variable names

Hi there, i'm quite new to matlab and I'm trying to process some data I have in several textfiles.

These textfiles have names that are very similar.

e.g. A_P1_pitchAmplitude A_P2_pitchAmplitude A_P3_pitchAmplitude A_P4_pitchAmplitude A_P5_pitchAmplitude B_P1_pitchAmplitude B_P2_pitchAmplitude B_P3_pitchAmplitude B_P4_pitchAmplitude B_P5_pitchAmplitude C_P1_pitchAmplitude C_P2_pitchAmplitude C_P3_pitchAmplitude C_P4_pitchAmplitude C_P5_pitchAmplitude

I suppose it most be possible to make some kind of a loop in which the name of the datafile is variabale. In this way I could make my code more simple instead of having to copy-paste the same code for each file. Can somebody help me on how to do this?

This is what I want to do to the different files:

A_P1_pitchMinFile = importdata('A_P1_pitchMin.txt');
A_P1_pitchMin = A_P1_pitchMinFile(1,2); %minimale pitch 
A_P1_pitchMaxFile = importdata('A_P1_pitchMax.txt');
A_P1_pitchMax = A_P1_pitchMaxFile(1,2); %maximale pitch 

A_P1_pitchAmplitude = importdata('A_P1_pitchAmplitude.txt');

A_P1_pitch = A_P1_pitchAmplitude(:, 1:2);

for i = 1 : length(A_P1_pitch(:,1))
   if A_P1_pitch(i,2) < A_P1_pitchMin
       A_P1_pitch(i,2) = NaN;
   end
   if A_P1_pitch(i,2) > A_P1_pitchMax
       A_P1_pitch(i,2) = NaN;
   end 

end

Thank you! Kind regards, Luc

Answer 1

[Fangjun Jiang]

To construct your file name:

for i=1:5
  FileName=sprintf('A_P%d_pitchAmplitude',i);
  disp(FileName);
end

[Luc Nijs]

Thank you! This works. Now I have to be able to also change the 'A' into 'B' and 'C'.

[Oleg Komarov]

Did you try my version?

Answer 2

[RoJo]

Not sure this is what you mean, but as a start.

You can use dir function to read in all the files in a directory

files = dir('C:\Test'); % Or whatever is your directory with the files

You then have a struct array where you can access each filename by

for i=3:size(files, 1)
fileName = strcat('C:\Test\', files(i).name);
a = importdata(fileName);
end

[Luc Nijs]

Thank you!

This is a way to prevent that files have to be in the "current directory"? And now it just loads al the files in the directory, but what If I want to select certain files from that directory?

Answer 3

[Oleg Komarov]

Variable parts of the name

num = '1':'5';
let = 'A':'C';

Combine the variable parts

nNum = numel(num);
nLet = numel(let);
let  = repmat(let,nNum,1);
num  = repmat(num,1,nLet);

Build subnames

subn = reshape(sprintf('%c_P%c_pitch',[let(:).'; num]),[],nNum*nLet).';

Process each set of three files (amplitude, min, max)

for f = 1:size(subn,1)
    % Import
    data.(subn(f,:))         = importdata([subn(f,:) 'Amplitude.txt']);
    data.([subn(f,:) 'Min']) = importdata([subn(f,:) 'Min.txt']);
    data.([subn(f,:) 'Max']) = importdata([subn(f,:) 'Max.txt']);
    % Set min < MIN = NaN
    idx = data.(subn(f,:))(:,2) < data.([subn(f,:) 'Min'])(1,2);
    data.(subn(f,:))(idx,2) = NaN;
    % Set max > MAX = NaN
    idx = data.(subn(f,:))(:,2) > data.([subn(f,:) 'Max'])(1,2);
    data.(subn(f,:))(idx,2) = NaN;
end

Are you sure that:

pitch(:,2) < min(1,2) --> pitch(:,2) = NaN
pitch(:,2) > min(1,2) --> pitch(:,2) = NaN

what contains the first column of pitch?

[Luc Nijs]

Thank you!

I've tried it but i get an error message, stating:

??? Error using ==> vertcat

CAT arguments dimensions are not consistent.

I'm trying to find out what I did wrong...

the first column of pitch is a timestamp. Here I don't realy need it. I just want to "clean" the dataset, in which values below the pitch minimum (lowest not an the instrument that was recorded) and above the pitch maximum (highest note played) are considered to be errors in the pitch tracking.

[Oleg Komarov]

Enclose the code into (commenting out with % the comments):

function foo

num = ...

...

for

...

end

end

Save the function somewhere and launch it from the cmd window:

foo

Then report the full message error (especially the line where the error occurs)