need to find local maxima and minima of each row of a 7886 * 321 matrix stored in xlsx file.

filename = 'filtereddata.xlsx';

data = xlsread(filename);

Diff1= diff(data,1,2);

[rows, columns] = size(data);

extrema = zeros(7886,321);

for i=1:rows

[Diff1max,imax1,Diff1min,imin1] = extrema(Diff1(i,:));

end

Error: Insufficient number of outputs from right hand side of equal sign to satisfy assignment.

Answer 1

[Voss]

extrema is a 7886-by-321 matrix of zeros, so this expression:

extrema(Diff1(i,:))

says to get the elements of extrema at the indices given by Diff1(i,:). Those elements will be in a single matrix, so trying to assign them to the four variables Diff1max,imax1,Diff1min,imin1 is what causes the error.

I suspect you intend to use the function extrema from the File Exchange (https://www.mathworks.com/matlabcentral/fileexchange/12275-extrema-m-extrema2-m), in which case you should avoid naming a variable the same name as that function. Use another name for the variable extrema, say my_extrema, if you need that variable (which it's not clear whether you do).

my_extrema = zeros(7886,321); % variable
for i=1:rows
    [Diff1max,imax1,Diff1min,imin1] = extrema(Diff1(i,:)); % function call
    % do something with the outputs from extrema function
    % Diff1max,imax1,Diff1min,imin1 
end

[NotA_Programmer]

I suspect you intend to use the function extrema from the File Exchange (https://www.mathworks.com/matlabcentral/fileexchange/12275-extrema-m-extrema2-m)

=Yes

Can you please help how to use that variable my_extrema in this function-

[Diff1max,imax1,Diff1min,imin1] = extrema(Diff1(i,:));

[Voss]

Well, my_extrema is all zeros at that point, so I don't think you need to find the minima and maxima of that. (The outputs from the extrema function applied to my_extrema will all be zeros.)

I don't know if finding the extrema of the diff of data is appropriate or not or whether you should find the extrema of data itself instead (I don't have your data or know what it represents or know what you are trying to do with it), but it's clear (according to the comments at the top of extrema.m) that the extrema function operates on a vector, so you seem to be doing the right thing by passing in a vector to the function extrema.

Maybe try a few things and make a few plots to see how the extrema function should be applied to data:

% filename = 'filtereddata.xlsx';
% data = xlsread(filename);

% I make up some data, you can use yours from the xlsx file:
data = 1000*sin(2*pi*(1:7886).'/4000)*cos(2*pi*(1:321)/500);


% then use this code below as-is to see what the 
% plots look like for your data:

subplot(2,2,1)
plot(data(1,:))
hold on
[xmax,imax,xmin,imin] = extrema(data(1,:));
plot(imax,xmax,'go') % mark maxima with green circles
plot(imin,xmin,'ro') % mark minima with red circles
title('1st row of data')

subplot(2,2,2)
plot(data(:,1))
hold on
[xmax,imax,xmin,imin] = extrema(data(:,1));
plot(imax,xmax,'go')
plot(imin,xmin,'ro')
title('1st column of data')
Diff1 = diff(data,1,2);

subplot(2,2,3)
plot(Diff1(1,:))
hold on
[xmax,imax,xmin,imin] = extrema(Diff1(1,:));
plot(imax,xmax,'go')
plot(imin,xmin,'ro')
title('1st row of diff(data,1,2)')

subplot(2,2,4)
plot(Diff1(:,1))
hold on
[xmax,imax,xmin,imin] = extrema(Diff1(:,1));
plot(imax,xmax,'go')
plot(imin,xmin,'ro')
title('1st column of diff(data,1,2)')

[NotA_Programmer]

It is working well when used with single row or column. But when I am using it in for loop it gives error, I want to apply this extrema function on every single row.

filename = 'filtereddata.xlsx';

data = xlsread(filename);

Diff1 = diff(data,1,2);

[rows, columns] = size(data);

[Diff1max,imax1,Diff1min,imin1] = extrema(Diff1(1,:));

However while applying on multiple rows below code seems to work fine but this "Diff1max,imax1,Diff1min,imin1" shows the matrix of "1 x 21 Double" instead of "7886 x 21 Double".To resolve this I used "extrema = zeros(7886,321);" but this does not seems to work.

filename = 'filtereddata.xlsx';

data = xlsread(filename);

Diff1= diff(data,1,2);

[rows, columns] = size(data);

for i=1:rows

[Diff1max,imax1,Diff1min,imin1] = extrema(Diff1(i,:));

end

[Voss]

The extrema function is intended to work on a vector, not a matrix (according to its documentation, which I have not verified), so on each iteration of your for loop, you will get vectors for Diff1max, imax1, Diff1min, imin1. The sizes of the vectors might be different on each iteration of the loop (i.e., maybe your Diff1 has different numbers of extrema in each row), so it may not be possible to construct a matrix of those extrema (and in general it will not be possible). One way to handle such a situation is to use cell arrays:

clearvars extrema

filename = 'filtereddata.xlsx';
data = xlsread(filename);

Diff1= diff(data,1,2);
[rows, columns] = size(data);

Diff1max = cell(rows,1);
imax1 = cell(rows,1);
Diff1min = cell(rows,1);
imin1 = cell(rows,1);

for i=1:rows
    [Diff1max{i},imax1{i},Diff1min{i},imin1{i}] = extrema(Diff1(i,:));
end

Then Diff1max{1} will be a vector of the max values of the first row of Diff1, and Diff1max{2} is the maxima of the second row, and so on.

[NotA_Programmer]

"One way to handle such a situation is to use cell arrays:"

- Yes, this worked.

Thanks a lot!!

[Voss]

You're welcome!