Draw multiple surfaces from different tables(.mat) files. Such that the different .mat files have different Color and names should be displayed on the surface

Hello,

I would like to have to read multiple table (.mat) files from a folder and plot all the graphs in one figure like the rough diagram image I have attached. I have sent in the input files and the matlab code which runs all surfaces but does not have different color nor the names on the figure. I just want it easy to distinguish the the diffferent surfaces clearly.

Regards

Aman

[dpb]

So, what have you done and where, specifically, did you get stuck?

Appears all need to do is to define a color for each surface and apply that to each.

[Amanullah Khan]

Yes mr. dpb, I have uploaded a plot_from_tables.m file where if all the (fuels(.mat) files) are kept in a folder. and if you select this folder it will draw the surface for all the fuels but with a common Parula colormap. Yes, i would like to have different color and also the name of the file to be displayed in figure.

Answer 1

[Walter Roberson]

I already gave you instructions on how to do this, but you did not follow the instructions.

Also, you did not deal with the fact that you had scattered data.

mat = dir('*.mat'); 
figure( 'Name', 'IDT VS PRESSURE AND TEMPERATURE' );

colormaps = {@parula, @turbo, @hsv, @hot, @cool, @spring, @summer, ...
             @autumn, @winter, @gray, @bone, @copper, @pink, @jet, ...
             @lines, @colorcube, @prism, @flag};
colormaps = colormaps(randperm(length(colormaps)));

N = 25;

for K = 1:length(mat) 
    load(mat(K).name);
    table= table2array(Surface_table);

    a =cell2mat(table(:,1)); %temperature
    b =cell2mat (table(:,2)); %pressure
    c = cell2mat( table(:,3)); %IDT
    
    xd = double(a);
    yd = double(b);
    zd = double(c);
    
    [Xg, Yg] = meshgrid(linspace(min(xd),max(xd),N), ...
                        linspace(min(yd),max(yd),N) );
                    
    F = scatteredInterpolant(xd, yd, zd);
    Zg = F(Xg, Yg);
   
    zdnor = (Zg-min(Zg(:)))/(max(Zg(:))-min(Zg(:)));
    zdp = max(0, min(1, zdnor)); 
    zres = zdp*(1-eps);

    cmap = colormap(colormaps{K}());
    cidx = floor(zres * size(cmap,1))+1;
    zcol = reshape(cmap(cidx,:),size(Zg,1),size(Zg,2),3);
    
    surf(Xg, Yg, Zg, 'edgecolor', 'none', 'CData', zcol)
    hold on
    
end
hold off

This code clearly shows that you can apply different colormaps to different surfaces by scaling the z values into colormap indices and converting to RGB directly... just like I instructed before.

[Amanullah Khan]

Dear Mr Roberson,

But the surfaces obatained from surf command aren't as smooth from the prepareSurfaceData ( which gives a polynomial option for the curve of the surface). Also Is it possible to name the surface in the figure?

(As you could see in the image I attached the difference between the smooth curve and one with surf command)

Regards,

Aman

[Walter Roberson]

prepareSurfaceData does not create plots. But if you insist:

mat = dir('*.mat'); 
figure( 'Name', 'IDT VS PRESSURE AND TEMPERATURE' );

colormaps = {@parula, @turbo, @hsv, @hot, @cool, @spring, @summer, ...
             @autumn, @winter, @gray, @bone, @copper, @pink, @jet, ...
             @lines, @colorcube, @prism, @flag};
colormaps = colormaps(randperm(length(colormaps)));

N = 25;

for K = 1:length(mat)
    thisfile = mat(K).name; 
    load(thisfile);
    table= table2array(Surface_table);

    a =cell2mat(table(:,1)); %temperature
    b =cell2mat (table(:,2)); %pressure
    c = cell2mat( table(:,3)); %IDT
    
    [xd, yd, zd] = prepareSurfaceData(a, b, c);

    
    [Xg, Yg] = meshgrid(linspace(min(xd),max(xd),N), ...
                        linspace(min(yd),max(yd),N) );
                    
    F = scatteredInterpolant(xd, yd, zd);
    Zg = F(Xg, Yg);
   
    zdnor = (Zg-min(Zg(:)))/(max(Zg(:))-min(Zg(:)));
    zdp = max(0, min(1, zdnor)); 
    zres = zdp*(1-eps);

    cmap = colormap(colormaps{K}());
    cidx = floor(zres * size(cmap,1))+1;
    zcol = reshape(cmap(cidx,:),size(Zg,1),size(Zg,2),3);
    
    [~, basename, ~] = fileparts(thisfile);
    surf(Xg, Yg, Zg, 'edgecolor', 'none', 'CData', zcol, 'DisplayName', basename)
    hold on
    
end
hold off
legend show

You will find that nothing has changed about the surface shape, even though prepareSurfaceData is called. I suggest that you re-read the documentation for prepareSurfaceData: https://www.mathworks.com/help/curvefit/preparesurfacedata.html as you will see clearly that it does not "gives a polynomial option for the curve of the surface"

Now, it would certainly be possible to take the data and do some kind of surface fitting on it, and plot an interpolated version of the surface. This would require having a mathematical model to fit the surface to, and it is not something you have shown as happening in the past.

[Amanullah Khan]

Dear Mr. Roberson,

Sorry, I meant to say plot with surface fit. (plots with sfit)

The following syntaxes :

%% find polynomial surface fit
[xData, yData, zData] = prepareSurfaceData(b,a,c);

% Set up fittype and options.
ft = fittype( 'poly22' );   .............(This equation decides the degree of polynomial)

% Fit model to data.
[fitresult, gof] = fit( [xData, yData], zData, ft );

%% Plot fit with data.
%subplot( 1,2,1);
h = plot( fitresult, [xData, yData], zData);

Regards,

Aman

[Walter Roberson]

... I was supposed to just guess that you used surface fitting??

mat = dir('*.mat'); 
figure( 'Name', 'IDT VS PRESSURE AND TEMPERATURE' );

colormaps = {@parula, @turbo, @hsv, @hot, @cool, @spring, @summer, ...
             @autumn, @winter, @gray, @bone, @copper, @pink, @jet, ...
             @lines, @colorcube, @prism, @flag};
colormaps = colormaps(randperm(length(colormaps)));

N = 25;

for K = 1:length(mat)
    thisfile = mat(K).name; 
    load(thisfile);
    table= table2array(Surface_table);

    a = cell2mat(table(:,1)); %temperature
    b = cell2mat (table(:,2)); %pressure
    c = cell2mat( table(:,3)); %IDT
    
    % find polynomial surface fit
    [xData, yData, zData] = prepareSurfaceData(b,a,c);
    % Set up fittype and options.
    ft = fittype( 'poly22' );   .............(This equation decides the degree of polynomial)
        % Fit model to data.
    [fr, gof] = fit( [xData, yData], zData, ft );
    
    [Xg, Yg] = meshgrid(linspace(min(xData),max(xData),N), ...
                        linspace(min(yData),max(yData),N) );
    Zg = fr.p00 + fr.p01 .* Xg + fr.p01 .* Yg + fr.p20 .* Xg.^2 + fr.p02 .* Yg.^2 + fr.p11 .* Xg .* Yg;
   
    zdnor = (Zg-min(Zg(:)))/(max(Zg(:))-min(Zg(:)));
    zdp = max(0, min(1, zdnor)); 
    zres = zdp*(1-eps);

    cmap = colormap(colormaps{K}());
    cidx = floor(zres * size(cmap,1))+1;
    zcol = reshape(cmap(cidx,:),size(Zg,1),size(Zg,2),3);
    
    [~, basename, ~] = fileparts(thisfile);
    surf(Xg, Yg, Zg, 'edgecolor', 'none', 'CData', zcol, 'DisplayName', basename)
    hold on
    
end
hold off
legend show