[R] calculating column difference in a matrix

[arun]

Hi,
Try this:
a1<-a

a1[rowSums(is.na(a1))==1][ is.na(a1[rowSums(is.na(a1))==1])]<-0
library(matrixStats) c1<- cbind(a,rowDiffs(a1))
 c1
#     [,1] [,2] [,3]
#[1,]    2    5    3
#[2,]    3    8    5
#[3,]    4   NA   -4
#[4,]   NA    8    8
#[5,]   NA   NA   NA

identical(c,c1)
#[1] TRUE

A.K.



----- Original Message -----
From: Pedro Mardones <mardones.p at gmail.com>
To: R-help at r-project.org
Cc: 
Sent: Wednesday, March 13, 2013 5:43 PM
Subject: [R] calculating column difference in a matrix

Dear R users;

Consider the following toy example:

a <- matrix(c(2,3,4,NA,NA,5,8,NA,8,NA), 5, 2)
b <- cbind(a,apply(a, 1, diff, na.rm = TRUE))

What I would like be able to get is:

c <- matrix(c(2,3,4,NA,NA,5,8,NA,8,NA,3,5,-4,8,NA), 5, 3)

i.e., for each row if both values (column 1 and 2) are NA then the
difference must return NA, but if any of those two values is different
from NA (!is.na), I would like to actually perform the difference
(something like assigning 0 instead to NA to the cell).

I'd appreciate any hint or comment

Best,

Pedro

BTW: the reason of this is that I'm trying to build an error checking
procedure for a big dataset I have

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.