R help - Jul 2006 - Vector extracted from a matrix. How can I specify dimensions in as.matrix?

Transpose vector extracted from a matrix

Hello,

I am doing a recursive analysis that uses every line (vector) of a matrix in 
a loop. In the model, I need to transpose those vectors that are extracted 
from a matrix.

Using simple vectors (no matrix involved) the transpose function works fine:

simplevector <-matrix(1:3,3,1)
tsimplevector <-t(simplevector) #transposed
dim(simplevector) #3x1 matrix (vector)
dim(tsimplevector)#1x3

PROBLEM: However, when the vector is extracted from a matrix, its dimension 
is NULL.  In this case the transposed dimension is correct, but you'll see 
the next example, that if a row was extracted, the dimension would be wrong:

initialmatrix <- matrix(1:9,3,3)
extractedvector <-initialmatrix[,1] # extract first columm as vector
textractedvector <-t(extractedvector)#transposed
dim(extractedvector) #NULL!!!! <- not working
dim(textractedvector)#1x3 as expected

I have tried to transform the extracted vector as.vector and as.matrix. 
as.vector does not give the what I want (still NULL) and as.matrix can't 
specify the number of row and columns so both vectors are vertical.

In this example, I extract a column and a row. Notice that both 
as.matrix(vector) have the same dimension. Notice that both transposed 
vectors have 1x3 dimensions.

initialmatrix <- matrix(1:9,3,3)
extractedvector3x1 <-initialmatrix[,1]  # extract first columm as vector
extractedvector1x3 <-initialmatrix[1,]  # extract first row as vector

#Both are NULL
dim(extractedvector3x1) #NULL!!!!
dim(extractedvector1x3) #NULL!!!!

#transposed dimensions both 1x3
dim(t(extractedvector3x1)) #1x3
dim(t(extractedvector1x3)) #1x3 <- not 3x1

#as.vector: still NULL
dim(as.vector(extractedvector3x1)) #NULL!!!!
dim(as.vector(extractedvector1x3)) #NULL!!!!

#as.matrix: Both dim at 3x1
dim(as.matrix(extractedvector3x1)) #3x1
dim(as.matrix(extractedvector1x3)) #3x1 <- Problem: not 1x3

#as.matrix transposed: Both dim at 1x3
dim(as.matrix(t(extractedvector3x1))) #1x3
dim(as.matrix(t(extractedvector1x3))) #1x3 <- Problem: not 3x1

How can I get correct dimensions? Is there a way to specify dimensions in 
as.matrix?

Neuro

Use the notation x[ , 1, drop = FALSE]
See ?"["

On 7/27/06, Neuro LeSuperH?ros <neuro3000 at hotmail.com>
wrote:
> Transpose vector extracted from a matrix
>
> Hello,
>
> I am doing a recursive analysis that uses every line (vector) of a matrix
in
> a loop. In the model, I need to transpose those vectors that are extracted
> from a matrix.
>
> Using simple vectors (no matrix involved) the transpose function works
fine:
>
> simplevector <-matrix(1:3,3,1)
> tsimplevector <-t(simplevector) #transposed
> dim(simplevector) #3x1 matrix (vector)
> dim(tsimplevector)#1x3
>
> PROBLEM: However, when the vector is extracted from a matrix, its dimension
> is NULL.  In this case the transposed dimension is correct, but you'll
see
> the next example, that if a row was extracted, the dimension would be
wrong:
>
> initialmatrix <- matrix(1:9,3,3)
> extractedvector <-initialmatrix[,1] # extract first columm as vector
> textractedvector <-t(extractedvector)#transposed
> dim(extractedvector) #NULL!!!! <- not working
> dim(textractedvector)#1x3 as expected
>
> I have tried to transform the extracted vector as.vector and as.matrix.
> as.vector does not give the what I want (still NULL) and as.matrix
can't
> specify the number of row and columns so both vectors are vertical.
>
> In this example, I extract a column and a row. Notice that both
> as.matrix(vector) have the same dimension. Notice that both transposed
> vectors have 1x3 dimensions.
>
> initialmatrix <- matrix(1:9,3,3)
> extractedvector3x1 <-initialmatrix[,1]  # extract first columm as vector
> extractedvector1x3 <-initialmatrix[1,]  # extract first row as vector
>
> #Both are NULL
> dim(extractedvector3x1) #NULL!!!!
> dim(extractedvector1x3) #NULL!!!!
>
> #transposed dimensions both 1x3
> dim(t(extractedvector3x1)) #1x3
> dim(t(extractedvector1x3)) #1x3 <- not 3x1
>
> #as.vector: still NULL
> dim(as.vector(extractedvector3x1)) #NULL!!!!
> dim(as.vector(extractedvector1x3)) #NULL!!!!
>
> #as.matrix: Both dim at 3x1
> dim(as.matrix(extractedvector3x1)) #3x1
> dim(as.matrix(extractedvector1x3)) #3x1 <- Problem: not 1x3
>
> #as.matrix transposed: Both dim at 1x3
> dim(as.matrix(t(extractedvector3x1))) #1x3
> dim(as.matrix(t(extractedvector1x3))) #1x3 <- Problem: not 3x1
>
> How can I get correct dimensions? Is there a way to specify dimensions in
> as.matrix?
>
> Neuro
>
> ______________________________________________
> 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
> and provide commented, minimal, self-contained, reproducible code.
>

I think the confusion relates to the fact that in R vectors are not 2D
arrays with one dimension set to 1.

So your 'simplevector' is not a vector, and your
'extractedvector3x1'
and 'extractedvector1x3' are in fact not 3x1 or 1x3, they are 3-element
vectors, which sometimes behave like 1x3 arrays, sometimes not.

I also found that very confusing in the beginning.  You have to either
keep that in mind in your code or use the 'drop' option if that is
convenient.

	Cyril.

On Thu, 2006-27-07 at 09:23 -0400, Neuro LeSuperH?ros
wrote:
> Transpose vector extracted from a matrix
> 
> Hello,
> 
> I am doing a recursive analysis that uses every line (vector) of a matrix
in
> a loop. In the model, I need to transpose those vectors that are extracted 
> from a matrix.
> 
> Using simple vectors (no matrix involved) the transpose function works
fine:
> 
> simplevector <-matrix(1:3,3,1)
> tsimplevector <-t(simplevector) #transposed
> dim(simplevector) #3x1 matrix (vector)
> dim(tsimplevector)#1x3
> 
> PROBLEM: However, when the vector is extracted from a matrix, its dimension
> is NULL.  In this case the transposed dimension is correct, but you'll
see
> the next example, that if a row was extracted, the dimension would be
wrong:
> 
> initialmatrix <- matrix(1:9,3,3)
> extractedvector <-initialmatrix[,1] # extract first columm as vector
> textractedvector <-t(extractedvector)#transposed
> dim(extractedvector) #NULL!!!! <- not working
> dim(textractedvector)#1x3 as expected
> 
> I have tried to transform the extracted vector as.vector and as.matrix. 
> as.vector does not give the what I want (still NULL) and as.matrix
can't
> specify the number of row and columns so both vectors are vertical.
> 
> In this example, I extract a column and a row. Notice that both 
> as.matrix(vector) have the same dimension. Notice that both transposed 
> vectors have 1x3 dimensions.
> 
> initialmatrix <- matrix(1:9,3,3)
> extractedvector3x1 <-initialmatrix[,1]  # extract first columm as vector
> extractedvector1x3 <-initialmatrix[1,]  # extract first row as vector
> 
> #Both are NULL
> dim(extractedvector3x1) #NULL!!!!
> dim(extractedvector1x3) #NULL!!!!
> 
> #transposed dimensions both 1x3
> dim(t(extractedvector3x1)) #1x3
> dim(t(extractedvector1x3)) #1x3 <- not 3x1
> 
> #as.vector: still NULL
> dim(as.vector(extractedvector3x1)) #NULL!!!!
> dim(as.vector(extractedvector1x3)) #NULL!!!!
> 
> #as.matrix: Both dim at 3x1
> dim(as.matrix(extractedvector3x1)) #3x1
> dim(as.matrix(extractedvector1x3)) #3x1 <- Problem: not 1x3
> 
> #as.matrix transposed: Both dim at 1x3
> dim(as.matrix(t(extractedvector3x1))) #1x3
> dim(as.matrix(t(extractedvector1x3))) #1x3 <- Problem: not 3x1
> 
> How can I get correct dimensions? Is there a way to specify dimensions in 
> as.matrix?
> 
> Neuro
> 
> ______________________________________________
> 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
> and provide commented, minimal, self-contained, reproducible code.