I did SPI and SPEI work through R studio. As you know, r program for each month; We draw results for 1, 3, 6, 9, 12 and 24 months. However, when interpreting the results, we interpret the 3-month results as seasonal. We interpret 12-month-results as normal years or water years. but, in what interval do you think we should interpret the results of 6 months, 9 months and 24 months? i saw the time intervals are taken differently according to the subject and purpose discussed in the articles. Ideally, is there a standard? The most i curious 9-month series, I think I will take the water year as the beginning. As the October-June range?
If I understood your question correctly, first of all try to study the global series by specifying the seasonality in parameters (e.g. months), according to the lags preview you can divide your series in several segments or apply preprocessing to make the series stationary.
Hi Serhat, In my opinion there is no a defined standard in this field and it depends on your purpose.
Jaafar Idrais thank you for answer. but, Can you give an example to the global series?
Seyed Mahdi Amir Jahanshahi yes I think the same. it depends on our own purpose. but I am open to different points of view. I'm trying to find out which range values would be right for my study. thank you for answer.
Serhat Şenol you can find an exemple in this link, I hope that will be helpful (Y)
https://www.analyticsvidhya.com/blog/2015/12/complete-tutorial-time-series-modeling/
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series