How to read txt file in matlab

Dear all Any one could help me reading the txt file 'specimen1.txt' to plot the load vs extension .For instance

    Time Extension Load
    (sec) (mm) (N)
    0,00000 0,00000 -0,06352
    0,00200 0,00000 -0,06571
    0,00400 0,00000 -0,06751
    0,10400 0,00697 -0,00243
    0,20400 0,02255 -0,00033
    0,30400 0,03494 -0,00515
    0,40400 0,04640 0,02844
    0,50400 0,05832 0,03087
    0,60400 0,06989 0,05052
    0,70400 0,08169 0,05281
    0,80400 0,09350 0,02599
    0,90400 0,10495 0,06418
    1,00400 0,11652 0,00930
    1,10400 0,12774 0,07347
    1,20400 0,14001 0,08514
    1,30400 0,15135 0,03269
    1,40400 0,16351 0,09253
    1,50400 0,17484 0,09940

I tried with load but i got five columns with random values but i need them in three columns as titled above.Then plot the load column as function as extension.Any one could help please as soon. Regards mahmoud

Answer 1

[per isakson]

See Import data file with comma decimal point to get comma2point_overwrite

&nbsp

Addendum

Then run

    comma2point_overwrite('specimen1.txt')

(I assume you have a backup) and

    fid = fopen('specimen1.txt');
    cac = textscan( fid, '%f%f%f', 'Headerlines,2, 'CollectOutput',1 );
    fclose(fid);

Answer 2

[mahmoud el majzoub]

the file attached here

Answer 3

[Image Analyst]

Did you try importdata() with 2 header lines as an optional input?

Answer 4

[mahmoud el majzoub]

Yeah i see , i have the version 2010 ,do you some other method please .Thank for your help

[Star Strider]

You should be able to read ‘specimen1d.txt’ without problems using textscan. It used decimal-point rather than decimal-comma formatting.

Answer 5

[Michael Haderlein]

When I have this issue, I usually go this way:

fid=fopen('test.txt');
txt=char(fread(fid));
fclose(fid);
headerlines=2;
txt(1:max(find(txt==char(13),headerlines,'first')))=[];
txt=strrep(txt',',','.');
data=str2num(txt);

As only low-level functions are used, I guess this works independently from the Matlab release.

Answer 6

[Stephen23]

Why make this complicated: without changing the original datafile we can simply change the commas for periods inside MATLAB, and then use textread on the resulting string:

>> str = fileread('specimen1.txt');
>> str = regexprep(str,',','.');
>> M = cell2mat(textscan(str,'%f%f%f'))
M = 
       0         0    0.0076
  0.0020         0    0.0067
  0.0040         0    0.0040
  0.1040    0.0069    0.0294
  0.2040    0.0228   -0.0259
  0.3040    0.0347    0.0345
...
 75.4040    8.7966   -0.1200
 75.5040    8.8082   -0.1276
 75.6040    8.8201   -0.1791
 75.7040    8.8314   -0.1739
 75.8040    8.8436   -0.1943
 75.9020    8.8547   -0.1583

[mahmoud el majzoub]

Thanks for all. Stephen, your code was what I need exactly. But I would like to ask you: if I need that for many files for example from 1 to 7, and plot load vs extension for these files on the same plot, what do I do?

[Image Analyst]

For many files, put your code in the middle of a loop over all files. See the FAQ for a couple of code examples: http://matlab.wikia.com/wiki/FAQ#How_can_I_process_a_sequence_of_files.3F

[Stephen23]

@mahmoud el majzoub: That comment is not very clear to me what you are asking.

If all of your files have commas as the decimal radix, then you will need to use this (or some) conversion on all of your files. You can process multiple files in a loop.

[Image Analyst]

Mahmoud's code moved here:

Hello Stefan I tried this but it gave me an emoty matrices

for i=1:8
        textFileName = ['specimen' num2str(i) '.txt'];
  if exist(textFileName, 'file')
          fid = fopen(textFileName, 'rt');
    str = fileread(textFileName);
    fclose(fid);
    
          str = regexprep(str,',','.');
          M = cell2mat(textscan(str,'%f%f%f'))
  else
    fprintf('File %s does not exist.\n', textFileName);
  end
end

[Stephen23]

@mahmoud el majzoub: When I wrote this code I used your example file "specimen1.txt", which contains no headers. If the files have headers, then this method will not work. If you try it with your own example file, it will work. If you have a few header lines, then this code will get rid of them before the conversion:

>> N = 2; % <- pick how many header lines.
>> X = find(str==10);
>> str = str(X(N):end);

A few tips about your code:

• Do not use i or j as loop variable names, as these are names of the inbuilt imaginary unit.
• Use sprintf to create the filenames: sprintf('specimen_%i.txt',k)
• Or instead of creating the filenames and checking if they exist, use dir instead:

S = dir('specimen_*.txt');
N = {S.name};
for k = 1:numel(N)
    filename = N{k};
    ... your code here
end

Answer 7

[mahmoud el majzoub]

Moved.