How to read from a text file after skipping few characters in the file ?

Suppose I have a file like this :

ID: 100 Strength: 800e+001 Time: 4536.83736

I want to read only the numerical values. How can I do this ?

[Stephen23]

Not enough information! Are the strings always the same? Do the number of values per line change, or is it constant? is there a header?

[Swarnava Pramanik]

Hi Stephen,

Thanks for the answer. Yeah there is a header and the string length every line is not the same and also the value changes for every line. Could you please help me with this

[Swarnava Pramanik]

Hi Stephen, I'm also getting an error

Error using textscan Invalid file identifier. Use fopen to generate a valid file identifier.

[Stephen23]

This error indicates that fopen did not manage to open the file. Probable reasons are

• the supplied filename is not the name of an actual file.
• the file is not located in the current directory and you have not provided the fullpath.

You don't tell us anything about how your files are arranged so it is impossible for me to say which of these are the problem. However you can fix it quite easily yourself, by providing the correct filename/filpath as the first argument to fopen:

fid = fopen(filename,'rt'); % file is in current directory

If you need to generate the fullpath from path and filename strings, then you should use fullfile to do this:

fid = fopen(fullfile(pathstr,filename),'rt');

Answer 1

[Stephen23]

Try this:

fid = fopen('temp.txt','rt');
C = textscan(fid,'%*s%f%*s%f%*s%f');
fclose(fid);

C{1} will contain the first column of numeric values, C{2} contains the second column, and C{3} contains the third. If you also decide that you want the field names, simply remove the asterisks from the format string.

A slightly more advanced solution is to keep the fieldnames, and turn them into a structure:

fid = fopen('temp.txt','rt');
C = textscan(fid,'%s%f%s%f%s%f');
fclose(fid);
C(1:2:end) = strrep(cellfun(@unique,C(1:2:end)),':','');
D = struct(C{:});

This lets you access the values by the same names as the have in the datafile, e.g.

D.Strength

With the addition of

C(2:2:end) = cellfun(@num2cell,C(2:2:end),'UniformOutput',false);

it could also be turned into a non-scalar structure, which might make your work even easier.

[Swarnava Pramanik]

Hi Stephen,

I have this file and I need to access the data from this text file for my function. But am not able to access the data. I'm relatively new to matlab and am not accustomed with different inbuilt function matlab has.

[Stephen23]

Thank you for providing the complete textfile: as we can't read minds or your computer screen, we need you to help us by giving as much information as possible so that we can actually know what you need to do. There are billions of possible "textfile" formats and ways of arranging and accessing data, and as you can see, what you wrote and what you have in the file are different. Keep that in mind when you ask questions in future.

Now that we have your actual file, it was easy to create the correct code to read the data using textscan into a cell array C:

fid = fopen('current.txt','rt');
C = textscan(fid,'%s%f', 'MultipleDelimsAsOne',true,...
    'HeaderLines',1, 'Delimiter',':');
fclose(fid);

You can access the variable names (strings) in C{1}, ans the numeric data are stored in C{2}. We can access these values using indexing:

>> C{1}{3}
ans =
Eccentricity
>> C{2}(3)
ans =
    0.0039239

Note that to get the string we use cell array indexing with curly braces {}, whereas to get the numeric values we must use array indexing with parenthses ().