将 EGARCH 风味添加到 fGARCH 风味模型的循环中

Question

假设我有以下“for”循环。具体来说，我构建的这段代码首先计算所有指定风格的 ARCH(1) 模型的 AIC，然后计算所述风格的 GARCH(1,1) 模型的 AIC。

library(rugarch)
bchain_2012_logreturns=diff(log(prices))
aic_2012=matrix(NA,14,1)
garch_flavor=c("GARCH","AVGARCH","TGARCH","GJRGARCH","NGARCH","NAGARCH","APARCH")
k=1
for (i in 0:1){
for (j in garch_flavor){
model_2012=ugarchspec(variance.model = list(model="fGARCH", submodel = j, garchOrder = c(1, i)),mean.model = list(armaOrder = c(0, 0)))
modelfit_2012=ugarchfit(spec=model_2012, data=bchain_2012_logreturns, solver="hybrid") 
aic_2012[k,]=infocriteria(modelfit_2012)[1]
k=k+1
}
}

如 R CRAN 上的“rugarch”包小插图中所述（参见链接：https://cran.r-project.org/web/packages/rugarch/vignette/Introduction_to_the_rugarch_package.pdf），“eGARCH”模型风格不包括在 GARCH 系列（即“fGARCH”）子模型中。使用“rugarch”包估计 eGARCH 模型所需的命令如下：

model_2012=ugarchspec(variance.model = list(model="eGARCH",garchOrder = c(1, 1)),mean.model = list(armaOrder = c(0, 0)))
modelfit_2012=ugarchfit(spec=model_2012, data=bchain_2012_logreturns, solver="hybrid") 

我需要同时做的是：

1) 将 eGARCH 估计的最后一个命令集成到循环中，这样我就只有一个命令来计算所有 8 个 GARCH 模型风格的 AIC。

2) 将“统一”循环设置为仅在 (1,1) 阶的 eGARCH 上进行迭代。

我是 R 编程的新手，这对我来说是个不小的问题。

提前谢谢你。

Answer 1

使用lapply 和你可以做的条件：

#Combine all model names into one vector
model_names =c("eGARCH","GARCH","AVGARCH","TGARCH","GJRGARCH","NGARCH","NAGARCH","APARCH")

#If model name equals "eGARCH" use formula1 else formula2, compute for order (1,1) only
#for other models compute for orders (1,0) and (1,1)
#calculate AIC for each model and rbind to form a combined data.frame


AIC_2012 = do.call(rbind,lapply(model_names,function( x ) {

   if ( x=="eGARCH" ){

   model_2012_0 = ugarchspec(variance.model = list(model= x,garchOrder = c(1, 1)),mean.model = list(armaOrder = c(0, 0)))
   modelfit_2012_0 =ugarchfit(spec=model_2012_0   , data=bchain_2012_logreturns, solver="hybrid")

   DF = data.frame(model_name = x,order= "1,1" ,AIC = infocriteria(modelfit_2012_0)[1])

   }else {

   model_2012_0 = ugarchspec(variance.model = list(model="fGARCH", submodel = x, garchOrder = c(1, 0)),mean.model = list(armaOrder = c(0, 0)))
   model_2012_1 = ugarchspec(variance.model = list(model="fGARCH", submodel = x, garchOrder = c(1, 1)),mean.model = list(armaOrder = c(0, 0)))

   modelfit_2012_0 =ugarchfit(spec=model_2012_0   , data=bchain_2012_logreturns, solver="hybrid")
   modelfit_2012_1 =ugarchfit(spec=model_2012_1, data=bchain_2012_logreturns, solver="hybrid")

   DF_0 = data.frame(model_name = x, order= "1,0" , AIC = infocriteria(modelfit_2012_0)[1])
   DF_1 = data.frame(model_name = x, order= "1,1" , AIC = infocriteria(modelfit_2012_1)[1])

   DF = rbind(DF_0,DF_1)

   }

   return(DF)

}))