R help - Sep 2005 - Crash with seasonal ARIMA

The following command crashes my Mac OS X version of R:
(I'm running R on a PowerMac G5, with 1 GB of RAM and dual processors.)

 > arima.0 <- arima(w3.ts,order=c(1,0,0),seasonal=list(order=c 
(1,0,0),period=365))

-David



Here is some background:

w3.ts is hourly temperature data with about 20% NA's.  It contains  
about 3.5 years of data.

I am using period = 365, which makes sense me to me.  Is this wrong?



 > summary(w3.ts)
    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's
   14.00   54.00   68.00   66.09   78.00  108.00 2976.00

 > length(w3.ts)
[1] 29542

 > frequency(w3.ts)
[1] 24

 > deltat(w3.ts)
[1] 0.04166667

 > w3.ts[300:400]
   [1] 61 63 64 68 66 64 61 59 62 56 53 49 43 45 47 46 44 42 42 40 42  
48 52 55
[25] 57 59 60 60 60 59 52 50 49 45 47 41 38 36 37 35 36 37 39 39 38  
42 49 54
[49] 58 61 63 64 63 62 59 55 54 51 55 53 53 NA NA NA NA NA NA NA NA  
NA NA NA
[73] NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA  
NA NA NA
[97] NA NA NA NA NA

 > w3.ts[720:770]
[1]       NA       NA       NA       NA       NA       NA        
NA       NA
[9]       NA       NA       NA       NA       NA       NA        
NA       NA
[17]       NA       NA       NA       NA       NA       NA        
NA       NA
[25]       NA 41.50000 46.50000 40.00000 42.50000 37.50000 41.00000  
36.50000
[33] 40.50000 38.50000 42.50000 41.00000 46.00000 43.50000 46.50000  
43.50000
[41] 47.50000 41.50000 47.50000 42.50000 47.33333 47.66667 51.00000  
41.50000
[49] 51.50000 40.00000 52.50000


-David

Actually, I think period = 365 * 24 = 8760 is really what I need.   
That crashes arima as well.
> I am using period = 365, which makes sense me to me.  Is this wrong?

	[[alternative HTML version deleted]]