Restructure Data in R

Question

Hi Everyone,

I am just starting to get beyond the basics in R and have come to a point where I need some help. I want to restructure some data. Here is what a sample dataframe may look like:

ID  Sex Res Contact
1   M   MA  ABR
1   M   MA  CON
1   M   MA  WWF
2   F   FL  WIT
2   F   FL  CON
3   X   GA  XYZ

I want the data to look like:

ID  SEX Res ABR CON WWF WIT XYZ
1   M   MA  1   1   1   0   0
2   F   FL  0   1   0   1   0
3   X   GA  0   0   0   0   1

What are my options? How would I do this in R? R is so powerful, I figure this is probably a breeze!

In short, I am looking to keep the values of the CONT column and use them as column names in the restructred data frame. I want to hold a variable set of columns constant (in th example above, I held ID, Sex, and Res constant). Also, is it possible to control the values in the restructured data? I may want to keep the data as binary, and sometimes, I may want to have the value be the count of times each contact value exists for each ID.

Any help will be very much appreciated,

Brock

Answer 1

The reshape package is what you want. Documentation here: http://had.co.nz/reshape/. Not to toot my own horn, but I've also written up some notes on reshape's use here: http://www.ling.upenn.edu/~joseff/rstudy/summer2010_reshape.html

For your purpose, this code should work

library(reshape)
data$value <- 1
cast(data, ID + Sex + Res ~ Contact, fun = "length")

Answer 2

model.matrix works great (this was asked recently, and gappy had this good answer):

> model.matrix(~ factor(d$Contact) -1)
  factor(d$Contact)ABR factor(d$Contact)CON factor(d$Contact)WIT factor(d$Contact)WWF factor(d$Contact)XYZ
1                    1                    0                    0                    0                    0
2                    0                    1                    0                    0                    0
3                    0                    0                    0                    1                    0
4                    0                    0                    1                    0                    0
5                    0                    1                    0                    0                    0
6                    0                    0                    0                    0                    1
attr(,"assign")
[1] 1 1 1 1 1
attr(,"contrasts")
attr(,"contrasts")$`factor(d$Contact)`
[1] "contr.treatment"