problem with table()

on 01/26/2009 12:23 PM Dominik Hattrup wrote:
  

Hey everyone,

I am looking for the easiest way to get this table

# Table2
Year / 2000 / 2002 / 2004
Julia / 3 / 4 / 1
Peter / 1 / 2 / 4
...   /   ...   /   ...   /   ...

out of this one? 

# Table1
name   /   year   /   cases
Julia   /   2000   /   1
Julia   /   2000   /   2  
Julia   /   2002   /   4
Peter   /   2000   /   1
Julia   /   2004   /   1
Peter   /   2004   /   2
Peter   /   2002   /   2
Peter   /   2004   /   2
...   /   ...   /   ...

Code for table1:
name <- c('Julia','Julia','Julia','Peter','Julia','Peter','Peter','Peter')
year <- c(2000,2000,2002,2000,2004,2004,2002,2004)
cases <- c(1,2,4,1,1,2,2,2)
table1 <- data.frame(name,year,cases)

Thanks! Dominik
    

table() generates frequencies from individual values, not from already
tabulated data.

In this case, you can use xtabs():

  

xtabs(cases ~ name + year, data = table1)
    

       year
name    2000 2002 2004
  Julia    3    4    1
  Peter    1    2    4


See ?xtabs

An alternative would be to use tapply():

  

with(table1, tapply(cases, list(name = name, year = year), sum))
    

       year
name    2000 2002 2004
  Julia    3    4    1
  Peter    1    2    4


See ?tapply

HTH,

Marc Schwartz