R help - Sep 2010 - class "matrix" lost when extracting a row from matrix

Good morning experts!

situation:

class(myMatrix)="matrix"
class(myMatrix[(1:2),]))="matrix"
class(myMatrix[1,])= "character"

consequences are far reaching as, for instance colnames(myMatrix[(1:2),]) != 
colnames(myMatrix[1,]) or names(myMatrix[(1:2),]) !=  names(myMatrix[1,])

My question: 
1. How can the coercion to character be avoided and the attributes of the
original object myMatrix of class "matrix" been preserved when slicing
out
one row of the matrix?
2. Is there a reason why there is a "loss of generality" in that sense
that
a matrix of dimension (1xn) is treated differently than a matrix of
dimension (mxn); m,n>1 ? in other words is there a reason why class vector
is needed at all?

Many thanks for your help!
Holger
-- 
View this message in context:
http://r.789695.n4.nabble.com/class-matrix-lost-when-extracting-a-row-from-matrix-tp2545719p2545719.html
Sent from the R help mailing list archive at Nabble.com.

On 09/19/2010 10:02 AM, hw at wiseadvice.eu wrote:
>
> Good morning experts!
>
> situation:
>
> class(myMatrix)="matrix"
> class(myMatrix[(1:2),]))="matrix"
> class(myMatrix[1,])= "character"
>
> consequences are far reaching as, for instance colnames(myMatrix[(1:2),])
!> colnames(myMatrix[1,]) or names(myMatrix[(1:2),]) !=  names(myMatrix[1,])
>
> My question:
> 1. How can the coercion to character be avoided and the attributes of the
> original object myMatrix of class "matrix" been preserved when
slicing out
> one row of the matrix?
> 2. Is there a reason why there is a "loss of generality" in that
sense that
> a matrix of dimension (1xn) is treated differently than a matrix of
> dimension (mxn); m,n>1 ? in other words is there a reason why class
vector
> is needed at all?
m <- matrix(1:4, 2)

m[1,, drop = FALSE]
#      [,1] [,2]
# [1,]    1    3

m[, 1, drop = FALSE]
#      [,1]
# [1,]    1
# [2,]    2

See ?Extract

HTH,
Tobias