To read a .dat file by replacing 'nil' s with zeros

I have a .dat file with 3 columns :

56  1221.9522403147  4709.4286506579    
57  1296.7593276229  2382.7334294788    
58  1298.014307685  5378.3218155863    
59  1672.5745510151  3586.6821218019    
60  1293.7368062958  2311.6447578082    
61  1066.5874471792  3652.8767627687    
62  nil  7015.2811482462    
63  1295.6502693748  12997.366157104    
64  1473.1366368752  3878.5975654487    
65  1893.552884693  4782.6543016909    
66  1601.2373040587  5157.4325835724

This is a part of my data. it has some 'nil' s appearing somewhere. How can I read the data into a 3 x n matrix by replacing those 'nil' s with zeros?

Answer 1

[Walter Roberson]

fid = fopen('YourFile.dat', 'rt');
yourdata = cell2mat( textscan(fid, '%f%f%f', 'Delimiter', '\t', 'TreatAsEmpty', 'nil', 'EmptyValue', 0, 'CollectOutput', 1) );
fclose(fid);

[aneps]

and if I want to read from nth line, how should I do it?

[Walter Roberson]

N = 19;

fid = fopen('YourFile.dat', 'rt');
yourdata = cell2mat( textscan(fid, '%f%f%f', 'HeaderLines', N, 'Delimiter', '\t', 'TreatAsEmpty', 'nil', 'EmptyValue', 0, 'CollectOutput', 1) );
fclose(fid);

[aneps]

Did not work :( .. getting error message "Error using textscan Invalid file identifier. Use fopen to generate a valid file identifier."

[Walter Roberson]

Did you change fopen('YourFile.dat', 'rt') to reflect your actual file name? And is your actual data file in the current directory?

[aneps]

I used my file name (Result_2.dat) instead of 'YourFile.dat'.

[Walter Roberson]

   filename = 'Result_2.dat';
   if ~exist(filename, 'file')
     error('Your data file %s does not exist in this directory', filename);
   end
   [fid, message] = fopen(filename, 'rt');
   if fid < 0
      error('Opening the file %s failed because %s', filename, message);
   end
   yourdata = cell2mat( textscan(fid, '%f%f%f', 'Delimiter', '\t', 'TreatAsEmpty', 'nil', 'EmptyValue', 0, 'CollectOutput', 1) );
   fclose(fid);

[aneps]

Yes, it worked.. thank you... I have one more question, if I have several data like Result_1.da, Result_2.dat.... Result_n.dat, how can I loop it to read all as yourdata1, yourdata2, .... yourdatan?

[Walter Roberson]

http://matlab.wikia.com/wiki/FAQ#How_can_I_process_a_sequence_of_files.3F

Answer 2

[Azzi Abdelmalek]

fid=fopen('file.dat')
v=textscan(fid,'%s')
fclose(fid)
w=reshape(v{:},3,[])'
w=strrep(w,'nil','0')
out=str2double(w)

[aneps]

if I want to read from nth line, how should I do it? Thanks

[Azzi Abdelmalek]

Use

n=5
v=textscan(fid,'%s','headerlines',n)

[aneps]

did not work:( .. getting error message "Error using textscan Invalid file identifier. Use fopen to generate a valid file identifier."