We have a "missing signal detector" that looks at our metadata to identify signals that are present in the database, but which lack associated semantic metadata information. We should find a way to run this on a schedule and report results to the team so we can act upon it, or at least task someone with running this manually on a periodic basis.
Note that a failure of a GH Actions job for this does not (necessarily) mean that the script did not run as intended... When no missing signal information is found, the job reports "success"; When missing signal info IS found, an artifact is generated by the job that lists offending signals, and the job is reported as a failure.
We have a "missing signal detector" that looks at our metadata to identify signals that are present in the database, but which lack associated semantic metadata information. We should find a way to run this on a schedule and report results to the team so we can act upon it, or at least task someone with running this manually on a periodic basis.
The python code that does the core functionality: https://github.com/cmu-delphi/delphi-epidata/blob/dev/scripts/report_missing_covidcast_meta.py
GH Action job histories: https://github.com/cmu-delphi/delphi-epidata/actions/workflows/missing_signals.yaml
GH Action definition: https://github.com/cmu-delphi/delphi-epidata/blob/dev/.github/workflows/missing_signals.yaml
Note that a failure of a GH Actions job for this does not (necessarily) mean that the script did not run as intended... When no missing signal information is found, the job reports "success"; When missing signal info IS found, an artifact is generated by the job that lists offending signals, and the job is reported as a failure.