Changing the format of output of data from a neural network

Hi there,

I'm using the code below to loop through files in a folder and classifying data with a neural network.

files = dir(fullfile(pwd,'*.txt'));
for k = 1 : numel(files)
    %%%%reading the txt file
    fid01 = fopen(fullfile(pwd,files(k).name));
    idx = 0;    tmparray = []; 
    tline = fgetl(fid01);
    while ischar(tline)
        idx = idx + 1;
        tmparray(idx,:) = str2num(tline);
        tline = fgetl(fid01);
    end
    fclose(fid01);
      %%%%simulate net
      output = sim(net, tmparray);   
      %%%%save the output
      filename = strcat(regexprep(files(k).name,'.txt',''),'-output.txt');
      fid02 = fopen(filename,'w');
      fprintf(fid02,'%.2f\n',output');    
      fclose(fid02);
  end

However the output is in the form of a 1x1 array, for example 1x1 array. For example

    output = 
    [ 
    0 
    0 
    0 
    0 
    1 
    1 
    1 
    0 
    0 
    0 
    0 
    1 
    0 
    0 
    0 
    0 
    0 
    0 
    0 
    0 
    0 
    0 
    0 
    0 
    0 
    0 
    0 
    0 
    ]

Would it be possible to convert it into a sparse matrix like this

output = full(output)

0  0  0  0 
1  1  1  0 
0  0  0  1 
0  0  0  0 
0  0  0  0 
0  0  0  0 
0  0  0  0

and then convert it to vectors , for example

Output = vec2ind(output)

2  2  2  3

I'm trying to get the output in integer format.

I would really appreciate any help

Thank you

John

Answer 1

[Greg Heath]

For a net with I-H-O node topology, an input matrix of N I-dimensional input column vectors has size(input) = [ I N ] . The command output = sim(net,input) will yield an output matrix of N O-dimensional output column vectors with size(output) = [ O N ]. If output is binary, with at least one "1" in each column, vec2ind(output) will yield a row vector containing the row numbers of each unit element. and full(ind2vec(result)) will recover the original matrix.

rand('state',4151941)

output = round(rand(3,4))

result = vec2ind(output)

output = [ 1 0 0 0; 1 1 1 0; 0 0 1 1]

result = [ 1 2 2 2 3 3 ]

full(ind2vec(result)) = [ 1 0 0 0; 1 1 1 0; 0 0 1 1]

Hope this helps.

Greg