2D color or surface plot based on 3 vectors of data

I want to do a 2D plot based on 3 vectors of data (X,Y and Z) coming from rig results, in order to have some color surfaces between the points that shows the trend to where the values are increasing or decreasing. Ideally should contain also some iso-lines appart from the Colors. Not sure how to achieve it but I have seen it published.

The closest I could get to what I want was to use a simple scatter (X,Y Z) plot. But when I tried to generate surfaces between them by using the ‘peaks’ function then I had the problem that my X and Y plots are not strictly increasing or decreasing, and that there were some repeated values, so I could not use the function ‘contourf’.

Any help is more than welcome!

Answer 1

[Pawel Jastrzebski]

How about this:

% interpolation data
% as in:
% https://uk.mathworks.com/help/matlab/ref/scatteredinterpolant.html?searchHighlight=scatteredInterpolant&s_tid=doc_srchtitle#responsive_offcanvas

F = scatteredInterpolant(datarig.X,datarig.Y,datarig.Z);
[xq,yq] = meshgrid(40:115,25:65);
F.Method = 'natural';
zq = F(xq,yq);

%colorbar data
limLow = min(datarig.Z);
limHi  = max(datarig.Z);
c = linspace(limLow,limHi,length(datarig.Z));

figure;
scatter3(datarig.X,datarig.Y,datarig.Z,[],c);
caxis([limLow, limHi])
colorbar('Limits',[limLow, limHi]);

hold on;
contourf(xq,yq,zq)
%surface(xq,yq,zq)

%view(90,90)

Output:

[Pawel Jastrzebski]

Switch the comments to see the 3D shape of this approximation:

contourf(xq,yq,zq)
%surface(xq,yq,zq)

[Jordi Riba]

Hi Pawel, that is very good !!! Thanks! Which view orientation should I take to only see the bottom part , and not the top? Thanks again !!!

[Pawel Jastrzebski]

Plot only the:

figure;
contourf(xq,yq,zq)

And nothing else. I added the other plots to see how the interpolation really worked.

[Jordi Riba]

Thanks Pawel!!! it works really well !!! Cheers! Jordi

Answer 2

[Pawel Jastrzebski]

In order to be able to use contour plot, your Z data need to be of X*Y size. What you can do is the following:

% C is a matrix of colours assigned to your data
% Color will get assigned based on the user specified condition
% Simple example with 3 colors RGB
% Red   = [1 0 0]
% Green = [0 1 0]
% Blue  = [0 0 1]
c = zeros(size(datarig,1),3);

cond1 = datarig.Z<=1.0e+09 *1; 
c(cond1,1) = 1; 
cond2 = (datarig.Z>1.0e+09 *1) & (datarig.Z<=1.0e+09 *1.15);
c(cond2,2) = 1;
cond3 = ~(cond1+cond2);
c(cond3,3) = 1;

% point size
s = 25;

figure
scatter3(datarig.X,datarig.Y,datarig.Z,s,c)
view(90,90)

To obtain something like:

[Jordi Riba]

Thanks Pawel, but this is not what I wanted. To get something similar I could just simple use a scatter, and I get what is shown on the attachment. I want to create extrapolated surfaces with iso-lines.