How to create contour plot to show height?

I want to create a countour plot such that my x axis and y axis are the coordinates of the plane and my z axis is height. I want to creat a 2b plot (contour) such that my height is differentiated with color. My z value has no relation with xy value other that the first value of z is corresponding to the cobination of first two values of x and y.

For example my x, y and z vectors are:

AAA=1:1:1000;

BBB=1:1:1000;

CCC=2:2:2000;

[DGM]

if there's no correlation between Z and X,Y, then you don't have a meaningful description of a 2D dataset.

If the x,y,z vectors are correlated, but scattered, then use griddata() to create gridded data that's compatible with what contour() expects.

[Muhammad Qaisar Fahim]

The correlation is that for each pair of x and y there is a corresponding value of z but the limitataion is that we only have one to one correspondance i.e., 1st value of x and first value of y corresponds to first value of z but first value of x and 2nd value of y corresponds to nothing.

[DGM]

Okay, yeah. That's just scattered data. Walter's suggestion looks pretty good.

Answer 1

[DGM]

You can always try just using griddata()

S = load('scattered.mat'); % load scattered data

n = 100;
x = linspace(min(S.x),max(S.x),n);
y = linspace(min(S.y),max(S.y),n);
[xg yg] = meshgrid(x,y); % create new grid
zg = griddata(S.x,S.y,S.z,xg,yg); % interpolate scattered data to grid

contour(xg,yg,zg) % now you can use contour(), contourf(), surf(), etc

[Muhammad Qaisar Fahim]

Let me explain my question again. I have 100 x values, 100 y values and 100 z values. Where x values represent x axis, y values represent y axis and z values represent z axis. Each z value correspond to pair of x and y value. Now I want to plot such that for different values of z I get different color. If z value is same e.g., 10 for 3 points then color should be same. I want to color code based on height. How can I do that ?

[DGM]

Are you sure you want an actual contour plot?

S = load('scattered.mat'); % load scattered data

n = 100;
x = linspace(min(S.x),max(S.x),n);
y = linspace(min(S.y),max(S.y),n);
[xg yg] = meshgrid(x,y); % create new grid

zg = griddata(S.x,S.y,S.z,xg,yg); % interpolate scattered data to grid

contourf(xg,yg,zg)

figure
pcolor(xg,yg,zg)
shading flat

Or maybe you want something else?

figure
S = load('scattered.mat'); % load scattered data

scatter3(S.x,S.y,S.z,10,S.z,'filled')

[Muhammad Qaisar Fahim]

Actually the data has latitude(x) and longitude(y) of a road and also I have height of the road (z). Now I want to construct a plot which has shows height too and that height should be color coded.

[Muhammad Qaisar Fahim]

I need something like this

[Sam Chak]

Hi @Muhammad Qaisar Fahim

You better share/attach the spatial data longitude A, lattitude B, elevation C (road height) here in the comments and edit your original post, so that you can get the proper help that you want.

You also need to confirm if the following conversion gives the geometry of the road.

x = (6371).*cos(B).*cos(A);
y = (6371).*cos(B).*sin(A);
scatter(x, y)

[DGM]

So you basically want variable-color lines. There are two basic ways to do that.

First, you could use a very dense scatter plot to create pseudolines. This is the simple way, and the data doesn't need to be ordered.

cmap = parula(64);

N = 1000; % how many points to plot?
x = linspace(0,10,N); % a straight line
y = linspace(0,10,N);
z = sin(linspace(-pi/2,7*pi/2,N)); % sinusoidally-varying z-height

h = scatter(x,y,10,z,'filled');

colormap(cmap);
colorbar

Alternatively, you can use a surf() plot. This will actually create lines, so the data needs to be ordered.

cmap = parula(64);

N = 100; % how many points to plot?
x = linspace(0,10,N); % a straight line
y = linspace(0,10,N);
z = sin(linspace(-pi/2,7*pi/2,N)); % sinusoidally-varying z-height

h = surf([x(:) x(:)],[y(:) y(:)],[z(:) z(:)]);
set(h,'facecolor','none','edgecolor','interp');
set(h,'linewidth',3); % make it fat so it's easier to demonstrate

view(2); % only show 2-D view
colormap(cmap);
colorbar

That said, I don't know how exactly this data is arranged -- how dense it is, etc. It may become complicated if the points are relatively sparse and unordered.

[Sam Chak]

Hi @DGM

His data of the position of the road (in Spherical coordinate system) was posted here in the first comment of the first answer (by @KSSV).

https://www.mathworks.com/matlabcentral/answers/1676344-how-to-plot-a-road-along-its-height

[DGM]

Oh. Well it doesn't make much sense spending time and confusion here then. Thanks for the heads up.

Answer 2

[Walter Roberson]

See https://www.mathworks.com/matlabcentral/fileexchange/38858-contour-plot-for-scattered-data

[Muhammad Qaisar Fahim]

Hi Walter,

I am using this link ( https://www.mathworks.com/matlabcentral/fileexchange/10408-contours-for-triangular-grids ) I am getting an error on it :

Error using tricontour (line 58)

Incorrect input dimensions

Error in runMatlogOSM_2nodes (line 25)

tricontour(A,B,C,1)

Eventhough I have downloaded the function and the path as

addpath(genpath('./tricontour'), '-end');

My Command line is

tricontour(A,B,C,1)

[Walter Roberson]

The first parameter to tricontour() needs to be the xy coordinates, each row being an x y pair.

The second parameter to tricontour() needs to be the triangular connectivity data.

You might want to use https://www.mathworks.com/help/matlab/ref/delaunaytriangulation.html and extract the Points and ConnectivityList properties of the result.