R: How to reshape a table into vectors

Question

I'm working through the examples in Kruschke's Doing Bayesian Data Analysis and need a bit of help understanding how to get data into the format that his code examples require. In chapter 22 he has a table like this

         Blue  Brown  Green Hazel
Black    20    68     5     15
Blond    94    7      16    10
Brunette 84    119    29    54
Red      17    26     14    14

I'm comfortable with inputting the table into R by entering it into a spreadsheet and using read.table("clipboard", header=T, sep="\t") or typing it into R like this

con.table2 <- matrix(c(20,68,5,15,94,7,16,10,84,119,29,54,17,26,14,14),nrow=4,byrow=TRUE)
dimnames(con.table2) <- list(c("Black","Blond","Brunette","Red"),c("Blue","Brown","Green","Hazel"))

But in his code, he presents this table like so, ready for analysis (full code is here http://www.indiana.edu/~kruschke/DoingBayesianDataAnalysis/Programs/PoissonExponentialJagsSTZ.R)

Freq = c(68,119,26,7,20,84,17,94,15,54,14,10,5,29,14,16)
Eye = c("Brown","Brown","Brown","Brown","Blue","Blue","Blue","Blue","Hazel" # runs off the page of his book
Hair = c("Black","Brunette","Red","Blond","Black","Brunette","Red","Blond","Black" # runs off the page of his book

It looks like the table has been converted into three vectors. What's the most efficient way to do this? I'd like to replace his data with my own, so it would be great to learn how to transform the data into the format needed for this analysis.

Answer 1

For this, I'd use melt() in the reshape2 package:

library(reshape2)
df <- melt(con.table2, varnames=c("Hair", "Eye"), value.name="Freq")

# df is a data frame, a list from which you can easily extract the
# component vectors "Hair", "Eye", and "Freq. 
# Try, for example:
str(df)
df$Hair    

Answer 2

There is a method in base R for converting objects of class "table" to data.frames. The reason that it does not succeed with your matrix is that you didn't tell R that it was a table. Once you do so the method succeeds:

class(con.table2) <- "table"
as.data.frame(con.table2)
#-----------------------
       Var1  Var2 Freq
1     Black  Blue   20
2     Blond  Blue   94
3  Brunette  Blue   84
4       Red  Blue   17
5     Black Brown   68
6     Blond Brown    7
7  Brunette Brown  119
8       Red Brown   26
9     Black Green    5
10    Blond Green   16
11 Brunette Green   29
12      Red Green   14
13    Black Hazel   15
14    Blond Hazel   10
15 Brunette Hazel   54
16      Red Hazel   14

The "table" class in R is expected to be a contingency table (just as you have constructed), i.e, one with counts in cells. In this case you could have had fractional values in there and there would be no problems but some methods that were expecting the values to be integer might choke on non-integer values.