[Rd] mean function on correlation matrices (PR#2540)

Peter Dalgaard BSA p.dalgaard@biostat.ku.dk
Tue Feb 11 12:24:03 2003


rs381@wbic.cam.ac.uk writes:

> Full_Name: Raymond Salvador
> Version: R 1.6.2
> OS: Windows ME
> Submission from: (NULL) (131.111.93.195)
> 
> 
> The mean function applied on individual components of several correlation
> matrices
> gives a wrong result (gives the first value instead of the mean).
> 
> Here there is a simple example
> 
> x1 <- rnorm(10,1,1)
> y1 <- rnorm(10,1,1)
> z1 <- cbind(x1,y1)
> w1 <- cor(z1)
> x2 <- rnorm(10,1,1)
> y2 <- rnorm(10,1,1)
> z2 <- cbind(x2,y2)
> w2 <- cor(z2)
> 
> w1[1,2]
> w2[1,2]
> mean(w1[1,2], w2[1,2])

Yes, and? 

> mean(2,6)
[1] 2

> mean(c(2,6))
[1] 4

The mean function just doesn't work the way you expected. Please read
the help page before reporting bugs.

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk)             FAX: (+45) 35327907