Independent Variables - % Change V/s Absolute Value
(3/5 from 1 vote)
I have a question on input data - independent variables. I have multiple time series as my input data, with some in % change form and the other in absolute value. My dependent variable time series is in absolute value form as well. Is it ok to have independent variable time series in 2 different forms - % change and absolute number? Should I change (manually) all the independent variables to % change form? My reasoning for keeping both forms is that some variables are index values while others are month over month changes.
You can keep the format of the independent values and have some in percentages and some in absolute values. Bear in mind that you will need to enter dats in the same formats in the forecast to be consistent with the regression equation.