The new function for making a composite model (combining outputs from multiple models) assumes that all predicted variables use the same model weights. In some cases it may be desirable to weight models differently when combining for flow versus temp. This would need to be added later.
Also, the function requires that the supplied model weights sum to 1. This was included to flag errors in manual weights, but it could be revisited if we want it to instead assume relative weights.
The new function for making a composite model (combining outputs from multiple models) assumes that all predicted variables use the same model weights. In some cases it may be desirable to weight models differently when combining for flow versus temp. This would need to be added later.
Also, the function requires that the supplied model weights sum to 1. This was included to flag errors in manual weights, but it could be revisited if we want it to instead assume relative weights.