问题
I'm currently downloading stock data using GetSymbols from the Quantmod package and calculating the daily stock returns, and then combining the data into a dataframe. I would like to do this for a very large set of stock symbols. See example below. In stead of doing this manually I would like to use a For Loop if possible or maybe use one of the apply functions, however I can not find the solution.
This is what I currently do:
Symbols<-c ("XOM","MSFT","JNJ","GE","CVX","WFC","PG","JPM","VZ","PFE","T","IBM","MRK","BAC","DIS","ORCL","PM","INTC","SLB")
length(Symbols)
#daily returns for selected stocks & SP500 Index
SP500<-as.xts(dailyReturn(na.omit(getSymbols("^GSPC",from=StartDate,auto.assign=FALSE))))
S1<-as.xts(dailyReturn(na.omit(getSymbols(Symbols[1],from=StartDate,auto.assign=FALSE))))
S2<-as.xts(dailyReturn(na.omit(getSymbols(Symbols[2],from=StartDate,auto.assign=FALSE))))
S3<-as.xts(dailyReturn(na.omit(getSymbols(Symbols[3],from=StartDate,auto.assign=FALSE))))
S4<-as.xts(dailyReturn(na.omit(getSymbols(Symbols[4],from=StartDate,auto.assign=FALSE))))
S5<-as.xts(dailyReturn(na.omit(getSymbols(Symbols[5],from=StartDate,auto.assign=FALSE))))
S6<-as.xts(dailyReturn(na.omit(getSymbols(Symbols[6],from=StartDate,auto.assign=FALSE))))
S7<-as.xts(dailyReturn(na.omit(getSymbols(Symbols[7],from=StartDate,auto.assign=FALSE))))
S8<-as.xts(dailyReturn(na.omit(getSymbols(Symbols[8],from=StartDate,auto.assign=FALSE))))
S9<-as.xts(dailyReturn(na.omit(getSymbols(Symbols[9],from=StartDate,auto.assign=FALSE))))
S10<-as.xts(dailyReturn(na.omit(getSymbols(Symbols[10],from=StartDate,auto.assign=FALSE))))
....
S20<-as.xts(dailyReturn(na.omit(getSymbols(Symbols[20],from=StartDate,auto.assign=FALSE))))
SPportD<-cbind(SP500,S1,S2,S3,S4,S5,S6,S7,S8,S9,S10,S11,S12,S13,S14,S15,S16,S17,S18,S19,S20)
names(SPportD)[1:(length(Symbols)+1)]<-c("SP500",Symbols)
SPportD.df<-data.frame(index(SPportD),coredata(SPportD),stringsAsFactors=FALSE)
names(SPportD.df)[1:(length(Symbols)+2)]<-c(class(StartDate),"SP500",Symbols)
Any suggestions?
Thanks!
回答1:
lapply
is your friend:
Stocks = lapply(Symbols, function(sym) {
dailyReturn(na.omit(getSymbols(sym, from=StartDate, auto.assign=FALSE)))
})
Then to merge:
do.call(merge, Stocks)
Similar application for the other assignments
回答2:
dailyReturn
uses close prices, so I would recommend you either use a different function (e.g. TTR::ROC
on the Adjusted column), or adjust the close prices for dividends/splits (using adjustOHLC
) before calling dailyReturn
.
library(quantmod)
Symbols <- c("XOM","MSFT","JNJ","GE","CVX","WFC","PG","JPM","VZ","PFE",
"T","IBM","MRK","BAC","DIS","ORCL","PM","INTC","SLB")
# create environment to load data into
Data <- new.env()
getSymbols(c("^GSPC",Symbols), from="2007-01-01", env=Data)
# calculate returns, merge, and create data.frame (eapply loops over all
# objects in an environment, applies a function, and returns a list)
Returns <- eapply(Data, function(s) ROC(Ad(s), type="discrete"))
ReturnsDF <- as.data.frame(do.call(merge, Returns))
# adjust column names are re-order columns
colnames(ReturnsDF) <- gsub(".Adjusted","",colnames(ReturnsDF))
ReturnsDF <- ReturnsDF[,c("GSPC",Symbols)]
回答3:
Packages are quantmod
for data download and PerformanceAnalytics
for analysis/plotting.
care must be taken with time series date alignment
Code
require(quantmod)
require(PerformanceAnalytics)
Symbols<-c ("XOM","MSFT","JNJ","GE","CVX","WFC","PG","JPM","VZ","PFE","T","IBM","MRK","BAC","DIS","ORCL","PM","INTC","SLB")
length(Symbols)
#Set start date
start_date=as.Date("2014-01-01")
#Create New environment to contain stock price data
dataEnv<-new.env()
#download data
getSymbols(Symbols,env=dataEnv,from=start_date)
#You have 19 symbols, the time series data for all the symbols might not be aligned
#Load Systematic investor toolbox for helpful functions
setInternet2(TRUE)
con = gzcon(url('https://github.com/systematicinvestor/SIT/raw/master/sit.gz', 'rb'))
source(con)
close(con)
#helper function for extracting Closing price of getsymbols output and for date alignment
bt.prep(dataEnv,align='remove.na')
#Now all your time series are correctly aligned
#prices data
stock_prices = dataEnv$prices
head(stock_prices[,1:3])
# head(stock_prices[,1:3])
# BAC CVX DIS
#2014-01-02 16.10 124.14 76.27
#2014-01-03 16.41 124.35 76.11
#2014-01-06 16.66 124.02 75.82
#2014-01-07 16.50 125.07 76.34
#2014-01-08 16.58 123.29 75.22
#2014-01-09 16.83 123.29 74.90
#calculate returns
stock_returns = Return.calculate(stock_prices, method = c("discrete"))
head(stock_returns[,1:3])
# head(stock_returns[,1:3])
# BAC CVX DIS
#2014-01-02 NA NA NA
#2014-01-03 0.019254658 0.001691638 -0.002097810
#2014-01-06 0.015234613 -0.002653800 -0.003810275
#2014-01-07 -0.009603842 0.008466376 0.006858349
#2014-01-08 0.004848485 -0.014232030 -0.014671208
#2014-01-09 0.015078408 0.000000000 -0.004254188
#Plot Performance for first three stocks
charts.PerformanceSummary(stock_returns[,1:3],main='Stock Absolute Performance',legend.loc="bottomright")
Performance Chart:
![](https://www.eimg.top/images/2020/03/23/128e660edee95903f5d555b76e0d7167.png)
来源:https://stackoverflow.com/questions/24377590/getsymbols-downloading-data-for-multiple-symbols-and-calculate-returns