how to read a txt file having alphanumeric data and special characters. ? My text file is like this. I tried textread but could not get a result. any suggestion.

     1        6:s        6:s          0      0      0      0.11   0     *0.889
     2        6:s        6:s          0      0.272  0      0.006  0     *0.722
     3        6:s        6:s          0      0.036  0      0      0.001 *0.963
     4        6:s        6:s          0      0      0      0      0     *1    
     5        6:s        6:s          0      0      0      0      0     *1    
     6        6:s        3:f      +   0      0.238 *0.365  0.009  0.033  0.355
     7        6:s        5:n      +   0      0      0      0     *0.999  0.001
     8        6:s        6:s          0      0      0      0      0     *1

[jonas]

What is your desired output?

[Muhammad Shoaib]

Thanx Jonas, I want the last six columns on output containing only the numeric data.

[Walter Roberson]

How do you want the * to be handled?

Answer 1

[Stephen23]

Here is an efficient solution using textscan. The test file is attached (you did not provide us with a sample file).

opt = {'Delimiter',' +*:','MultipleDelimsAsOne',true,'CollectOutput',true};
fmt = ['%*d%*d%*s%*d%*s',repmat('%f',1,6)];
[fid,msg] = fopen('test.txt','rt');
assert(fid>=3,msg)
C = textscan(fid,fmt,opt{:});
fclose(fid);

Which imports that last six columns of numeric data:

>> C{1}
ans =

     0.00000   0.00000   0.00000   0.11000   0.00000   0.88900
     0.00000   0.27200   0.00000   0.00600   0.00000   0.72200
     0.00000   0.03600   0.00000   0.00000   0.00100   0.96300
     0.00000   0.00000   0.00000   0.00000   0.00000   1.00000
     0.00000   0.00000   0.00000   0.00000   0.00000   1.00000
     0.00000   0.23800   0.36500   0.00900   0.03300   0.35500
     0.00000   0.00000   0.00000   0.00000   0.99900   0.00100
     0.00000   0.00000   0.00000   0.00000   0.00000   1.00000

[Muhammad Shoaib]

thanks Stephen Cobeldick, My actual file is attached. I need only the last six columns in this file with only numeric data. Please help me in this regard.

[Stephen23]

@Muhammad Shoaib: did you try my answer with your file? My code already works with your file, without any changes. Here are the first and last ten lines from the file, as imported into MATLAB:

>> C{1}(1:10,:)
ans =
   0.00000   0.00000   0.00000   0.11000   0.00000   0.88900
   0.00000   0.27200   0.00000   0.00600   0.00000   0.72200
   0.00000   0.03600   0.00000   0.00000   0.00100   0.96300
   0.00000   0.00000   0.00000   0.00000   0.00000   1.00000
   0.00000   0.00000   0.00000   0.00000   0.00000   1.00000
   0.00000   0.23800   0.36500   0.00900   0.03300   0.35500
   0.00000   0.00000   0.00000   0.00000   0.99900   0.00100
   0.00000   0.00000   0.00000   0.00000   0.00000   1.00000
   0.00000   0.00000   0.00000   0.00000   1.00000   0.00000
   0.00000   0.00000   0.00000   0.00000   0.00000   1.00000
>> C{1}(end-9:end,:)
ans =
   0.00000   0.00000   0.00000   1.00000   0.00000   0.00000
   0.00000   0.00100   0.00000   0.94200   0.05700   0.00000
   0.00000   0.26800   0.00000   0.73200   0.00000   0.00000
   0.00000   0.00000   0.00000   1.00000   0.00000   0.00000
   0.00000   0.00000   0.00000   1.00000   0.00000   0.00000
   0.00000   0.00000   0.00000   1.00000   0.00000   0.00000
   0.00000   0.00000   0.00000   0.00400   0.99600   0.00000
   0.00000   0.00000   0.00000   1.00000   0.00000   0.00000
   0.00000   0.00000   0.00000   1.00000   0.00000   0.00000

[Muhammad Shoaib]

thank Stephen Cobeldick, the code worked for me.

[jonas]

Don't forget to accept the answer so that future readers can see that the issue was resolved.

Answer 2

[jonas]

str=fileread('notes.txt');
str=regexprep(str,'[:+*a-zA-Z]',' ')
opt = {'Delimiter',' ','MultipleDelimsAsOne',true,'CollectOutput',true};
out = textscan(str,'%f%f%f%f%f%f%f%f%f',opt{:})

[Stephen23]

Note that str2num relies on eval, which makes this a risky method to use with input data, especially with data containing weird random characters.

However note that textscan also works on character vectors (not just files), so you could easily use textscan to efficiently parse the char vector into a numeric array.

[jonas]

Oh, didn't know that. Thanks. I like to overuse regexp for everything and figured this could work.

[Muhammad Shoaib]

thanks Jones, My actual file is attached. I need only the last six columns in this file with only numeric data. Please help me in this regard.

[jonas]

Ok, I've updated the solution. This really proved Stephens point, as str2num failed miserably. Just grab the last 6 columns of the results.

I think the other answer worked with your file. Did you even try it?

[Muhammad Shoaib]

Thanks Jonas, for your kind support and cooperation.

[jonas]

No problem! I'll refrain from deleting this answer as it shows how to use regexprep to clean up a textfile but make sure to accept the other one :)

[Stephen23]

+1 this is a good idea, very useful for complex/unusual file formats. Do not delete!