R编程：用于自回归时间序列模型的自适应套索

Question

我需要有人给我一些有关自动回归时间序列模型的自适应套索的R编程的提示：

Y_{t}=\phi_{1}Y_{t-1}+\epsilon_{t}, where \phi_{1}=0.5  and \epsilon_{t } \sim N(0,1) 

我知道在回归模型的情况下该怎么做。

我已经发布了我所做的回归分析。

library(glmnet)

set.seed(100)

beta_{1} = 2

beta_{2}=3

n=50

epsilon =rnorm(n,0,1) 

x = (rnorm(n,0,9))

y =(beta_{1 }+ beta_{2} * x + epsilon)

d=data.frame(x,y)

执行ols回归

ols<-lm(y\sim x, data=d)

best.ols.coef=coef(ols)

以10倍CV进行自适应LASSO

alasso1.cv <- cv.glmnet(x = cbind(0,x), y = y,
                    type.measure = "mse",
                    nfold = 10,
                    alpha = 1,
                    penalty.factor = 1 / abs(best.ols.coef),
                    keep = TRUE)

alasso1.cv$lambda.min

best.alasso.coef <- coef(alasso1.cv, s = alasso1.cv$lambda.min)

best.alasso.coef