I download historical prices with quantmod's getSymbols function for multiple tickers and convert them into either a list or a multivariate XTS with the following code:
library(quantmod)
myenv <- new.env()
tickers <- c("^GSPC", "AAPL", "MSFT", "GOOG", "^RUT")
getSymbols(tickers, env=myenv)
ll <- eapply(myenv, function(x) x) # Convert to list
ts <- do.call(merge, (eapply(myenv, Ad))) # Convert to multivariate XTS and extract only the adjusted price
The problem I have with this approach is that the order of the tickers in the list and the XTS are not the same as what I specified in tickers:
> names(ll)
[1] "AAPL" "GSPC" "GOOG" "RUT" "MSFT"
> names(ts)
[1] "AAPL.Adjusted" "GSPC.Adjusted" "GOOG.Adjusted" "RUT.Adjusted"
[5] "MSFT.Adjusted"
I think this is because eapply performs the operations in a random order, as explained in the help pages of eapply:
Note that the order of the components is arbitrary for hashed environments.
How can I perform the same operations above but have an output that is in the same order as specified in my tickers vector? I.e. first item of the list / first column of XTS should correspond to first element of the tickers vector and so on.
You could just subset the results of eapply into the order you want.
library(quantmod)
tickers <- c("^GSPC", "AAPL", "MSFT", "GOOG", "^RUT")
myenv <- new.env()
symnames <- getSymbols(tickers, env=myenv) # getSymbols returns adjusted names
ts <- do.call(merge, eapply(myenv, Ad)[symnames])