Matrix indexing, getting back my original matrix.

Hi,

I have a matrix

mat= [ 1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...       
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9];

and I sub-sample it using the simple command

newMat= mat(1:4:end, 1:4:end);

Is there a way later on to get back the indexes I lost from the original matrix, if I want to reconstruct it in a new matrix likewise

newMat2= zeros(size(mat, 1), size(mat, 2));

newMat2(1:4:end, 1:4:end) = newMat;

So what I want is to replace the columns and rows I lost with lets say NaN.

[Jan]

I do not get the problem. In your example you fill the missing values with zeros. So all you want to do is using nan() instead of zeros()?

Answer 1

[Joseph Cheng]

Were you looking to get something like this?

mat= [ 1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9; ...       
       1 2 3 4 5 6 7 8 9; ...
       1 2 3 4 5 6 7 8 9];
selCol = 1:4:size(mat,2);
selRow = 1:2:size(mat,1);
newMat= mat(selRow, selCol);
newMat2= NaN*zeros(size(mat));
newMat2(selRow,selCol) = newMat;

newMat2 =

       1   NaN   NaN   NaN     5   NaN   NaN   NaN     9
     NaN   NaN   NaN   NaN   NaN   NaN   NaN   NaN   NaN
       1   NaN   NaN   NaN     5   NaN   NaN   NaN     9
     NaN   NaN   NaN   NaN   NaN   NaN   NaN   NaN   NaN
       1   NaN   NaN   NaN     5   NaN   NaN   NaN     9
     NaN   NaN   NaN   NaN   NaN   NaN   NaN   NaN   NaN
       1   NaN   NaN   NaN     5   NaN   NaN   NaN     9
     NaN   NaN   NaN   NaN   NaN   NaN   NaN   NaN   NaN
       1   NaN   NaN   NaN     5   NaN   NaN   NaN     9

[Joseph Cheng]

There may be a way to figure out the increments based on the new mat size and original mat but need to think of the conditions.

For instance if you go 1:4:100 you get a 1x25. which you can see 4. but you can just round or floor or ceil +1 for all cases.

Answer 2

[Azzi Abdelmalek]

You haven't lost anything

mat=repmat(1:9,6,1)  % Example
new_mat=mat(1:4,1:4)

You have both the original matrix and the new one, maybe you can give more details about loosing indices

[Raldi]

No this was just an example to illustrate my problem, assume the matrices are not on the same function and there is no way to access my original matrix. What I actually want is to find a way to select all the rows and columns that I left out in my original matrix and populate them with some new values (for simplicity here NaN)