time series - ARIMA component with seasonality - Using R -


i trying use auto.arima function in r on time series data.i used both original series , log transformed series , below outputs:

with original timeseries:

> auto.arima(inflowts) series: inflowts  arima(1,1,0)(0,1,0)[12]                      coefficients:           ar1       -0.6812 s.e.   0.1431  sigma^2 estimated 16565:  log likelihood=-137.88 aic=279.77   aicc=280.4   bic=281.95 

with log transformed series got:

> auto.arima(inflowts1) series: inflowts1  arima(1,1,0)(0,1,0)[12]                      coefficients:           ar1       -0.6695 s.e.   0.1488  sigma^2 estimated 0.008878:  log likelihood=20.97 aic=-37.93   aicc=-37.3   bic=-35.75 

the difference aic , bic parameters have become better when transformed variable used.

since data has seasonality question how handle seasonal component here giving both p , q values 0 [arima(1,1,0)(0,1,0)[12] ]

any pointer on how handle such scenario highly appreciated.


Comments

Popular posts from this blog

python - How to insert QWidgets in the middle of a Layout? -

python - serve multiple gunicorn django instances under nginx ubuntu -

module - Prestashop displayPaymentReturn hook url -