時(shí)間序列分析r語言程序(完整版)
【正文】 =39。)acf(x)pacf(x)錯(cuò)誤于x[i] = rnorm[i] 4/5 * rnorm[i 1] + 16/25 * rnorm[i 2] : 更換參數(shù)長(zhǎng)度為零 根據(jù)書上64頁來判斷（1，1）模型，x(t)(t1)=u(t)*(u1)，并直觀觀察該模型自相關(guān)系數(shù)和偏自相關(guān)系數(shù)的拖尾性。o39。o39。l39。o39。o39。o39。l39。LjungBox39。l39。o39。o39。o39。o39。LjungBox39。l39。l39。l39。LjungBox39。o39。LjungBox39。o39。o39。LjungBox39。LjungBox39。Ljung39。o39。o39。b39。b39。mean39。o39。l39。)acf(x)pacf(x)錯(cuò)誤于rnorm[i] : 類別為39。,xlim=c(1950,2010),ylim=c(60,100))=[,2]=acf()白噪聲檢驗(yàn)(,type=LjungBox,lag=6)(,type=LjungBox,lag=12)acf()pacf()根據(jù)自相關(guān)系數(shù)圖和偏自相關(guān)系數(shù)圖可以判斷為AR（1）模型 P81 口徑的求法在文檔上 P83arima(,order=c(1,0,0),method=ML)極大似然估計(jì)ar1=arima(,order=c(1,0,0),method=ML)summary(ar1)ev=ar1$residualsacf(ev)pacf(ev)參數(shù)的顯著性檢驗(yàn)t1=p1=pt(t1,df=48,=F)*2ar1的顯著性檢驗(yàn)t2=p2=pt(t2,df=48,=F)*2殘差白噪聲檢驗(yàn)(ev,type=LjungBox,lag=6,fitdf=1)(ev,type=LjungBox,lag=12,fitdf=1)predict(arima(,order=c(1,0,0)),=5)=predict(arima(,order=c(1,0,0)),=5)U=$pred+*$seL=$*$seplot(c(,$pred),type=l,col=1:2)lines(U,col=blue,lty=dashed)lines(L,col=blue,lty=dashed) ((n=100,list(ar=)))方法二x0=runif(1)x=rep(0,1500)x[1]=*x0+rnorm(1)for(i in 2:length(x)){x[i]=*x[i1]+rnorm(1)}plot(x[1:100],type=l)acf(x)pacf(x)擬合圖沒有畫出來x0=runif(1)x=rep(0,1500)x[1]=*x0+rnorm(1)for(i in 2:length(x)){x[i]=*x[i1]+rnorm(1)}plot(x[1:100],type=l)acf(x)pacf(x)方法一((n=100,list(ar=c(1,))))方法二x0=runif(1)x1=runif(1)x=rep(0,1500)x[1]=x1x[2]=*x0+rnorm(1)for(i in 3:length(x)){x[i]=x[i1]*x[i2]+rnorm(1)}plot(x[1:100],type=l)acf(x)pacf(x)x0=runif(1)x1=runif(1)x=rep(0,1500)x[1]=x1x[2]=x1+*x0+rnorm(1)for(i in 3:length(x)){x[i]=x[i1]+*x[i2]+rnorm(1)}plot(x[1:100],type=l)acf(x)pacf(x)又一個(gè)式子x0=runif(1)x1=runif(1)x=rep(0,1500)x[1]=x1x[2]=*x0+rnorm(1)for(i in 3:length(x)){x[i]=x[i1]*x[i2]+rnorm(1)}plot(
公司管理相關(guān)推薦
鄂ICP備17016276號(hào)-1