This PR builds a data dictionary into our code base and replaces upstream names with names from that data dictionary.
It also moves functions into a package infrastructure and adds population adjustment to state level Rt forecasts. Apologies for the permission noise.
Once merged I suggest we rename to covid.german.forecasts. This PR blocks updating this weeks forecast.
In a future PR I/we should remove the remainder of the none standard infrastructure and clean up within model code. Once this is complete we should be able to write up our evaluation more easily and extract some learning.
This PR does not in the main update code in either of the crowd forecasting folders as how code works in those folders is not clear to me.
Based on https://github.com/KITmetricslab/covid19-forecast-hub-de/issues/632 the Hub data dictionary cannot be assumed to be static or uniquely link codes with locations.
This PR builds a data dictionary into our code base and replaces upstream names with names from that data dictionary.
It also moves functions into a package infrastructure and adds population adjustment to state level Rt forecasts. Apologies for the permission noise.
Once merged I suggest we rename to
covid.german.forecasts. This PR blocks updating this weeks forecast.In a future PR I/we should remove the remainder of the none standard infrastructure and clean up within model code. Once this is complete we should be able to write up our evaluation more easily and extract some learning.
This PR does not in the main update code in either of the crowd forecasting folders as how code works in those folders is not clear to me.