How can I delete header from a .txt file and save it?

I have a text file like this one:

element, unit, refTime, validTime, (62.000000,-6.250000)
HTSGW, [m], 197901010000, 197901010000, 0.000
HTSGW, [m], 197901010000, 197901010300, 0.220
HTSGW, [m], 197901010000, 197901010600, 0.420
HTSGW, [m], 197901010000, 197901010900, 0.560
HTSGW, [m], 197901010000, 197901011200, 2.190

I just want to remove the header and then save the file in the same .txt file without the header.

I may use the next script, shown in other forums, however I cannot save the generated list.

       fid = fopen(Text);

       tline = fgetl(fid);
       tline = fgetl(fid);

            while ischar(tline)
                disp(tline)
                tline = fgetl(fid);

            end

       fclose(fid);

I wish someone could help me. Thanks.

[José-Luis]

Windows or *nix?

[Guillaume]

The best way to not have to worry about windows or *nix, is to tell matlab to take care of line endings for you:

fopen(filename, 'rt');  %for reading
fopen(filename, 'wt');  %for writing

[José-Luis]

I was mostly thinking that this is a one liner for sed and those utilities tend to be more efficient than what Matlab has to offer.

[Guillaume]

Ah yes, certainly it would be more efficient to do that outside of matlab with the OS tools (assuming there is one that does that, not sure about windows...).

Answer 1

[Azzi Abdelmalek]

fid=fopen('file.txt')
s=textscan(fid,'%s%s%s%s%s','headerlines',1)
fclose(fid)
ss=[s{:}]
fid1=fopen('file.txt','w')
for k=1:size(ss,1)
fprintf(fid1,'%s%s%s%s%s\r\n',ss{k,:})
end
fclose(fid1)

[Javier Martinez]

It worked great thanks

[Javier Martinez]

The only problem is that I am reading many .txt files, and your solution only lets me read a unique .txt file. As I am making a loop, I wish I could generate a unique .txt file with the different data. If you have any idea of that it would be great. I am extremely grateful for your help.

[Guillaume]

Why bother parsing each line into 5 different strings, thus slowing down the program. You actually don't care about the content of each line, so

s = textscan(fid, '%s', ...);

should be enough. Same for the fprintf call.

Answer 2

[Guillaume]

Note: when using fopen always specify the permissions ('r' or 'w' mostly with optional 't' for text mode).

Your code only display the text. You also need to open a file for writing if you want to save it.

Probably the easiest way to do it all would be:

filecontent = fileread('somefile');
newcontent = regexp(filecontent, '(?<=[^\n\r].*[\n\r]+).*', 'match');  %keep everything but first line (defined as any number of characters but newline or carriage return followed by any number of newlines or carriage returns
fid = fopen('newfile', 'w');
fprintf(fid, '%s', newcontent);  
fclose(fid);

[Javier Martinez]

There is an error:

Error using fprintf Function is not defined for 'cell' inputs.

[Guillaume]

D'oh! either add , 'once' after the 'match' in the regexp call. or simply dereference the cell as it should only have one element:

fprintf(fid, '%s', newcontent{1});

Adding 'once' would be more correct.