时间序列例子.pptVIP

par(mfrow=c(2,4)) acf(e) acf(prod) acf(rw) acf(U) pacf(e) pacf(prod) pacf(rw) pacf(U) par(mfrow=c(1,1)) library(vars) data(Canada) VAR(Canada, p = 2, type = none) VAR(Canada, p = 2, type = const) VAR(Canada, p = 2, type = trend) VAR(Canada, p = 2, type = both) #看结果 (a=arima(test, order = c(9, 0, 9))) a$coe[c(9,18)] acf(a$res[-1],xaxp=c(0,20,10)) # /TSDL/books/odonovan.dat #SERIES 4.1 : CONSEC. READINGS BATCH CHEMICAL PROCESS R.B.MILLER et al. x=scan() 40 54 48 52 41 52 38 56 48 45 66 17 62 50 38 59 51 55 48 51 50 52 44 65 40 65 41 64 53 48 53 43 66 48 52 42 44 56 44 58 41 54 51 56 38 56 49 52 33 52 59 34 57 39 60 40 52 44 65 43 48 44 49 44 49 69 40 54 58 49 tem=ts(x,frequency = 1) par(mfrow=c(2,4)) plot(tem) qqnorm(tem);qqline(tem) acf(tem);pacf(tem) plot(diff(tem,1))#差分 a=arima(tem,order = c(1, 1,1)) a$aic #458.053 acf(a$res,xaxp=c(0,20,10));pacf(a$res,xaxp=c(0,20,10)) res=armasubsets(y=diff(tem,1),nar=14,nma=14,=temprature,ar.method=ols) plot(res) par(mfrow=c(1,1)) #Parameter Estimation # ARMA(1,1) set.seed(101044) x11=arima.sim(n = 100, list(ar = c(0.6), ma=c(-.3),sd = sqrt(0.1796))) par(mfrow=c(1,2)) acf(x11);pacf(x11) par(mfrow=c(1,1)) arima(x11, order=c(1,0,1),method=CSS) # conditional sum of squares arima(x11, order=c(1,0,1),method=ML) # maximum likelihood #Residual Analysis #plot residuals library(TSA) set.seed(101044) x=arima.sim(n = 100, list(ar = c(-0.5), sd = sqrt(0.1796))) par(mfrow=c(1,2)) acf(x);pacf(x) par(mfrow=c(1,1)) a=arima(x,order=c(1,0,1)) par(mfrow=c(2,2)) plot(a$res,ylab=Residuals,type=o);abline(h=0) acf(a$res);pacf(a$res) qqnorm(a$res);qqline(a$res) par(mfrow=c(1,1)) #3 diagnostic Display set.seed(101044) x=arima.sim(n = 100, list(ar = c(-0.3897),ma=-.4, sd = sqrt(0.1796))) par(mfrow=c(2,3)) acf(x);pacf(x) (a=arima(x,order=c(1,0,1)))#look at aic, log likelihood plot(a$res,ylab=Residuals,type=o);abline(h=0) acf(a$res);pacf(a$res) qqnorm(a$res);qqline(a$res) par(mfrow=c(1,1)) #tsdiag(a,gof=15,omit.initial=F) tsdiag(a)# This is a gene