Create a panel from two data frames

Question

I have two data frames -- one with stock closing prices arranged by date (rows) and ticker symbol (columns):

> head(data.stocks)
        date     A   AAPL ABAT    AB    ABV
1 2010-10-04 32.59 278.64 3.65 26.42 125.89
2 2010-10-05 33.04 288.94 3.66 27.10 129.05
3 2010-10-06 32.67 289.19 3.59 26.99 129.90
4 2010-10-07 33.20 289.22 3.66 27.04 129.94
5 2010-10-08 33.80 294.07 3.84 26.76 132.66
6 2010-10-11 33.75 295.36 3.87 26.95 133.37

The other data frame has open interest (oi) arranged by one row per ticker symbol per date:

> head(data.oi)
        date symbol  oi close
1 2010-10-04      A   6     0
2 2010-10-04     AA 104     0
3 2010-10-04   AAPL 940     0
4 2010-10-04     AB   0     0
5 2010-10-04   ABAT   0     0
6 2010-10-04    ABB   0     0

I would like to add another column to this second data frame so that I end up with a "panel" data frame with one ticker symbol per date per row with oi and closing price.

I thought this loop would work, but I get the following error:

> for (i in seq(length(data.oi$date))) {
+ row <- which(data.stocks$date == data.oi$date[i])
+ col <- which(colnames(data.stocks) == data.oi$symbol[i])
+ data.oi$close[i] <- data.stocks[row, col]
+ }
Error in data.oi$close[i] <- data.stocks[row, col] : 
  replacement has length zero

FWIW, it stores the first value in data.oi, then fails:

> head(data.oi)
        date symbol  oi close
1 2010-10-04      A   6 32.59
2 2010-10-04     AA 104  0.00
3 2010-10-04   AAPL 940  0.00
4 2010-10-04     AB   0  0.00
5 2010-10-04   ABAT   0  0.00
6 2010-10-04    ABB   0  0.00

Thanks! I feel like reshape or aggregate could do it, but I can't figure out how to merge these two.

Answer 1

hi, melt in reshape2 package and merge would be useful:

library(reshape2)
merge(subset(data.oi,select=-close), 
      melt(d,id.vars="date",variable.name="symbol",value.name="close"))