Index in position 2 exceeds array bounds (must not exceed 1). Error in dlmread (line 159) result= result(:,1:ncols);

In my try.txt file, there are 9 columns with 300000 float values(300 times 1000 samples). In between every 1000 samples and in the beginning, there is some text written which I tried to skip by incrementing i appropriately. My aim is to find variance of every column with 1000 samples and repeat till the end(i.e will 300000), which will enable me to get 300 by 9 matrix.

clc
clear
close all
file=('try.txt');
%M = dlmread(testtime.txt,' ',50,50)
i=23;
%variences=zeros(300,1);
for k=1:300
    
    j=i+1000-1;
A = dlmread(file,' ',[i 0 j 8]);
variences(k)= var(A(:,3));
i=i+1000+11;
end
variences';
plot(1:300,variences)

variences=variences';

Please help me asap. I am new to matlab

Answer 1

[Jon]

You may have other problems too, but I noticed that since i does not change inside of your loop, you always read from the same starting row. Note that in your call i which is the first (starting row) does not change, but you ending row j does. Instead I would think you want something more like

A = dlmread(file,' ',[j 0 j+1000 8]);

[Jon]

This may give you some ideas about good ways of dealing with data files that have mixed data and headers https://www.mathworks.com/help/matlab/import_export/import-block-of-mixed-data-from-text-file.html

[anshu rai]

Thank you for the reply sir. But is it because of the indentation inside the loop which doesn't let i to change. If yes then I even tried indenting it but there wasn't any change. Also the above code snippet did not help me. Rather it generated another error saying,

Mismatch between file and format character vector.

[Jon]

Sorry, I did not look at your code carefully enough before I responded.I didn not notice that you increment i at the bottom of the loop. Indentations don't matter in MATLAB, it is easier to read if indented properply but how the code runs is only based upon matching end statements with other control flow elements (for, while, if etc). So changing the indentation will not fix anything. To format indent the code nicely, do not manually use tabs or added spaces. Just select all of the code and then right click and select smart indent. Can also select all with ctrl-A, and then smart indent with ctrl-I.

Now I will try to look more closely at your code and see if I can give you any further insight

[Jon]

When you calculate the variances you want to calculate them for each of the 9 columns in A. Also you need to assign the whole resulting 9 element vector of variances to an array to hold the results for that block of data . You should preallocate your array variances before the loop to efficiently allocate space and also to ensure that there isn't anything using already assigned to that array (even though you do a clear at the beginning, still good practice) See if this gets closer. Indexing is tricky with the offsets so please confirm that I got this right and it is consistent with what is in your files

clc
clear
close all
% assign parameters at beginning to avoid "magic numbers" appering in code
numHeaderRows = 23; % number of initial header rows
numBlocks = 300; % number of data blocks
numDataRows = 1000; % number of data rows in a block
numCols = 9; % number of data columns
numTextRows = 12; % number of text rows between blocks

% assign name of file holding the data
file='try.txt';

% prepare to loop through data
% assign initial offset to first row of data to be read 
i=numHeaderRows - 1; % initial offset (zero based)
variances=zeros(numBlocks,numCols); % preallocate array to hold calculated variances
for k=1:numBlocks
    % assign ending offset for block
    j=i+1000-1;
    
    % read in block of data
    A = dlmread(file,' ',[i 0 j numCols-1]);
    
    % calculate variances
    variances(k,:)= var(A(:,3)); % assign to current row in result matrix
    
    % increment starting offset for next block
    i = i + numDataRows + numTextRows;
end
% plot results
plot(1:numBlocks,variances)

% save variances as numberCols by numBlocks array
variances = variances';

[Jon]

If this still is not working then please attach a portion of your data file with the first couple of blocks of data (so about 2000+ rows, not too big but big enough to see exactly what is in there and the repetitive pattern)