How to Read Data from one type of (.sema) file and to convert it into .txt file

I have 20000 (.sema) file and want to get .txt file from .sema file by excluding 1st column and first 11 rows. For example file name is acc_HXX_X_0_Y_0.

clear all

DATA_DIR = 'Desktop/subshear/subshear_acc/';

out_DIR = 'EQ/subshear/acc_files_subshear/';

x = [-30:5:30].*1e3;

y = [-30:5:30].*1e3;

[xx yy]=meshgrid(x,y);

nstat=length(xx);

nx = length(x);

ny = length(y);

kk=0;

for i=1:nx

for j=1:ny

kk=kk+1;

st_name=sprintf('X%dkY%dk',xx(kk)./1e3,yy(kk)./1e3);

filename = sprintf('%s/IND.%s.HXX.sema',DATA_DIR,st_name);

% fprintf(fid,'%s \t %s \t %f \t %f \t %f \t %f\n',st_name,'IND',yy(kk), xx(kk), 0.0, 0.0);

t_acc=load(filename,'r');

filename_acc = sprintf('%s/acc_%s.txt',out_DIR,st_name);

fid=fopen(filename_acc);

%fprintf(fid,'%g',t_acc(:,2));

%fclose(fid);

end

end

Answer 1

[Cris LaPierre]

Quick inspection suggests that your files are just text files with a custom extension. At least I can open them in a text editor and read them just fine.

MATLAB does try to determine file properties based on the extension. That will not be possible with this extension. However, you can still import the data. You just need to set some settings manually.

The file you mention is not in the zip file you shared. The 5 files you share all are 800x2. Here's how I might do it:

data = readmatrix("IND.X0kY0k.HXX.sema","FileType","text","Range",[12 2])

data = 789×1    
              7.151661e-13
              2.853065e-12
             4.9193176e-12
            -3.5160106e-12
            -4.2815907e-11
            -1.1427437e-10
            -1.2858667e-10
              1.392735e-10
             8.8575636e-10
             1.7722569e-09

[Cris LaPierre]

Alas, I see you are using 2017b. I tested my code in 20b. Readmatrix wasn't introduced until R2019a. If you can update, that would be my first recommendation.

Otherwise, I suggest using the Import Tool to interactively import a file. From there, you can generate an import script or function to use on the rest of the files.

With some help from the code generated by the import tool, I was able to come up with this code (tested in R2017a).

% import assuming file is fixed width
fileID = fopen('IND.X0kY0k.HXX.sema','r');
dataArray = textscan(fileID, '%*17s%f%[^\n\r]', 'Delimiter', '',  'WhiteSpace', '', 'HeaderLines' ,11, 'EndOfLine', '\r\n');
fclose(fileID);
% Convert cell array to an array
IND = [dataArray{1:end-1}]

[Cris LaPierre]

Use this table to determine the best function for saving the array to a text file in R2017b. The options appear to be:

• Comma delimited numbers - csvwrite
• Delimited numbers - dlmwrite
• Delimited numbers, or a mix of text and numbers - none
• Column-oriented delimited numbers or a mix of text and numbers - writetable

[Sumit Saha]

@Cris LaPierre Thank you so much. But there are 20,000 .sema files (here I've uploaded only 5 files) which I need to convert it into .txt file. With the help of your code it's quite easy to convert each file into txt file (one at a time) or I've to write 20000 line to read sema file. Is there any possibility to extract the data with the help of 'for loop' with only second column?

[Cris LaPierre]

Loading multiple files is a completely separate problem unrelated to loading the 2nd column of data. You'll find that has been adequately discussed on the forum already. I would point you to the 2 options provided in this recent post.

The bigger question is what is the end goal? Each one being converted to a text file, or one text file containing the 20,000 files * 800 rows of data?

[Sumit Saha]

@Cris LaPierre Each one being converted to a text file ---that is my ultimate goal.

"Loading multiple files is a completely separate problem unrelated to loading the 2nd column of data"- yeaahh that's definately true.

[Cris LaPierre]

Then a for loop is probably the most straight-forward solution. The trick is capturing all the files. If they are all in the same folder, then something like this should work (untested)

S = dir('*.sema'); % assumes the files are in your current folder
for k = 1:numel(S)
    file = fullfile(S(k).folder, S(k).name);
    
    % import assuming file is fixed width
    fileID = fopen(file,'r');
    dataArray = textscan(fileID, '%*17s%f%[^\n\r]', 'Delimiter', '',  'WhiteSpace', '', 'HeaderLines' ,11, 'EndOfLine', '\r\n');
    fclose(fileID);
    % Convert cell array to an array
    IND = [dataArray{1:end-1}];
    
    % Write IND to a text file. 
    % ## Probably best to use a different folder than your source folder. ##
    [filepath,name,ext] = fileparts(file);
    dlmwrite(fullfile(filepath,[name '.txt']),IND,'precision','%9.7e');
end

[Sumit Saha]

@Cris LaPierre Thank you so much...it's working now

[Sumit Saha]

@Cris LaPierre how can i rename these files ?

example- IND.X0kY0k.HXX.txt to 1.txt like wise for every files

[Cris LaPierre]

Use the movefile function (Move or rename file or folder).

[Sumit Saha]

@Cris LaPierre

I've used that part but somehow these initian files are replaced by one file.

clear all
close all
% f = dir('*.txt');
% for kk = 1:numel(f)
%   fileFrom = f(kk).name;
%   fileTo = [erase(f(kk).name,'.txt'),'.txt'];
%   movefile(fileFrom,fileTo);
% end
files = dir('*.txt');
for i=1:length(files)
  [pathname,filename,extension] = fileparts(files(i).name);
  % the new name, e.g.
  newFilename = filename(1:5);
  % rename the file
  movefile([filename extension], [newFilename extension]);
end

[Sumit Saha]

Actually I just want to rename the files serially by 1.txt,2.txt,3.txt.....

[Cris LaPierre]

I don't understand your code. What is this supposed to be doing?

newFilename = filename(1:5);

[Sumit Saha]

@Cris LaPierre

I've just tried to delete the old name and put numeric character. My aim is to delete the old file name and replace it with numbers ( serially).

[Sumit Saha]

@Cris LaPierre Is there any way to do this renaming part serially?

[Cris LaPierre]

That line of code is not doing what you think it is doing.

You already have a counter that increases with each new file. Why not just use that? This code in untested.

files = dir('*.txt');
for i=1:length(files)
  [pathname,filename,extension] = fileparts(files(i).name);
 
  % rename the file
  movefile([filename extension], [num2str(i) extension]);
end

[Sumit Saha]

@Cris LaPierre yeah...It's working . Thanks a lot to sort out this issue.

[Sumit Saha]

@Cris LaPierre How can I collect the names of files before converting these into serial order?

[Cris LaPierre]

You can use the dir function to collect the contents in a folder. Perhaps my answer in this post might be helpful.