How to find X and Y coordinates of maximum gap between curves?

Please see the attached figure. In this figure, I want to find the values of X and Y coordinates where the two curve has maximum separation in Y-direction. In this figure, we can see that maximum gap is somewhere around 0.7 in the x-scale.

To do so, I simply take the subtraction of red data and blue data and obtain the gap value and corresponding X-value.

The gap value is delta Y is 6.7*10^7 and X-value is 0.027451.

However, the gap is not maximum at X = 0.027451. So, is there any way I can find the X-Y values of maximum gap between curves?

The data is also attached here with. It may not be same dataset as shown in figure, but a similar data.

I attach fig file for the same data for refrence.

Here is the code to obtain this type of plot.

But how to find maximum gap between curves?

clear; close all

load('rd.mat');
load('bl.mat');
X = flux1.vapa_flx_E
plot(X,rd,'-or');
hold on;
plot(X,bl,'-ob');
set(gca,'YScale','log','YTick',[10^2 10^3 10^4 10^5 10^6 10^7 10^8 10^9 10^10 10^11],'LineWidth',1.5,'FontWeight','bold',...
    'FontSize',14, 'YTicklabel',{'10^2','','10^4','','10^6','','10^8','','10^1^0',''});
dif1 = rd-bl;
[val_maxdif1, idx_maxdif1] = max(dif1);
delta_y = val_maxdif1;
x1 = X(idx_maxdif1);

Any help is greatly appriciated.

Answer 1

[Karim]

It depends on what you mean with the gap, notice that you plot it on a logarithmic axis. As a result from a graphical point of view the gap seems larger at 0.7 however if you make the axis linear you will see that the gap is bigger at the small X values

% load mat files
load(websave('rd', "https://www.mathworks.com/matlabcentral/answers/uploaded_files/1093100/rd.mat"))
load(websave('bl', "https://www.mathworks.com/matlabcentral/answers/uploaded_files/1093105/bl.mat"))
load(websave('X' , "https://www.mathworks.com/matlabcentral/answers/uploaded_files/1093090/X.mat" ))

% create figure
figure
plot(X,rd,'-or');
hold on;
plot(X,bl,'-ob');
set(gca,'YScale','log','YTick',[10^2 10^3 10^4 10^5 10^6 10^7 10^8 10^9 10^10 10^11],...
    'LineWidth',1.5,'FontWeight','bold','FontSize',14,...
    'YTicklabel',{'10^2','','10^4','','10^6','','10^8','','10^1^0',''});
grid on
title('Logarithmic Y axis')

figure
plot(X,rd,'-or');
hold on;
plot(X,bl,'-ob');
set(gca,'LineWidth',1.5,'FontWeight','bold','FontSize',14);
grid on
title('Linear Y axis')

dif1 = rd-bl;

figure 
plot(X,dif1)
grid on

update after extra comment of the OP

here i made an attempt to identify the 'gap' that the OP looks for

dif2 = (rd-bl)./bl;

figure
plot(X,dif2)
grid on

[maxGap, ixd] = max( dif2 );
maxGap_X = X(ixd);

disp(['max gap is found at X = ', num2str(maxGap_X)])

max gap is found at X = 0.692

[MP]

@Karim: "It depends on what you mean with the gap, notice that you plot it on a logarithmic axis. As a result from a graphical point of view the gap seems larger at 0.7 however if you make the axis linear you will see that the gap is bigger at the small X values"

I truly understand this. But, I want to find the difference in the gap.

So, is there any way (changing scale) we can find the difference in the gap between two curves.

[Karim]

hi, no need to create an "answer" you can just comment below the previous answer.

so if i understand what you mean you somehow want to find a sort of normalized gap?

@MP i updated the original answer, so that the max gap is found at 0.692

[Star Strider]

@Karim — Your websave solution to opening .mat files with the online Run feature is brilliant!

I wish I’d thought of that!

[Karim]

Thank you, I had some issues a while back because we are limited to the amount of files we can upload to the answers per day. Hence I looked a bit for an workaround.

For completness, I also use the matfile command to see which variables are inside the mat file.

load(websave('rd', "https://www.mathworks.com/matlabcentral/answers/uploaded_files/1093100/rd.mat"))
matfile('rd.mat')

ans = 
  matlab.io.MatFile

  Properties:
              Properties.Source: '/users/mss.system.KkVnZh/rd.mat'
            Properties.Writable: false                            
    Properties.ProtectedLoading: false                            
                             rd: [64x1 double]                    

  Methods