got error 'Attempt to execute SCRIPT pca as a function'

w = 1./var(jointdata);
[wcoeff,score,latent,tsquared] = pca(jointdata,...
'VariableWeights');
Attempt to execute SCRIPT pca as a function:
C:\Users\Adnan Gujjar\Desktop\MSProject\pca.m

i want to apply PCA in it why i got error can any one help me??

[Stephen23]

What do you get when you type this in the command window?:

which pca

Did you write this script/function? What is the first line of code in that Mfile?

[MUHAMMAD ADNAN]

dear sir i got Attempt to execute SCRIPT pca as a function: C:\Users\Adnan Gujjar\Desktop\MSProject\pca.m

Error in classifyexample (line 8) [wcoeff,score,latent,tsquared,explained] = pca(jointdata,...

[MUHAMMAD ADNAN]

this is whole ccode that i run in to matlab i have 8 variable numerial data.

load jointdata  
figure()
boxplot(jointdata,'orientation','horizontal');
C = corr(jointdata,jointdata);  
w = 1./var(jointdata);
[wcoeff,score,latent,tsquared,explained] = pca(jointdata,...
'VariableWeights',w);
c3 = wcoeff(:,1:3);
 coefforth = inv(diag(std(jointdata)))*wcoeff;
 I = c3'*c3;
 cscores = zscore(jointdata)*coefforth;
 figure()
plot(score(:,1),score(:,2),'+');
xlabel('1st Principal Component');
ylabel('2nd Principal Component');

[Jan]

@Muhammad: Please answer Stephens question: What do you get when to type this command in the command window:

which pca

Or even better:

which pca -all

Answer 1

[Walter Roberson]

Your file C:\Users\Adnan Gujjar\Desktop\MSProject\pca.m does not have its first non-comment line starting with

function SOMETHING = pca(SOMETHING)

It needs to do so if you want to call upon it to return a calculated result.

[MUHAMMAD ADNAN]

dear sir i can not understand please here is my all code can u correct it. i ma very new in Matlab

load jointdata

figure()
boxplot(jointdata,'orientation','horizontal');
C = corr(jointdata,jointdata);

w = 1./var(jointdata);
[wcoeff,score,latent,tsquared,explained] = pca(jointdata,...
'VariableWeights',w);
c3 = wcoeff(:,1:3);
 coefforth = inv(diag(std(jointdata)))*wcoeff;
 I = c3'*c3;
 cscores = zscore(jointdata)*coefforth;
 figure()
plot(score(:,1),score(:,2),'+');
xlabel('1st Principal Component');
ylabel('2nd Principal Component');

[Walter Roberson]

That part is not obviously wrong. The difficulty is that you are calling upon pca(), which requires that pca() be a function. Where is the code for the pca() function?

Is it possible that you stored this code you show here in a file named pca.m ? If so then when it got to the call to pca() it would attempt to use the current file to satisfy the call to pca (a recursive call) rather than calling into the Statistics Toolbox.

If you named your code pca.m then you need to rename it. Don't call it inv.m or zscore.m or plot.m or xlabel.m or ylabel.m either.

[MUHAMMAD ADNAN]

can u make some change in code where it is necssary because i can not understand what u say?. i show my whole code here make changing please. then paste here i run what u send me.

[Stephen23]

MATLAB has two kinds of M-files: scripts and functions, which are very different to use.

What you currently have may be script which you are trying to use like a function, which causes the error. You will need to show us the complete code for us to know this.

To convert a script into a function simply a line at the start of the M-file:

function [out1,out2,..outN] = pca(inp1, inp2.. inpN)

You can read lots of examples in the function documentation.

[Walter Roberson]

Do not put the line Stephen suggests at the top of your file!

Just take your file and rename it from pca.m to runpca.m

[MUHAMMAD ADNAN]

Dear i change filename i still face same problem. please look it and correct me code.

[Walter Roberson]

What filename are you using now? And exactly what error message are you getting now?

Answer 2

[Hassan Macanga]

 
clc
clear
 
input_dir = 'user/mac/desktop/python/eigenfacematlab/eigenface24/train';
testdir=  'user/mac/desktop/python/eigenfacematlab/eigenface24/test';
image_dims = [200, 180];
classes=[0;1;2;3;4;5;6;7;8;9];
 
 
num_images = 170;
images = [];
imageclasses=[]; 
disp(num_images);
counter = 1;
c = cell(170,1);
c{1}={''};
file=cell(170,1);
file(1)={''};
paths = fullfile(testdir,{'0','1','2','3','4','5','6','7','8','9'});
pathsA = fullfile(testdir,{'0','1','2','3','4','5','6','7','8','9'});
for i = classes(1):classes(10)
    
    filenames = dir(fullfile(input_dir,int2str(i),'\*.jpg'));
    num_images = size(filenames,1);
    for n = 1:num_images
     %filenames = dir(fullfile(input_dir, '*.jpg'));
     %disp(filenames)
     filename = fullfile(input_dir,int2str(i), filenames(n).name); 
     disp(filename)
     ij=cellstr(filename);
     k=cellstr(filenames(n).name);
     file{counter}=(k);
     disp(filename)
     img = imread(filename);
     img = rgb2gray(img);
     img = im2double(img);
     c{counter}=(1i);
     images(:, counter) = img(:);
     imagesclasses(:,counter)=i;
     counter=counter+1;
    end
     
end
    
 
 
num_images = 170;
counter = 170;
mean_face = mean(images, 2);
shifted_images = images - repmat(mean_face, 1, num_images);
 
 
[u, s, v] = pca(images');
num_eigenfaces = 20;
u = u(:, 1:num_eigenfaces);
 
features = u' * shifted_images;
testlabel=[];
cc=1;
testimages=[];
featuresvector=[];
predicition=[];
%testing
for i = classes(1):classes(10)
    filenames1 = dir(fullfile(testdir,int2str(i),'\*.jpg'));
    for n = 1:3
     filename1 = fullfile(testdir,int2str(i),filenames1(n).name);
     
     img = imread(filename1);
     img = rgb2gray(img);
     img = im2double(img);
     figure(1)
     imshow(img)
   % if n == 1
      %  images = zeros(prod(image_dims), num_images);
  %  end
     
     input_image(:, cc) = img(:);
     testlabel(:,cc)=i;
     
     feature_vec = u' * (img(:) - mean_face);
 
     similarity = arrayfun(@(n) 1 / (1 + norm(features(:,n) - feature_vec)), 1:num_images);
 
     % find the image with the highest similarity
     [match, matchidx] = max(similarity);
     A = reshape(images(:,matchidx), image_dims);
     Predicition(:,cc)=imagesclasses(matchidx);
     % display the result
     %figure
     %A=cell2struct(file,'name',1)
      if 3%n=0
         figure
         imshow([img reshape(images(:,matchidx), image_dims)]);
         title(sprintf('matches %f, score %0.2f', 1-match));
      end
    
     cc=cc+1;
     
    end
    
end
 
confusionMatrix=confusionmat(testlabel,predicition);
 
%show mean face
a = reshape(mean_face, image_dims);
figure
imshow(a)
 
 
% display the eigenfaces
figure;
for n = 1:num_eigenvalues
    subplot(2, ceil(num_eigenvalues/4), n);
    ejector = reshape(u(:,n), image_dims);
    images(ejector);
    colormap(gray);
end
 

[Hassan Macanga]

Attempt to execute SCRIPT eigenfaces as a function:

/Users/mac/Desktop/python /eigenfacematlab /eigenfaces.m

i keep in getting this error anyone to help out

[Rik]

This is not an answer, but a question. Please post it as your own question. Feel free to post a link to your question here.

For your reposted question, consider the following changes:

• either attach your m-file or use proper layout to make your question more readable
• post an exact description of what you are doing that produces this error (what is your current directory and how are you calling this script)

[Walter Roberson]

You have a script (not a function) named eigenfaces.m which you are attempting to invoke as if it were a function. However it is not clear to us where you are doing that, as you do not use eigenfaces in your code. Perhaps the file you posted is itself named eigenfaces.m and you are somehow invoking it as if it were a function.