n births per month in NYC Decomposition of additive
Stat 333, Knudson Decomposing TS and seasonal ARIMA Apr 26, 2018 Learning Log for 04/26/2018 We talked about Decomposition of Additive Series today. Take a look through this R code to see what we did. All the R-code is commented with what we learned and its application. Decomposing Time Series
n births per month in NYC #install.packages(forecast) births seasonal AR (p >= 1) This stuff here is not a difinitive thing. You more gather information and then do some testing. # Our model thus far is P>=1, D>=1, Q>=1 STEP 4: ESTIMATE MODEL
ARIMA (p, d, q) x (P, D, Q) S ARIMA (1, 0, 1) x (1, 1, 1) 12 #mod1 =1 and we tried all EQUAL to 1
ARIMA (2, 0, 0) x (1, 1, 1) 12 #mod3 = 1) From the graphs, it looks to be that q = 0. This is because there are no significant spikes in the Acf(diff1) plot in the first points. However, we see that there is a big spike at time 12. This means that Q = 1. Since there are no significant spikes in the Pacf(diff1) plot, we conclude that p = 0 . This is because there are no significant spikes in the first points. We see that there is a big spike at time equal to 12. Thus, we can set P = 1.
Step 4 – Estimate Model p = 0 , P = 1 q = 0, Q = 1 d = 1 , D = 1 c(p,d,q) - nonseasonal c(P,D,Q) - seasonal part #mod1
n births per month in NYC #install.packages(forecast) births seasonal AR (p >= 1) This stuff here is not a difinitive thing. You more gather information and then do some testing. # Our model thus far is P>=1, D>=1, Q>=1 STEP 4: ESTIMATE MODEL
ARIMA (p, d, q) x (P, D, Q) S ARIMA (1, 0, 1) x (1, 1, 1) 12 #mod1 =1 and we tried all EQUAL to 1
ARIMA (2, 0, 0) x (1, 1, 1) 12 #mod3 = 1) From the graphs, it looks to be that q = 0. This is because there are no significant spikes in the Acf(diff1) plot in the first points. However, we see that there is a big spike at time 12. This means that Q = 1. Since there are no significant spikes in the Pacf(diff1) plot, we conclude that p = 0 . This is because there are no significant spikes in the first points. We see that there is a big spike at time equal to 12. Thus, we can set P = 1.
Step 4 – Estimate Model p = 0 , P = 1 q = 0, Q = 1 d = 1 , D = 1 c(p,d,q) - nonseasonal c(P,D,Q) - seasonal part #mod1
