I'm doing ARIMA with several variables on a time series.
Arima value decimalvar1 decimalvar2, ar(1)
Decimalvar1 and decimalvar2 end up being really close to value. Decimalvar1 coefficient is like 30 000 while value let's say 70 000.
I get them in the results as coefficients in the same box as the constant. The lags are in a different box showing decimal values. I'm supposed to get the decimalvars
as decimal values aswell. More like 1.01 rather than 30 000
THe decimalvars are a list of changes from normal I think? It's supposed to show interest based on how many searched for it, and it starts with 0. It shows values like. -0.01 0.03 - 0.12. Almost all of them negative
So my question is, should I format these decimalvars to something else before the regression? for example with start value 1, then 0,99? Or how?
Arima value decimalvar1 decimalvar2, ar(1)
Decimalvar1 and decimalvar2 end up being really close to value. Decimalvar1 coefficient is like 30 000 while value let's say 70 000.
I get them in the results as coefficients in the same box as the constant. The lags are in a different box showing decimal values. I'm supposed to get the decimalvars
as decimal values aswell. More like 1.01 rather than 30 000
THe decimalvars are a list of changes from normal I think? It's supposed to show interest based on how many searched for it, and it starts with 0. It shows values like. -0.01 0.03 - 0.12. Almost all of them negative
So my question is, should I format these decimalvars to something else before the regression? for example with start value 1, then 0,99? Or how?