To create Polar Shaded plot

X,Y,Z DATA ..all n*1 array

X is angle, with NEWS convention, Y is radius, Z is data.

I need shaded plot of data, onto a polar plot of radius information in Y angle information in X and data information in Z

Basically a rose plot sans the discreet histogram.

Sorry cant figure out any solution.

[Joachim Schlosser]

Could you please go to https://de.mathworks.com/products/matlab/plot-gallery.html and give the closest resemblance of what you need?

[Utsav]

Sorry I could not find the exact match. However I want to reproduce something like this exactly.

Answer 1

[Mike Garrity]

If you don't see a polar plot which does what you want, you can use the pol2cart function to get to any of the Cartesian techniques, such as the ones I described in this blog post.

% Made up data
npts = 200;
theta = 2*pi*rand(npts,1);
r = rand(npts,1);
v = cos(theta) .* sin(pi*r);

% Convert to cartesian
[x,y] = pol2cart(theta,r);

% Interpolate onto grid
[xg,yg] = meshgrid(linspace(-1,1,125));
F = scatteredInterpolant(x,y,v);

% Mask out extrapolation
b = boundary(x,y);
inmask = inpolygon(xg(:),yg(:), x(b),y(b));
vg = F(xg,yg);
vg(~inmask) = nan;

% Call pcolor
polar(nan,nan)
hold on
h = pcolor(xg,yg,vg);
h.EdgeColor = 'none';
colorbar

[Utsav]

Fantastic Sir...it solved my problem nearly. just two things I want the circles of constant radius visible and 0 degree as true north. Any help? Sorry am a newbie to this field.Thanks for your help though

[Mike Garrity]

First, I'm afraid I cut & pasted the wrong version when I saved that answer the first time, so you'll want to get the updated version. I left out the call to polar the first time!

For moving the 0 angle, you really want the new polarplot function as I talked about in this blog post. The problem is that it doesn't really work with this pol2cart trick, so I don't think that's going to work for you.

As for the grid lines being behind the pcolor, that's because I added the pcolor object last. To fix this, you need to move it down in the list. It is possible, but it's rather hacky. It look something like this:

% Made up data
npts = 200;
theta = 2*pi*rand(npts,1);
r = rand(npts,1);
v = cos(theta) .* sin(pi*r);

% Convert to cartesian
[x,y] = pol2cart(theta,r);

% Interpolate onto grid
[xg,yg] = meshgrid(linspace(-1,1,125));
F = scatteredInterpolant(x,y,v);

% Mask out extrapolation
b = boundary(x,y);
inmask = inpolygon(xg(:),yg(:), x(b),y(b));
vg = F(xg,yg);
vg(~inmask) = nan;

% Call polar & pcolor
polar(nan,nan)
hold on
h = pcolor(xg,yg,vg);
h.EdgeColor = 'none';
uistack(h,'down',29)
colorbar

That 29 in the call to uistack is a magic number involving the number of grid lines and labels, and you'd probably need to fiddle with it.

If you're really going to need a lot of customizing, you'd probably be better off looking at some of the polar functions on the file exchange. For example, this one looks like it might be a good starting point, but I've never tried it.

[Utsav]

Thanks Sir. It was helpful a lot.