Closed peterdudfield closed 2 months ago
decided actually to go with option 3. This will make the read function the faster I think
Once #85 is deployed, it would be great to back fill the last month of forecasts, attaching them to the model. This can be done with some SQL code
Ill makr this as complete for now, but might need to run migrations still
Detailed Description
We now want to run multiple ml models from one site, so we need to tag the forecast with the model name
Context
foreacstvalue
table, so dont want to do many joinsPossible Implementation
There are a few options
forecast
tableforecastvalue
tablemodel_name
andmodel_version
toforecastvalue
tableI think we should go for 3, to keep it simple for the moment
We would also need to