How do I turn a maxtrix to a vector

So I have the following 167x6 matrix:

A B C D E F
G H I J K L
M N O P …

and so on to make a 167x6. I want it to look like (1x1002):

A B C D E F G H I J K L M N O P …

however, when I use reshape, it gives me this:

A G M B H N C I O D J P E K …

It sorts it by columns instead of rows. Please help

[Jan]

I've formatted the code for you today. Please use the "{} Code" button in the future.

[Greg Smith]

Thank you. Sorry, I'm brand new to this.

Answer 1

[John D'Errico]

So, you cannot use reshape in a different way? Suppose you first transpose the matrix, and THEN used reshape? (Hint.)

The point is, MATLAB stores data in a matrix going down the columns in memory. So, IF you transpose the matrix first, reshape will give you what you are looking to do.

[Greg Smith]

What I don't understand is that it's saying that the number values don't match, yet I haven't added or removed any of the data points. A 167x6 matrix has the same value points as a 1x1002 vector. Why am I receiving this error?

[Stephen23]

"Transpose made my data out of order though,..."

Can you please show us what you tried, and the output you got.

"Is there a way to simply take row2 and put it at the end of row1?"

Yes. Transpose and reshape, just like John D'Errico wrote.

[Greg Smith]

>>> B = transpose(CubeSat);
>>> C = reshape(B, [6:167,1,1002]);
error using reshape
to RESHAPE the number of elements must not change.

I apologize, I'm very new at this

[Stephen23]

reshape(CubeSat.',1,[])

[Jan]

@Greg: No need for apologies. This is a forum for Matlab questions. You do have Matlab questions. Then the members of the forum post answers.

[Walter Roberson]

C = reshape(B, 1, [])

Answer 2

[Jan]

All resorting of arrays, which keeps the neighborhood in a specific dimension, can be performed by

Y = reshape(permute(reshape(X, dim1), order2), dim3)

All you have to find is the correct definitions of the parameters. For 2D arrays, permute(T, [2,1]) can be replaced by a transpose. In some cases one or both reshapes can be omitted.

In your case:

S = ['A' 'B' 'C' 'E' 'F'; ...
     'G' 'H' 'I' 'K' 'L'; ...
     'M' 'N' 'O' 'P' 'Q'];
T = reshape(S.', 1, [])

PS. Inserting the quotes to produce running codes took the most time while typing this answer. Please post the data such, that they can be used by copy&paste. This is cuter than leaving the tedious work for the ones who answer.

[Greg Smith]

Thank you for the help. I will remember this for the future. What I still don't understand is the reshape is saying that I need to have the same values in order to reshape the matrix. From my understanding though a 167x6 is the same as 1x1002. Do you know why I am receiving this error?

[Jan]

C = reshape(B, [6:167,1,1002]);

This tries to reshape the elements of B to an array with the dimensions

 [6,7,8,9,10,...,166,167,1,1002]

This would be huge.

You mean:

C = reshape(B, 1, 1002)

or

C = reshape(B, [1, 1002])

Another solution would be:

C = B(:).'

or the automatic determination of the needed value:

C = reshape(B, 1, [])
%                  ^ replaced automatically by 1002

[Stephen23]

@Greg Smith: because you specified the dimensions incorrectly. You specified the output size like this:

>> [6:167,1,1002]
ans =
    6    7   8    9   10   11  ... lots of numbers here ... 166  167    1  1002

That is a really big array, with size 6x7x8x9x10x11x...x166x167x1x1002. In fact that array would have many more elements in it than there are particles in the observable universe:

>> prod([6:167,1,1002])
ans =   1.2555e+301

You probably meant to specify the size like this:

[1,1002]