Hello, i got trouble when i try identify a intervention analysis about disaster impact on night time light intensity. i have 65 month data from January 2015 until may 2020. based on pre-intervention analysis, i have Arima (0,0,0)(0,0,0). but when i try to do pulse intervention analysis i don't know how to do fitted the arima intervention analysis? anyone know solution for the problem?
Guy Mélard
There is no problem of combining a pulse with a random noise model. It is a simple regression model with a dummy variable as the explanatory variable. There no need to use a software package for ARIMA models. But look closely at the residuals to be sure there is no delayed effect.
1 votes
1 thanks
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series
Similar questions and discussions