Adding values to a cell array in a loop for wav files

I am reading in a wav file in a loop, condcuting and power spectral analysis, and then identifying peaks (by frequency) in a specific band. At the end I need to concatenate the values and spit out a full table that includes all files analyzed. I'm having trouble with " peaks(i) = [pks freq]". I am getting this error "Unable to perform assignment because the indices on the left side are not compatible with the size of the right side." See code:

for i = 941:24:1608; % Start at 4am to avoid chorusing fish or vessels
    fprintf('Processing %s\n', char(strcat(Dirin,"\",filelist(i))))
    filename = filelist(i);
    [filepath,name,ext] = fileparts(filename);
    [y,Fs] = audioread(char(strcat(Dirin,"\",filelist(i)))); % read in wav file
    samples = length(y);
    dt = 1/Fs;
    t = (0:dt:(samples-1)*dt);
    [sensor_spectrum, freq] = pwelch(y,w,NOVERLAP,NFFT,Fs);
    sensor_spectrum_dB = 20*log10(sensor_spectrum) - sensitivity - gain ;
    
    sensor_spec_adj = sensor_spectrum_dB(292:1758); %subsetting db
    freq_adj = freq(292:1758); %subsetting frequency
    [pks, freq] =  findpeaks(sensor_spec_adj, freq_adj, 'MinPeakProminence', 10);
    peaks(i) = [pks freq]
    name_add = repmat({name},size(peaks(i),1),1)    
    peaks(i) = [name_add num2cell(peaks)]
end

Answer 1

[Star Strider]

Save it to a cell array instead —

peaks{i} = [pks freq]

That should work, since the two vectors being cncatenated will have the same sizes (both assumed to be column vectors).

.

[Benjamin Colbert]

This returns a 1x941 cell, as (i think) i=941

[Star Strider]

As I mentioned ‘(both assumed to be column vectors)’ however apparently, they‘re row vectors, so instead use:

peaks{i} = [pks; freq]

Testing this —

M = randi(9,5,10)

M = 5×10

4 5 6 2 9 4 9 6 4 1 2 9 1 2 2 2 6 4 6 3 4 1 6 2 6 9 3 1 1 3 3 7 3 4 5 9 6 3 7 9 3 8 6 2 2 3 5 6 8 3

freqvctr = linspace(0, size(M,2)-1, size(M,2))*0.1

freqvctr = 1×10

0 0.1000 0.2000 0.3000 0.4000 0.5000 0.6000 0.7000 0.8000 0.9000

for i = 1:size(M,1)
    [pks,freq] = findpeaks(M(i,:),freqvctr, 'MinPeakProminence',2.5);
    peaks{i} = [pks; freq];
end

peaks

peaks = 1×5 cell array
    {2×2 double}    {2×3 double}    {2×2 double}    {2×2 double}    {2×2 double}

peaks{1}

ans = 2×2

9.0000 9.0000 0.4000 0.6000

peaks{end}

ans = 2×2

8.0000 8.0000 0.1000 0.8000

The slight changed needed to work with row vectors seems to work.

.

Answer 2

[Yuqing Li]

Hello Benjamin,

[pks freq] as the output from function "findpeaks" will be an Nx2 matrix (N being the number of peaks found), but peaks(i) is a scalar. You can iniitialize peaks as a cell array:

peaks = cell(number of iterations,1);

and save [pks freq] in each iteration by:

peaks{i} = [pks freq];

[Benjamin Colbert]

peaks{i} = [pks freq]; ends up creating a 1x941 cell

[Yuqing Li]

What form of output do you want to get?

Is it a table with filenames in the first column, pks and freq in the second and third column respectively?