Read text file without header and separate columns

Dear Matlab Team,

my problem is as follows. I could not find a solution in other forums, but none seem to work for my issue:

I have a text file with a big headliner, which I want to skip. The data I need starts in line 194 and has 50.000 values. There are 3 values in each row, sperated by ",". Basically, what I need is to extract the 50 000 values and put it in matrix with 3 columns for further analyzation.

This is how the file looks like (starting from below "DATA BLOCK..."]

...
[DATA BLOCK 1-1-1]
0.00000,20.47668,0.13959
0.00010,20.46021,0.16467
0.00020,20.47668,0.11914
0.00030,20.48767,0.19208
...

How can I separate this data and create a matrix from it?

Thanks in advance,

Philipp

Answer 1

[Rik]

You can use the readmatrix function.

A = readmatrix(filename,'NumHeaderLines',193);

Edit:

As Walter suggested below: releases older than R2019a are missing the readmatrix function. On R2013b or newer, the readtable function can help out:

A = table2array(readtable(filename,'NumHeaderLines',193, 'readvariablenames', false));

[Philipp Reh]

Thanks for your quick response.

It does not recognize the "readmatrix" command; "Undefinded function or variable 'readmatrix'. (Spelling is correct).

Is there any code beforehand necessary and do the "," have no influence?

[Walter Roberson]

You did not specify your MATLAB release. readmatrix() needs R2019b or later.

For earlier releases,

A = table2array(readtable(filename,'NumHeaderLines',193, 'readvariablenames', false));

readtable() should be able to detect the comma delimiter. If you want to be sure, you can use

'Delimiter', ','

[Philipp Reh]

I downloaded the 2020a version and it worked (was 2018).

Thanks a lot.

Philipp

[Philipp Reh]

I have a follow up question, which is a bit off topic. I hope you can still help me. I have more than one txt file in the main file. I want to add the other files to the same array, so that I have a 50.000 (rows) x 90 (columns) (for 30 files) Matrix. I have the main code ready, however I don't know how it adds up all three columns and not only one:

input_folder = 'folder'; 
files = dir(fullfile(input_folder, '*.txt'));
file_paths = fullfile({files.folder}, {files.name});
for i = 1 : numel(file_paths)
    data =  readmatrix(file_paths{i},'NumHeaderLines',193);
    A(:,i) = data(:,3);
    
end

I think the Mistake is in the last line, can you help me here again please?

Best,

Philipp

[Rik]

Assuming you have 3 columns per file, something like the code below should help. The point is that you need to compute the indices

cols=3*(i-1)+(1:3);
A(:,cols)=data;

[Walter Roberson]

sum_by_3 = squeeze(sum(reshape(data, size(data,1), 3, []),2));

This sums groups of 3 columns, so like

[data(1,1)+data(1,2)+data(1,3), data(1,4)+data(1,5)+data(1,6), ...
 data(2,1)+data(2,2)+data(2,3), data(2,4)+data(2,5)+data(2,6), ...

[Walter Roberson]

Looks like I misinterpreted what was being requested.

You should be pre-initializing A.

input_folder = 'folder'; 
files = dir(fullfile(input_folder, '*.txt'));
file_paths = fullfile({files.folder}, {files.name});
nfile = numel(file_paths);
for i = 1 : nfile
    data =  readmatrix(file_paths{i},'NumHeaderLines',193);
    if i == 1
      A(:,3*nfile) = 0;     %grow array to largest size
    else
      A(:, i*3-2:i*3) = data;
    end
end

[Philipp Reh]

Thank you.

Answer 2

[B]

I think summarizing you've got something like this:

input_folder = 'folder'; 
files = dir(fullfile(input_folder, '*.txt'));
file_paths = fullfile({files.folder}, {files.name});
nfile = numel(file_paths);
for i = 1 : nfile
    data=  readmatrix(file_paths{i},'NumHeaderLines',1);% a display of data in a single file
    cols=3*(i-1)+(1:3)
    A(:,cols)=data; % A is your desired result outpout file
end