This setting specifies the maximum number of predictors included into the regression model. Specifying:
-1 – No regression at all is estimated. The ”raw” hourly day type quotients will be used solely to define the relation between the auxiliary and the forecast series. As a simple equation this can be written:
F = QA
where F denotes the forecast series, Q the hourly day type quotients and A the auxiliary series
1 – Only one predictor will be included. Since the hourly day type quotients in this setup are the main explanatory factors, these will most surely be picked. The result can be written
F=αQA+c
where α and c are the estimated coefficient and constant.
2-5 – Also weather parameters can be included. This can in simplified form be written as
F=αQA+βW+c
where βW denotes the estimated weather adjustments in a compact form.