empty matrix when using textscan

Hi, this is my file and it is called «1.dat»:

bou=200
bou=30480
bou=22860
bou=0
bou=497

number  xfe     fee     prefue       cefee        ewee          wef                 wfe        wefn      
1              13767          11061          0              208963921855920      74822197             208963921855920      0              900            
2              13767          11061          0              208963949100942      74822207             208963949100942      450            890            
3              13767          11061          0              208963981822687      74822217             208963981822687      450            860            
4              13757          11051          0              208964003640637      74822227             208964003640637      390            840

i want to ignore the 6 first lines and i used:

D = cell2mat(textscan('1.dat', '%f%f%f%f%f%f%f%f%f', 'headerlines', 6))

but it keeps on returning me «empty matrix: 0-by-6» i've also tried putting 7 instead of 6 (because of the empty line but still doesnt work, same error)

what is wrong?

Answer 1

[dpb]

textscan uses fopen/fclose to open/close the file and the subsequent file handle, NOT the actual file name. By using the name of the file in the location for the file handle, it is trying to read the string '1.dat' as the data stream which obviously doesn't match the expected data format.

fid=fopen('1.dat','r');  % check for success in real life...
D=cell2mat(textscan(fid, '', 'headerlines', 7));  % if the blank record is there, must count it...
fid=fclose(fid);  % close the file after reading...
% do whatever following...
...

NB:

You'll note I used an empty string for the format string -- this has the effect of textscan will internally scan the input record and ascertain the number of fields and automatically return the shape of the resulting cell array to match the file structure. This saves a lot of effort in building format strings and is especially useful if the number of columns isn't known a priori.

[Van Ly]

it worked! thank you very much!

[Jeremy Hughes]

You could also try READTABLE, it should be able to do most of this automatically in recent releases.