train <- read.csv("train.csv", header=TRUE) as.numeric(train$ppi) #as.Date(train$date, "%m/%d/%Y") myts=ts(train[,-1], start=c(2003,1), frequ