How to Imorting a text file containig indexes and the coressponding elements of a matrix to matlab?

want to get a matrix from a text file that contains the corresponding elements and there indexes.

Answer 1

[Pavan Sahith]

Hello Himanshu,

I see that you are looking to import a matrix from a text file containing corresponding elements and their indices. Let's create a sample text file and see how to import it into MATLAB.

As I dont have any text file , I am creating sample.txt with some random data

% Create a sample file
data = [1, 1, 5.5; 1, 2, 6.3; 2, 1, 7.8; 2, 2, 8.1];
writematrix(data, 'sample.txt', 'Delimiter', ',');

To import this text file into MATLAB, you can use the 'readmatrix' function:

data = readmatrix('sample.txt');

then you can extract the indices and values, then populate a matrix accordingly by using a simple 'for' loop :

% Extract indices and values
row_idx = data(:, 1);
col_idx = data(:, 2);
values = data(:, 3);

% Determine the size of the matrix
max_row = max(row_idx);
max_col = max(col_idx);

% Initialize the matrix
matrix = zeros(max_row, max_col);

% Populate the matrix
for i = 1:length(values)
    matrix(row_idx(i), col_idx(i)) = values(i);
end

% Display the matrix
disp(matrix);

you can refer to following MathWorks documentation to know more about

'readmatrix'-https://www.mathworks.com/help/matlab/ref/readmatrix.html

I hope this helps you complete your task.

[HIMANSHU]

Hello Pavan!

Thanks for answering

Your code is work fine but the matrix i am getting is too big, so it is showing a message that the matrix is too big(101 gb) for the physical memory.

Is there any way to create a sparse matrix without creating a full matrix.

[Stephen23]

https://www.mathworks.com/help/matlab/tall-arrays.html

Answer 2

[Walter Roberson]

Use spalloc -- pass in the number of rows, number of columns, and the total number of entries, and get out a sparse matrix that you can then set individual elements using a loop.

Or, just use sparse passing in the row indices, column indices, and corresponding data values: this will construct the entire array in one call.

Just don't sparse() a partial matrix and then loop setting more elements than were previously allocated -- growing a sparse matrix is expensive.

[HIMANSHU]

Thanks Walter for your answer

This is exactly what I was looking for.