兩個時間序列的相關性的自助置信區間

Question

我有兩個時間序列（ts_a和ts_b），我想獲得兩個序列之間相關性的自舉置信區間（在R中）。有關如何進行的任何建議？

感謝IND提前

Answer 1

這不是我的代碼，你可以找到原來here

require(boot) 
xy <- data.frame(cbind(ts_a,ts_b)) 

# Bootstrap the Pearson correlation coefficient

pearson <- function(d,i=c(1:dim(xy)[1]){ 
     d2 <- d[i,] 
     return(cor(d2$ts_a,d2$ts_b)) 
    } 
bootcorr <- boot(data=xy,statistic=pearson,R=999) 
bootcorr 
#Making a result table for output using 95% BCa CI 
resultTable <- summary(bootcorr) 
resultTable <- cbind(resultTable , CI95pc = NA) 
for (i in 1:dim(resultTable)[1]) { 
##Extracting CI from boot.ci output 
    a <- boot.ci(bootcorr,index = i, conf=.95, type = 'bca')$bca[4:5] 
##Adding the CI to your output table 
    resultTable $CI95pc[i] <- paste(round(a[1],3),"-",round(a[2],3),sep=" ") 
    rm(a) 
} 
print(resultTable)