Evaluating your model side-by-side with a range of other models

Since Chap integrates a large number of models under a unified interface, it is easy to compare the predictions of your model against those of alternative models using a variety of input data, metrics and visualisations.

The most powerful way of comparing models is to install the Chap modelling platform alongside a DHIS2 instance, allowing you to run your and other models through the GUI of the "Modelling App", which includes interactive side-by-side comparison of predictions by different model.

As a simpler setup, you can evaluate your own and other models using the Chap command line interface with the chap eval command, and then compare the resulting predictions using chap export-metrics. See the Evaluation Workflow for the recommended approach.

Note: The legacy chap evaluate command is deprecated and will be removed in v2.0. See the legacy examples for reference.