# auto.arima not working for seasonal=TRUE

I'm not sure whether I've just messed up something syntactically, if so I can't figure out what. I have the following data (res):

I can run

aa = auto.arima(res, seasonal=FALSE)


which successfully (returns ARIMA(5,0,0), but if I try

aa = auto.arima(res, seasonal=TRUE)


it runs instantaneously and returns a (0,0,0) model. Any advice what I am doing wrong?

• Can you possibly post your data? – Stats Mar 21 '19 at 14:23
• In stats.stackexchange.com/questions/317734/… @Adamo wisely pointed out the acf and pacf of series ( like this one !) that have been affected by deterministic structure should NOT be used to identify the SARIMA model.. Post your data like as asked and I will try to figure out the problem. It appears to be weekly data is that right ? – IrishStat Mar 22 '19 at 0:30
• Yes it's weekly data. I restarted R and it's seems to be working now, though I'm not sure why that would be. Thanks for stopping by to help. – James Mar 23 '19 at 14:28