search

robjhyndman / forecast

Forecasting Functions for Time Series and Linear Models
http://pkg.robjhyndman.com/forecast
1.12k stars 342 forks source link

How to determine seasonality in auto.arima #139

Closed zhuzelong closed 9 years ago

zhuzelong commented 9 years ago

Wish you would not mind me asking questions here. I am a bit confused of the Arima() and auto.arima().

I have an irregular time series (unit is day) stored in a zoo object. It is irregular because no data on holidays and weekends.

The function as.ts() converted the zoo into a regular ts with NA interpolated. But the resulting ts has a frequency 1. So auto.arima() seemed to exclude all seasonal terms even though I specified D=1, max.D=30.

If I converted the zoo explicitly by ts(zoo, start=, frequency=340) (340 is due to exclusion of holidays and weekends), then auto.arima would include seasonal terms.

The question is, the ts is irregular and conversion by ts(..., freq=340) led to wrong time series. Is there any approaches that forecast can deal with the issue?

robjhyndman commented 9 years ago

This is not a help site. It is intended for bug reports only. Please ask questions on crossvalidated.com