Need help with indexing in a nested for loop

Hi everyone. I have to write a code that will apply a narrow band-pass filter around each frequency and stack the ccfs for different test velocities: This is my attempt at writing the code but it gives me the error Assingment has more non-singleton rhs dimensions than non-singleton subscripts. I would appreciate any kind of help. Thank you.

% ccf is a NxM matrix (153x4001), with N the amount of cross-correlation pairs and M the amount of time samples.
%ccf are cross-correlation functions

load ccf

%Define the testing velocity range
testVelocity = 100:1:300; 
freqVec = 7:40;
sampling_rate = 400;
% dist is a Nx1 vector with the interstation distance for each of the N pairs
dist = nonzeros(flipud(tril(repmat((85:-5:5)',1,85/5))))';
DispCurve=zeros(length(testVelocity), length(freqVec)); 

% Loop over frequencies
for fInd = 1:length(freqVec)
  freq=freqVec(fInd);
   [A,B]=butter(4,[7, 40]./(sampling_rate/2)); 
    ccf_filtered=filtfilt(A,B,ccf); % Narrow band-pass filter applied to each ccf
  % Loop over all velocities
   for ind = 1:length(testVelocity)
    % find the sample in each ccf that corresponds to the test velocity
    testVsamplePositiveLag=(M-1)/2 + round(dist./testVelocity(ind));
    testVsampleNegativeLag=(M-1)/2 - round(dist./testVelocity(ind));
    DispCurve(ind,fInd)=sum(ccf_filtered(:,testVsamplePositiveLag)) + sum(ccf_filtered(:,testVsampleNegativeLag)); % sum positive and negative lag times
   end
end

Answer 1

[Star Strider]

The simplest solution is just to add an extra dimension to ‘DispCurve’.

Preallocate as:

DispCurve=zeros(length(testVelocity), length(freqVec), size(ccf,1)); 

and the assignment in the loop is then:

DispCurve(ind,fInd,:)=sum(ccf_filtered(:,testVsamplePositiveLag)) + sum(ccf_filtered(:,testVsampleNegativeLag)); % sum positive and negative lag times

You can sort out the dimensions later to plot them or whatever you want to do with them.

If you want to keep it as a 2D matrix, adding second sum calls or, if you have a recent MATLAB release, adding 'all' to the sum call will create it as a scalar:

DispCurve(ind,fInd)=sum(ccf_filtered(:,testVsamplePositiveLag),'all') + sum(ccf_filtered(:,testVsampleNegativeLag),'all'); % sum positive and negative lag times

Experiment to get the result you want.