How to plot dynamic amount of contour plots within one figure

Hello,

I have a Problem and I need your help.

I have a Matlab Script that calculates three data tables that I need for plotting with contour().

Now I've managed to find a way that Matlab saves each of those 3 Datatables within a folder so that I am able to regain controll over those Informations after I've already plottet other plots with different input arguments.

My goal is that, e.g. I have plotted 5 Plots from which I've also have its data saved for each individual plot. Now I would be able to plot these again but I want to combine all of these Plots into one single Figure.

In the end I just want all 5 Plots within 1 Figure because I need to find out a way to create one new Plot possibly containing informations from all other plots combined before, therefore I wanted to use interpolation. From those 5 Plots that I've put over each other within one Figure I need to find out all of the 'lowest' points within this combined figure in order to create a new final plot containing all the relevant information that I wanted.

Thanks in advance

[Dyuman Joshi]

You can extract the coordinates of the plots from the figures, and use those to plot them together and get the lowest points.

[Mathieu NOE]

maybe you should share your data and code if you have one

[Frederic von Altrock]

@Dyuman Joshi I might have formulated my demands wrong but you've helped me by saying that I can extract all the points. In order to get the final plot that is combined from multiple other ones I not only need the lowest points but If you would lay all the plots over each other I just want the Line that is the lowest. F.e.: The red-Line would be the resulting plot from all the 3 different ones that you can see:

@Mathieu NOE I could, but I'm not sure how helpful that would be (please correct me if I'm wrong). Right now I've got a script, not written by me, that gives me the appropriate matrices for each plot. You can change the parameters and the plot is different every time. The way I store the data is by using 'writematrix(...)' for each of the three 21x21 double matrices. At the moment, this only helps me to replot the plots, but I still haven't found a way to combine them the way I want.

[Mathieu NOE]

@Frederic von Altrock

see my little demo in the answer section

all the best

[Frederic von Altrock]

Thank you in advance for your help! I will take a look at it right away.

Answer 1

[Mathieu NOE]

@Frederic von Altrock

I modified a bit your combine function (see attachment)

I am not sure I generated the best scenario here so there maybe still some work to get exactly what you want

so I get these 3 original curves

and the final result (for the time being) looks like this

please double check and get back to me to further refinement

[Mathieu NOE]

it could be a good idea to work with your parameters (and generate the correct curves)

can you provide the parameters (that are in main.m) for different tools ? (or create several main.m functions : main1, main2 , etc...) ?

[Frederic von Altrock]

Hey, thank you so much for your extensive help! I will take a look at it shortly.

All the best

[Frederic von Altrock]

@Mathieu NOE it could be a good idea to work with your parameters (and generate the correct curves)

Changing only the w0x and w0y parameters for 3 different toolnames would give you the kind of curves and functionality of the script I am looking for.

I will try to get my hands on different parameters if that would make a difference in the plot and let you know as soon as I get them. Still, your modifications worked great for my plots. The red line works correctly and marks all the minimum points. I didn't know you could extract all the data from C1 like you did, plotting is kinda new to me.

[Mathieu NOE]

hello again

it's exactly what I did : change only w0x and w0y parameters (but maybe not with the correct amount, add or retrieve ? )

[Frederic von Altrock]

Hey, I'm sorry. I did not want to blame you. Maybe a range would have been helpfull ^^ I've only changed both Values between 850 and 1.550

Otherwise (havent tried that one out yet) you can also change the zeta x and y value between 2-4%.

Additionaly I've also found some for-loop iteration bugs in my code, maybe that also intervened with your plots. Anyways, you've helped me quite much already. I will continue to work with what I've got now and come back when I've got a more concrete Question again.

[Mathieu NOE]

ok , at your disposal for further development if needed !

[Frederic von Altrock]

Hey, thank you for continuing to offer your help! There might be one thing you can help me with. I have attached a file to this comment that contains the corrected code. You also have some simulations of mine so you can create a plot that looks the same as mine. I also attached the main.m script, which is only used to create new plots and write them to your local directory.

So to plot my plots you can just open combine and run it. When you do, three different plots should open (you can plan on not showing them in the future, but for now they are there). Be sure to clear all windows before you run it again, as it starts bugging when I do.

The legend you have in the combined plot cannot be displayed in the min plot. Matlab sees it as a single line, so it cannot assign the legend to the plots. Maybe you have an idea how to solve this.

Anyway, your modifications have already helped me a lot. Thank you in advance, have a nice weekend!

[Mathieu NOE]

hello again Frederic

see my suggestion in attachment

I think you can create the last plot with a much simpler code (as I suggest)

all the best

this is my result for the 3rd plot

[Mathieu NOE]

hello Frederic

did my last suggestion fullfill your needs ?

[Frederic von Altrock]

Hey, sorry I'm only getting back to you now. I'm approaching my exams and haven't had much time these days. I already saw your suggestion last Friday and was surprised how simply and concisely you summarized my code part, simply forgot to reply.... I was able to take a lot from your suggestions :) and yes, my requirements for the program are more than fullfilled since the legend is fixed now. Many thanks again for your help, time and effort!

Best regards, Frederic

[Mathieu NOE]

ok , wish you success for your exams !!

Answer 2

[Dyuman Joshi]

In that case, concatenate the extracted points and use min() -

%Random data
x = 0:0.01:5;
y1 = sin(x);
y2 = cos(x);
y3 = abs(2.5-x)-1;

%plot the lines
plot(x, y1, 'g--', 'LineWidth', 2.5)
hold on
plot(x, y2, 'c-', 'LineWidth', 2.5)
plot(x, y3, 'b-.', 'LineWidth', 2.5)

%Concatenate and find the minimum
out = min([y1; y2; y3], [], 1);

%Plot the minimum line
plot(x, out, 'r', 'LineWidth', 2)
ylim([-2 3])

[Frederic von Altrock]

Hey, thanks for your help and additional answer, but unfortunately I'm not able to use the min(..) function like you did. I have already explained my problem to @Mathieu NOE in one of the answers below, thanks in advance!

Answer 3

[Mathieu NOE]

here's a small demo code for you

adapt to your data (horizontal or vertical concatenation depending if your data is row or column oriented)

% create some dummay data
x = (1:100)';
y1 = 0.5*exp(-(x-20).^2/100);
y2 = exp(-(x-40).^2/50);
y3 = 1.3*exp(-(x-60).^2/50);

% pick min values of all data 
y_all = [y1 y2 y3]; % concat data
ymin = min(y_all,[],2); % pick the min of them

%plot
semilogy(x,y1,x,y2,x,y3)
hold on 
semilogy(x,ymin,'r-*','linewidth',2)

[Frederic von Altrock]

Hey, thanks for your answer! Unfortunately, this does not help me with my task.

I've been trying to work with min(..), but fortunately the values within the matrices I'm using are apparently structured and read differently, and therefore I can't just use "concat" and "min(..)" in the same way you do.

First of all, there are 3 matrices I'm using for each plot. Each matrix contains 21x21 double values. My plots are all created like this: contour(matrice1, matrice2, matrice3,[1, 1], 'k');

matrice1 contains x-axes, matrice2 contains the Range from 0.5 to 10 and matrice3 contains the values resulting in the lines.

As I said, I didn't write the script that returns me these matrices, I've been told that it's not relevant for me to know, but I'm not so sure about that anymore. I've managed to create a structure that allows me to dynamically save plots as .csv files on my PC. I'm able to replot them with the information they contain, and I've also figured out a way to combine all the plots. See it here:

Now I've tried to use the "min(..)" function as you described. My Code looks like this (Figure1 is created within a while(..) loop before this step) :

% Gemeinsame Darstellung aller Plots am Ende
figure;
hold on;
for k = 1:counterWithin
    C1 = contour(plotsInfo(k).dataOutput1, plotsInfo(k).dataOutput2, plotsInfo(k).dataOutput3, [1, 1], 'k');
    title('Combined Plot');
end

% Initialisiere leere Arrays für die Daten
concatenatedDataOutput3 = [];
for k = 1:numel(plotsInfo)
    % Concaten der Daten für das 3te Feld
    concatenatedDataOutput3 = [concatenatedDataOutput3; plotsInfo(k).dataOutput3(:)];
end
% pick minimum values
concatenatedDataMin = min(concatenatedDataOutput3,[],2);

hold off;

figure;
contour(plotsInfo(1).dataOutput1, plotsInfo(1).dataOutput2, concatenatedDataMin, [1, 1], 'r-*', 'LineWidth',2);

It just opens a plain "Figure3" window containing nothing. I guess I concated my values wrong.

I might I have to get my information about the lowest points with a different kind of method. Or in this case I am doing it wrong.

[Mathieu NOE]

well, as i said in my first comment, it would definitively help to access your data (as mat file)

[Frederic von Altrock]

Put them all in a separate folder if you want to run them.

Within main you can change the toolname, w0x and w0y to simulate random plots.

When you run combine, it will make the combined plot from all the simulated plots you ran in main before.

Thanks in advance!

[Mathieu NOE]

see my answer below

Answer 4

[Shubham]

Hi Frederic,

To achieve your goal, you can follow these steps:

1. Load the saved data tables for each of the individual plots from the folder.
2. Create a figure and use subplot or tiledlayout and nexttile (for MATLAB R2019b or later) to arrange the plots in a single figure.
3. Use contour to plot each data table in its respective subplot.
4. To find the 'lowest' points across all plots, you can combine the data into one matrix and find the minimum values.
5. Use interpolation if necessary to align the data points from different plots.
6. Create a new plot with the combined information.

Here's an example MATLAB script that outlines this process:

% Step 1: Simulate loading the saved data tables
% For this example, I'll create some synthetic data and save it as if they were your actual data files
for i = 1:5
    [X, Y] = meshgrid(linspace(-3, 3, 100), linspace(-3, 3, 100));
    Z = peaks(X, Y) + i; % Slightly different peaks for each dataset
    filename = sprintf('data%d.mat', i);
    save(filename, 'X', 'Y', 'Z');
end

% Step 2: Create a figure and arrange subplots
figure;
t = tiledlayout(3, 2); % We have 5 individual plots and 1 combined plot

% Step 3: Plot each data table in its respective subplot
% Initialize a matrix to store the combined lowest points
combinedZ = inf(size(Z)); % Initialize with infinities for finding minimum

for i = 1:5
    filename = sprintf('data%d.mat', i);
    loadedData = load(filename);
    
    nexttile; % Create a new tile for each individual plot
    contour(loadedData.X, loadedData.Y, loadedData.Z);
    title(sprintf('Plot %d', i));
    
    % Step 4: Combine the data to find the 'lowest' points
    combinedZ = min(combinedZ, loadedData.Z); % Element-wise minimum
end

% Step 6: Create a new plot with the combined information
% Span the new plot across two tiles at the bottom of the layout
nexttile(5, [1, 2]); % This will place the combined plot in the last row, spanning two columns
contour(loadedData.X, loadedData.Y, combinedZ); % Use the last loaded X and Y for plotting
title('Combined Plot with Lowest Points');

% Optionally, clean up by deleting the synthetic data files
for i = 1:5
    filename = sprintf('data%d.mat', i);
    delete(filename);
end

In this script, we first simulate the situation by creating 5 synthetic datasets and saving them as .mat files. We then create a figure with a tiled layout to accommodate all the individual plots and one combined plot. Each individual plot is created in a loop where we also keep track of the lowest points across all datasets. Finally, we create a combined contour plot that shows the lowest points.

Please note that in the combined plot, combinedZ contains the minimum value from each corresponding point across all datasets. This is visualized in the final contour plot titled "Combined Plot with Lowest Points".

[Frederic von Altrock]

Hey, as I mentioned in my answer above, I may have misspoken my requirements for the final plot. Thanks for your reply though! I was able to use some of your code and implement it in my script which resulted in these plots:

You can see an image in my answer above that shows how I would like to combine my plots to create my desired result plot. I'm currently working on a way to combine all of these into a single plot and use e.g. interpolation to get all my corresponding points to create the plot I want.

Thanks in advance for your help!