Question

Suppose I have a two column matrix. How do I pack the columns into a pair/tuple so that they can be assigned to a one column matrix?

> A = matrix(NA,nrow=5,ncol=1)
> B = matrix(runif(10),ncol=2)
> A
     [,1]
[1,]   NA
[2,]   NA
[3,]   NA
[4,]   NA
[5,]   NA

> B
          [,1]      [,2]
[1,] 0.1886287 0.6995596
[2,] 0.1576875 0.9792369
[3,] 0.9056386 0.1640904
[4,] 0.9125812 0.7003167
[5,] 0.9327778 0.8149431

> A[,1] = B   # need this to work

I have a n-col matrix of prices, a column for each stock. I am trying to compute a moving MACD statistic for each stock. I am using a n-col MACD matrix to contain the results. When I feed a one col of prices to MACD function (from package TTR), it returns a 2-col matrix of signal and macd, so I need to way to contain this statistic within the same dimension.

Answer 1

You can do that with lists.

> matrix(apply(B,1,list))
     [,1]  
[1,] List,1
[2,] List,1
[3,] List,1
[4,] List,1
[5,] List,1

That said, this is a very un-R-like way to do things and is probably more trouble than it’s worth. If you describe what you’re actually trying to do, someone could show you a more appropriate approach.

UPDATE:
Based on the updated question, the code below will put the macd and signal in a 2n matrix. You may want to write a more elaborate function (e.g. one that identifies the macd and signal columns with their respective instruments).

If you want the macd and signal columns in separate matrices, you could just grep the columns from the out object.

library(quantmod)
getSymbols("SPY;IWM;QQQ")
Data <- Cl(merge(SPY,IWM,QQQ))
out <- do.call(merge, lapply(1:NCOL(Data), function(i) MACD(Data[,i])))