Question

我正在尝试将脚本放入For-Loop中，以便计算多只股票的平均每日收益。

但是当我设置前循环时，出现以下错误：

getSymbols（Symbols = stock，src =“ yahoo”，from =“ 2005-01-01”，错误：必须对多个符号请求使用auto.assign = TRUE

因此，我删除了env = NULL并将其替换为auto.assign=TRUE，但是当我运行脚本时，结果是NaN。

任何建议将不胜感激。谢谢。

install.packages('quantmod')
library(quantmod)

stock <- c("AAPL") 
{
stock.xts <- getSymbols(Symbols = stock, src = "yahoo", 
                       from = "2005-01-01",
                       to = "2019-09-10", 
                       env = NULL)

stock.xts <- as.data.frame(stock.xts)

N <- nrow(stock.xts)
todays.price <- stock.xts[2 : N,4]
yesterdays.price <- stock.xts[1 : N-1,4]
stock_dailyreturn <- (todays.price - yesterdays.price)/yesterdays.price
stock.xts$daily.return <- c(NA,(todays.price - yesterdays.price)/yesterdays.price)
print(mean(stock_dailyreturn)) 
}

使用前循环：

stock <- c("AAPL", "MSFT", "CRM", "ORCL", "NFLX", "GOOG")

for(i in 1:length(stock)){
  stock.xts <- getSymbols(Symbols = stock, src = "yahoo", 
                          from = "2005-01-01",
                          to = "2019-09-10",
                          auto.assign = TRUE)
  stock.xts <- as.data.frame(stock.xts)

  N <- nrow(stock.xts)
  todays.price <- as.numeric(stock.xts[2 : N,4])
  yesterdays.price <- as.numeric(stock.xts[1 : N-1,4])
  stock_dailyreturn <- (todays.price - yesterdays.price)/yesterdays.price
  stock.xts$daily.return <- c(NA,(todays.price - yesterdays.price)/yesterdays.price)
  print(mean(stock.xts$daily.return))
}

Answer 1

您的整个问题都在代码的最后一行mean(stock.xts$daily.return)中，以及紧接其后的一行中。如果平均的向量包含mean值NA，则NaN默认返回NA或stock.xts$daily.return <- c(NA, ...。可以通过在均值内设置na.rm = TRUE来更改。

但是，可以进一步降低循环的复杂度。首先，您将在每次迭代中导入符号。这是不必要的。您也可以使用diff和lag函数，通过使用diff(...)/lag(...)计算收益。

例如，您可以使用类似于以下示例的内容：

library("quantmod")
stock <- c("AAPL", "MSFT", "CRM", "ORCL", "NFLX", "GOOG")
#Create a place to store the symbols
stockenv <- new.env()
stuff <- getSymbols(Symbols = stock, src = "yahoo", 
                   from = "2005-01-01",
                   to = "2019-09-10",
                   env = stockenv) #<=== note here, i import into the environment
ls(stockenv) #<== check that all the symbols were imported into the environemnt.
#[1] "AAPL" "CRM"  "GOOG" "MSFT" "NFLX" "ORCL"
priceChange <- vector("list", length(ls(stockenv)))
meanPrice <- vector("list", length(ls(stockenv)))
#overwrite names
names(priceChange) <- names(meanPrice) <- ls(stockenv)
for(i in ls(stockenv)){
    priceChange[[i]] <- diff(stockenv[[i]][,4]) / lag(stockenv[[i]][,4])
    meanPrice[[i]] <- mean(priceChange[[i]], na.rm = TRUE)
}
simplify2array(meanPrice)
#        AAPL          CRM         GOOG         MSFT         NFLX         ORCL 
#0.0012548717 0.0013114559 0.0008412977 0.0005783129 0.0019348552 0.0005236143

Answer 2

您超级亲密！尝试以下操作：

更改您的for循环以在每次迭代中保留单个值（请参见下面的for循环）
更新getSymbols函数的用法，使其正确自动分配（否则默认为NULL-请查阅文档以获取更多信息）
使用占位符矢量进行getSymbols函数创建对象（我现在称呼它为“ hello

祝你好运！

stock <- c("AAPL","MSFT","CRM")

hello <- c()

for(value in stock){

stock.xts <- getSymbols(Symbols = value, src = "yahoo",
                    from = "2005-01-01",
                    to = "2019-09-10",
                    env = hello,
                    auto.assign=TRUE)
stock.xts <- as.data.frame(stock.xts)

N <- nrow(stock.xts)
todays.price <- stock.xts[2 : N,4]
yesterdays.price <- stock.xts[1 : N-1,4]
stock_dailyreturn <- (todays.price - yesterdays.price)/yesterdays.price
stock.xts$daily.return <- c(NA,(todays.price - yesterdays.price)/yesterdays.price)
print(mean(stock_dailyreturn))

}

使用For-Loop计算平均每日收益

2 个答案: