1从1976年1月到2011年9月加利福尼亚州和美国的月失业率，数据在文件mCAUS-7611.txt中给出(变量为:year、mon、CA和US).
(a)为加利福尼亚州月失业率建立一个纯时间序列模型，进行模型检验，并写出拟合的模型.(b)对于加利福尼亚州月失业率，建立一个时间序列模型，使用美国月失业率的1阶滞后作为解释变量，进行模型检验，并写出拟合的模型.
©预测期从2008年1月到2011年9月，使用样本外预测，并比较这两个模型.这个是题目，da=read.table(“w-petroprice.txt”,header=T)
da1=read.table(“w-gasoline.txt”)
head(da)
pgs=log(da1[,1])
pus=log(da$US)
tdx=c(1:717)/52+1997 # calendar time
par(mfcol=c(2,1))
plot(tdx,pgs,xlab=‘year’,ylab=‘ln(price)’,type=‘l’)
title(main=‘(a) Gasoline’)
plot(tdx,pus,xlab=‘year’,ylab=‘ln(price)’,type=‘l’)
title(main=‘(b) Crude oil’)
acf(pgs)
dpgs=diff(pgs)
acf(dpgs,lag=20)
pacf(dpgs,lag=20)
m1=ar(diff(pgs),method=‘mle’)
m1$order
t.test(dpgs)
m1=arima(dpgs,order=c(5,0,0),include.mean=F)
m1
m1=arima(dpgs,order=c(5,0,0),include.mean=F,fixed=c(NA,NA,NA,0,NA))
m1
tsdiag(m1,gof=20)
#3.1.2
dpus=diff(pus)
m3=lm(dpgs~-1+dpus)
summary(m3)
acf(m3$residuals,lag=20)
pacf(m3$residuals,lag=20)
m4=ar(m3$residuals,method=‘mle’)
m4$order
m4=arima(dpgs,order=c(6,0,0),include.mean=F,xreg=dpus)
m4
m4=arima(dpgs,order=c(5,0,0),include.mean=F,xreg=dpus)
m4
m4=arima(dpgs,order=c(5,0,0),include.mean=F,xreg=dpus,fixed=c(NA,NA,NA,0,NA))
m4
tsdiag(m4,gof=20)
c1=c(NA,NA,NA,0,NA)
source(“backtest.R”)
pm1=backtest(m1,dpgs,316,1,fixed=c1,inc.mean=F)
c4=c(NA,NA,NA,0,NA,NA)
pm4=backtest(m4,dpgs,316,1,xre=dpus,inc.mean=F,fixed=c4)
tdx=tdx[2:717]
pm4fit=dpgs[317:716]-pm4$error
pm1fit=dpgs[317:716]-pm1$error
plot(tdx[317:716],dpgs[317:716],xlab=‘year’,ylab=‘growth’,type=‘l’)
points(tdx[317:716],pm1fit,pch=‘*’)
plot(tdx[317:716],dpgs[317:716],xlab=‘year’,ylab=‘growth’,type=‘l’)
points(tdx[317:716],pm4fit,pch=‘*’)
legend(par(‘usr’)[1], par(‘usr’)[4], xjust = 0,c(“*为预测”,“实线为真实值”),pch=16,col=c(“orange”,“green”))
#3.1.3
m6=lm(dpgs[2:716]~-1+dpus[1:715])
summary(m6)
acf(m6$residuals,lag=20)
pacf(m6$residuals,lag=20)
m7=ar(m6$residuals,method=‘mle’)
m7$order
m7=arima(dpgs[2:716],order=c(9,0,0),include.mean=F,xreg=dpus[1:715])
m7
m7=arima(dpgs[2:716],order=c(9,0,0),include.mean=F,xreg=dpus[1:715],fixed=c(NA,NA,NA,0,NA,0,0,0,NA,NA))
m7
tsdiag(m7,gof=20)
c7=c(NA,NA,NA,0,NA,0,0,0,NA,NA)
pm7=backtest(m7,dpgs[2:716],315,1,xre=dpus[1:715],inc.mean=F,fixed=c7)
Global temperature
Gt=scan(file=‘m-GLBTs.txt’)
Gtemp=ts(Gt,frequency=12,start=c(1880,1))
plot(Gtemp,xlab=‘year’,ylab=‘temperature’,type=‘l’) # Plot the data
acf(diff(Gt),lag=36)
pacf(diff(Gt),lag=36)
m1=arima(Gt,order=c(1,1,2))
m1
acf(m1$residuals,lag=36)
m1=arima(Gt,order=c(1,1,2),seasonal=list(order=c(0,0,1),period=24))
m1
tsdiag(m1,gof=36)
trend-stationarity
time=c(1:1568) # time index
m2=lm(Gt~time)
summary(m2)
par(mfcol=c(2,1))
acf(m2$residuals,lag=36)
pacf(m2$residuals,lag=36)
m2=arima(Gt,order=c(2,0,1),xreg=time)
m2
tsdiag(m2,gof=36) # Significant ACF at lag 24.
m2=arima(Gt,order=c(2,0,1),seasonal=list(order=c(0,0,1),period=24),xreg=time)
m2
tsdiag(m2,gof=36) # model checking
Comparison
source(“backtest.R”)
pm1=backtest(m1,Gt,1368,1)
time=as.matrix(time)
pm2=backtest(m2,Gt,1368,1,xre=time)

Gt=scan(file=‘m-GLBTs.txt’)
time=c(1:1568)
time1=c(rep(0,1212),time[1213:1568])
mm1=lm(Gt~time+time1)
summary(mm1)
x1=cbind(time,time1)
mm1=arima(Gt,order=c(2,0,1),seasonal=list(order=c(0,0,1),period=24),xreg=x1)
mm1
tsdiag(mm1,gof=36)
Box.test(mm1$residuals,lag=8,type=‘Ljung’)

da=read.table(“m-ncdc-noaa-glbtemp.txt”)
head(da)
tail(da)
da=da[1:1568,]
temp=da[,3]
m3=arima(temp,order=c(1,1,2),seasonal=list(order=c(0,0,1),period=24))
m3
tsdiag(m3,gof=36)
m4=arima(temp,order=c(2,0,1),seasonal=list(order=c(0,0,1),period=24),xreg=time)
m4
m4$coef
sqrt(diag(m4$var.coef))
m4$coef/sqrt(diag(m4$var.coef)) # Compute t-ratios
tsdiag(m4,gof=36)
Backtesting
pm3=backtest(m3,temp,1368,1)
pm4=backtest(m4,temp,1368,1,xre=time)
U.S. Monthly Unemployment Rates
da=read.table(‘m-unrate.txt’,header=T)
dim(da)
head(da)
unemp=da$rate
unrate=ts(unemp,frequency=12,start=c(1948,1))
plot(unrate,xlab=‘year’,ylab=‘unrate’,type=‘l’)
par(mfcol=c(2,2))
acf(unemp,lag=36)
pacf(unemp,lag=36)
acf(diff(unemp),lag=36)
pacf(diff(unemp),lag=36)
m1=arima(unemp,order=c(1,1,5),seasonal=list(order=c(1,0,1),period=12))
m1
tsdiag(m1,gof=36)
c1=c(NA,NA,NA,0,0,NA,NA,NA)
m1=arima(unemp,order=c(1,1,5),seasonal=list(order=c(1,0,1),period=12),fixed=c1)
m1
tsdiag(m1,gof=36)
Box.test(m1$residuals,lag=24,type=‘Ljung’)
Box.test(m1$residuals,lag=36,type=‘Ljung’)

mm=arima(unemp,order=c(0,1,0),seasonal=list(order=c(1,0,1),period=12))
mm
par(mfcol=c(2,1))
acf(mm$residuals,lag=24)
pacf(mm$residuals,lag=24)
mm1=arima(unemp,order=c(5,1,0),seasonal=list(order=c(1,0,1),period=12))
mm1
cc1=c(0,NA,NA,NA,NA,NA,NA)
mm1=arima(unemp,order=c(5,1,0),seasonal=list(order=c(1,0,1),period=12),fixed=cc1)
mm1
tsdiag(mm1,gof=36)
Backtesting
source(“backtest.R”)
pm1=backtest(m1,unemp,700,1,fixed=c1,inc.mean=F)
pmm1=backtest(mm1,unemp,700,1,fixed=cc1,inc.mean=F)
Use of initial jobless claims
da=read.table(“m-unrateic.txt”,header=T)
head(da)
unrate=da$rate
x=da[,5:9]/1000
ts.plot(ts(da$rate,start = c(1967,2),frequency = 12),ylab = “Unrate”)
ts.plot(ts(x$icm1,start = c(1967,2),frequency = 12),ylab = “ic-m”)
nm1=lm(unrate~icm1,data=x)
summary(nm1)
par(mfcol=c(2,1))
acf(nm1$residuals,lag=36)
pacf(nm1$residuals,lag=36)
nm1=arima(unrate,order=c(2,0,3),xreg=x[,5],seasonal=list(order=c(1,0,1),period=12))
nm1
nm1=arima(unrate,order=c(2,0,2),seasonal=list(order=c(1,0,1),period=12),xreg=x[,5])
nm1
tsdiag(nm1,gof=36)

nm2=lm(unrate~w1m1+w2m1+w3m1+w4m1,data=x)
summary(nm2)
nm2=lm(unrate~w1m1+w2m1+icm1,data=x)
summary(nm2)
nm2=lm(unrate~w1m1+w2m1,data=x)
summary(nm2)
acf(nm2$residuals,lag=36)
pacf(nm2$residuals,lag=36)
nm2=arima(unrate,order=c(2,0,2),seasonal=list(order=c(1,0,1),period=12),xreg=x[,1:2])
nm2
tsdiag(nm2,gof=36)

这个是老师给的代码，在这个代码的基础上帮我写一下这abc三个题，用R软件写，由于我的电脑文件命名有中文导致输出不了结果，可以帮我把这个数据全部写出来，不用用到文件地址吗，
文章回顾了过去一年的生活与工作经历，技术发展迅速，作者在项目管理方面积累了经验，计划继续学习新技术，提升英语口语和日语能力。展望未来，作者将专注于项目管理和技术积累。
扫一扫
6032
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
