[R] NA's?

[Bill.Venables@csiro.au]

Your message doesn't help us very much.  You haven't said what kind of
calculation it is you want to do, and that certainly matters.  For
example, for some kinds of computations the solution you started below
would work fine:

> M <- matrix(1:16, 4, 4)
> is.na(diag(M)) <- TRUE
> M
     [,1] [,2] [,3] [,4]
[1,]   NA    5    9   13
[2,]    2   NA   10   14
[3,]    3    7   NA   15
[4,]    4    8   12   NA
> rowSums(M, na.rm = TRUE)
[1] 27 26 25 24
> colSums(M, na.rm = TRUE)
[1]  9 20 31 42

You can also use apply( ) with functions that will accept missing values
(and ignore them) for computations on either the rows or the columns.

Hoping for a general mechanism that would somehow signal the diagonal
values as values to be "ignored" in a general way is not a possibility.
Just as a curiosity, what were you hoping that na.omit(M) would do?

V.

-----Original Message-----
From: r-help-bounces at stat.math.ethz.ch
[mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Asha Jayanthi
Sent: Thursday, 31 March 2005 11:22 AM
To: r-help at stat.math.ethz.ch
Subject: [R] NA's?


I have a large matrix of data .

The size of the matrix ranges from 100 x 100 to 1000 x 1000

Now i have to do computations on that. And should not consider the
diagonal 
elements.

I tried setting diag(M) = NA  and M = na.omit(M).

But this omits all the rows. I only want to omit that diagonal elements
only 
but consider the whole row.

diag(M) = 0 seems like a good option but this will affect my result.

How to proceed with this. How to just ignore some specific values. what
if i 
want to consider only the upper / lower triangular matrix

Asha


http://adfarm.mediaplex.com/ad/ck/4686-26272-10936-31?ck=RegSell Start
your 
business.

______________________________________________
R-help at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html