weighting means

What about:

?weighted.mean
?tapply
Q8 <- rnorm(100)
Faculty <- factor(c(rep(1,50), rep(2,50)))
w <- 1:100
tapply(Q8, Faculty, weighted.mean, w)

         1          2 
-0.0976302 -0.1457895 


Kjetil Halvorsen

Tim Wilson wrote:

Hi everyone,

I've got a dataframe called 'faculty'. I want to do a weighted mean on
the column called 'Q8' weighted by the contents of column 'CETP'. In
addition, I need to operate on the result of splitting 'faculty'
according to the contents of a column 'FACULTY'. For example

lapply(split(faculty$Q8, faculty$FACULTY), mean)

$"1"
[1] 2.2

$"2"
[1] 1.888889

gives the mean of 'Q8' after splitting by 'FACULTY'. When I try this
with weighted.mean, it doesn't work because I don't know how to specify
the result of the split as the argument for weighted.mean. For example,

lapply(split(faculty$Q8, faculty$FACULTY), weighted.mean)

$"1"
[1] 2.2

$"2"
[1] 1.888889

gives the same results as above because I haven't specified the vector
to use for weighting. The following doesn't work at all for the obvious
reason that weighted.mean is not operating on the result of the split.

lapply(split(faculty$Q8, faculty$FACULTY), weighted.mean(faculty$Q8,

faculty$CETP))
Error in match.fun(FUN) : not function, character, or symbol:
"weighted.mean(faculty$Q8, faculty$CETP)"

Any ideas?

-Tim

--
Tim Wilson      |   Visit Sibley online:   | Check out:
Henry Sibley HS |  http://www.isd197.org   | http://www.zope.com
W. St. Paul, MN |                          | http://slashdot.org
wilson at visi.com |  <dtml-var pithy_quote>  | http://linux.com