with main:
git merge origin/main -
Run unit tests + QA locally:
.venv/bin/python -m pytest tests/ .venv/bin/python -m tools.qa -
(Optional) Rebuild the merged GeoJSON locally to sanity-check your changes:
.venv/bin/python -m tools.build_geojson --skip-readmeDo not commit the resulting
build/,qa/qa_log.csv,qa/matrix_log.csv,qa/reports/, orREADME.mdupdates. Those land onmainautomatically when an admin merges your PR; including them in your PR causes merge conflicts and gets flagged in review. -
Open a PR. Fill in the
## What's newsection in the PR body (template provided) — that text becomes the GitHub Release description and the README "what's new" block when this PR is released. CI runspytest+tools.qaand blocks merge on any failures. -
Wait for admin review and merge. You don't run a release — CI does that automatically.
Admins (maintainers with write access to main) review PRs and merge.
-
Review the PR: data diff, CI green,
## What's newsection populated and accurate, contributor checklist ticked. -
Merge to
main. That's it for the common case — the release workflow takes over.
Escape hatches:
-
Suppress release for a trivial change (e.g. typo fix in a metadata file): include
[skip release]in the merge commit message. CI will skip the release step. -
Force a release without a data change (e.g. after fixing
tools/build_geojson.py): go to the Actions tab → "Release on data merge" → "Run workflow", and supply a description via the manual input. -
Emergency local release (CI is down): pull
main, then run the same sequence the CI workflow runs:.venv/bin/python -m tools.qa .venv/bin/python -m tools.build_geojson .venv/bin/python -m tools.release # interactive; packs dist/<tag>.tar.gz + updates README git add build/ qa/qa_log.csv qa/matrix_log.csv qa/reports/ README.md git commit -m "New build YYYY-MM-DD" git push .venv/bin/python -m tools.publish # creates the GitHub Release pointing at HEADThe publish step is separate from the pack step so the GitHub Release tag points at the commit that contains the build artifacts (the push above), not the pre-build merge commit.
Maintainers who will cut emergency local releases also need:
ghCLI installed and authenticated (gh auth login) — required bytools.publish, not bytools.release.$EDITORset (used bytools.releasefor the interactive description prompt).
The release workflow (.github/workflows/release.yml) runs on push to main when data/** changes (and on manual workflow_dispatch).
What it does, in order:
- Bails if the HEAD commit message contains
[skip release]. - Extracts the
## What's newsection from the merge commit's PR body (viagh api). - Runs
python -m tools.qa. - Runs
python -m tools.build_geojson. - Runs
python -m tools.release --description-file <tmp> --non-interactive, which packsbuild/asdist/<tag>.tar.gz, persists the description asdist/<tag>.description.md, and updates the README. This step does NOT publish anything. - Commits and pushes the resulting
build/,qa/, andREADME.mdback tomainwith[skip release][skip ci]in the commit message to prevent recursive triggering. - Runs
python -m tools.publish, which callsgh release create <tag> dist/<tag>.tar.gz --target $(git rev-parse HEAD) .... Because this runs after the commit-back, the release tag points at the commit that contains the build artifacts in its tree — not at the pre-build merge commit. The release URL is determined by<tag>and matches whattools.releasewrote into the README in step 5. - Dispatches a dashboard rebuild to
BDBV2026-Epidemic_Dashboardonmain, using the commit SHA from step 6 (the build-artifact commit). RequiresDASHBOARD_DISPATCH_TOKENon this repo; skips with a warning if unset.
The pre-existing qa.yml workflow runs pytest + tools.qa on PRs as the merge gate; it does not trigger on build/, qa/, or README.md changes, so the release workflow's commit-back does not retrigger it. The separate Trigger dashboard rebuild workflow is manual-only (escape hatch); production dashboard updates come from step 8 above.
Please cite the original data providers (links above) and this repository if any code or derived data is reused.
The repository code is licensed under the terms in LICENSE. We do not claim ownership of or the right to license the third-party data or software tools used. Please pass forward any existing license/warranty/copyright information when redistributing.
THE DATA AND SOFTWARE ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT.