Malaria
Monthly surveillance for each district. Models estimate confirmed cases for an age and sex group from the case numbers of recent months, and can also use outside factors such as rainfall and temperature.
Typical inputs
- Past months: district and case counts
- Outside factors: rainfall, temperature, vegetation
- One value per month
Targets
- Confirmed cases by age and sex group
- District totals
Available models
All malaria models →Open a model to see what it needs and test it with your own numbers.
Quick API examples
Each model needs different values. Fetch its ready-made example, send it back, then change the numbers. Replace <model> with a name from the list.
# 1. Save a ready-made example request curl -s "https://comp-api.ocular-project.com/models/<model>/example?op=forecast" \ | python3 -c "import json,sys; print(json.dumps(json.load(sys.stdin)['requests'][0]))" \ > request.json # 2. Send it curl -s -X POST "https://comp-api.ocular-project.com/forecast/<model>" \ -H "Content-Type: application/json" -d @request.json
Tuberculosis
Estimates of future TB case numbers for a district, to support planning and surveillance.
Typical inputs
- Some models learn the past case numbers during training and need no input
- Others use recent case numbers and outside factors you provide
- One value per month or per quarter, depending on the model
Targets
- Positive TB cases per period
- District totals
Available models
All TB models →Open a model to see what it needs and test it with your own numbers.
Quick API examples
Check what a model can do first: capabilities in its details says whether it can forecast.
# 1. Details and capabilities curl -s "https://comp-api.ocular-project.com/models/<model>" # 2. What a forecast request must contain (JSON Schema) curl -s "https://comp-api.ocular-project.com/models/<model>/schema?op=forecast" # 3. Ready-made example requests to start from (send one item of "requests" as-is) curl -s "https://comp-api.ocular-project.com/models/<model>/example?op=forecast"
Model types
Machine-learning (e.g. XGBoost) and statistical (e.g. Prophet) models. Forecasts use past values and optional outside factors such as rainfall.
API endpoints
/predict/<model>,
/forecast/<model>,
/evaluate/<model>, depending on what each model supports.
Data governance
Only send aggregated, de-identified data (for example monthly district totals), never patient details. Results support planning and surveillance, not decisions about individual patients.