Using get() to reference columns in quantmod arrays with R?

Question

I'm new to R, using the quantmod() package for a project. The following block of code works:

require(quantmod) 
stocks<-c("MMM", "MSFT", "BP")
for(i in 1:length(stocks)){
getSymbols(stocks[i], from= "2013-07-01")
s<-get(stocks[i])
dr<-dailyReturn(s)
print(paste(dr))
}

However, I need to reference specific columns to calculate some technical analysis indicators from the TTR package. For example:

open<-MMM$MMM.Open
RSI(open, n=14)

When I check:

identical(s, BP) #TRUE

And this works:

BP$BP.Open

However, this does not work:

s$s.Open #NULL

To provide adequate context, my goal is to go through a vector of stocks, check for a ]condition, then calculate some technical analysis and time series figures for that day and copy it into an ARFF file for use as training examples for a machine learning environment (Weka). Thanks.

Answer 1

It's generally easier to use the extractor functions Op et al. See ?OHLC.Transformations. Also, if you only have one symbol, you can use auto.assign=FALSE in your call to getSymbols to avoid the get call all together.

s <- getSymbols("BP", auto.assign=FALSE)

If you have multiple symbols, it's easier to store them in an environment and then loop over them with eapply:

e <- new.env()
getSymbols(stocks, env=e)
dr <- eapply(e, dailyReturn)

You can also apply TTR functions to each symbol this way.

rsi <- eapply(e, function(x) RSI(Op(x), n=14))

And you can use do.call with cbind to put them into a single object.

rsi_all <- do.call(cbind, rsi)