I am getting an empty figure for a given code

Hello, I have been given the following code and the data as well. However when i run it, the figure is empty. I cant find the problem and there is no error. (maybe the problem is with the findpeaks function)

Can someone help me?

Thank you !!

load modulation_125911_0dBAtt_1kOhm_100kOhmPort6.mat

set(0, 'DefaultAxesFontSize',14)
set(0, 'defaultLineLineWidth', 1.5)

R = data{1}+1i*data{2};
Rs = squeeze(R);

phi = scan.loops(1).rng;
phi = linspace(phi(1), phi(2), scan.loops(1).npoints);
freq = scan.loops(2).rng;
freq = linspace(freq(1), freq(2), scan.loops(2).npoints);

ref1x = [1, 402];
ref2x = [1, 402]; % just so i dont get an error... should be adjusted when i see the plot!
w = .5*tanh((freq-3e9)/5e7)+.5;
phase = exp(1i*(freq-freq(1))*-7e-10);


ref = w.*mean(data{1}(:, ref2x(1):ref2x(2))') ...
    + (1-w).*mean(data{1}(:, ref1x(1):ref1x(2))') ...
    +1i * ( w.*mean(data{2}(:, ref2x(1):ref2x(2))') ...
    + (1-w).*mean(data{2}(:, ref1x(1):ref1x(2))'));

offset = 0;
R = data{1}+1i*data{2};
Rcorr = (R.*repmat(phase, length(phi), 1)'  - offset) ...
    ./repmat(ref.*phase - offset, length(phi), 1)';

% Load flux/frequency data
phi = scan.loops(1).rng;
phi = linspace(phi(1), phi(2), scan.loops(1).npoints);
ph = phi(1:end-1);

freq = scan.loops(2).rng;
freq = linspace(freq(1), freq(2), scan.loops(2).npoints)./1e9;

% Calculate crude estimate for mA to Phi_0 conversion
%[~, ind] = max(diff(Rs, 1, 2));%
%[~, ind] = min(Rcorr);%
%osc = freq(ind);
%[~, ind] = min(osc(lB:uB));

Rdiff = diff(Rs,1,2);
osc = [];
phi_red = [];
for i=1:length(phi)-1
    [a, locs] = findpeaks(abs(Rdiff(:, i)), 'MinPeakProminence', 0.07);
    
    if(~isempty(locs))
        if(length(locs) > 1)
            [~, locs] = max(abs(Rdiff(locs, i)));
        end
        
        % cut away a lower frequecies
        if(freq(locs) > 5.8)
            osc = vertcat(osc, freq(locs));
            phi_red = vertcat(phi_red, phi(i));
        end
    end
end

osc = smooth(osc)';
% osc = vertcat(osc(50:160), osc(597:706));
figure(1);
plot(phi_red, osc,'.');
drawnow;

Answer 1

[Mara]

Hello,

Your variable phi_red is empty. Because your variable loc is empty, it does not run the if(~isempty(locs) command and replace your variable that you previously defined as phi_red = [ ];

The problem seems to be the specification 'MinPeakProminence', 0.07, with this, no 'relevant' peaks are found.

[Rik]

To expand on this answer: the largest prominence that is found is 0.00220830. If you lower that value you will see a plot.

[Roudy Hanna]

Hello Mara,

Thank you for your answer. I tied to change the threshold of 0.07 to different values(smaller and bigger)... but no image still.

Is there a way around this? (I'm relatively new to Matlab)

[Mara]

If you leave away 'MinPeakProminence', 0.07 completely, or , as Rik suggested lower the threshold to 0.002 or lower, you will get one!

[Mara]

But if this helps you depends on what you want to find out in your dataset, too.

As your threshold was much higher than the highest peak in your data, I suppose that data values like yours were not expected by who wrote the script. So I suggest, even if you get a plot, you should check if everything is right with your values!

[Roudy Hanna]

Hello Rik,

Thank you for your help. However, I am still not getting anything :/ !!! I also tried to leave away this section as Mara suggested, but also nothing...

[Roudy Hanna]

Hey Mara,

you're totally right !! The thing is I was hesitant to change it since the script as well as the data are part of an experimental course. I also have plenty of other "modulation" data like the one i uploaded; but not a single one is working with this script. I will check everything again but it's been nearly 4 days that I am stuck on this part

Thank you for everything !!

[Mara]

I got a plot, but I also left away these commands, since I do not have the necessary toolbox

% osc = smooth(osc)';
% osc = vertcat(osc(50:160), osc(597:706));

No worries, good luck!

[Roudy Hanna]

Hey, i closed matlab and re-opened it again, and followed your suggestions, now it works !!

Thanks to both of you Mara and Rik.

[Mara]

That's good to hear! No problem!

(Sometimes it is good to have clear; at the start of the code so no old variables are in the workspace, maybe that was the problem)

[Rik]

It is better to work with functions, which will automatically keep your workspace clear, and will even warn you about variables you're not using (which might point you to a typo).

[Roudy Hanna]

Deal, I will make sure to define it as a function then!!

Many thanks!!