Obtaining mean values contourf

Hi,

I am using the contourf function to plot data inside a specific level. After this, I would like to evaluate the mean value at each level (not the centroid) to be then used for further analysis.

So far I am able to exctract the data inside a specific level but then I am not able to find the mean values.

Can someone help?

Thank you,

Carola

Answer 1

[DGM]

Consider the following example

% some fake data
A = reshape(1:256,16,[]);
A = A + 10*randn(size(A));
[min(A(:)) max(A(:))] % just show what the data range is

ans = 1×2

-16.3551 266.1712

% unless levels are known explicitly
% get them from the contour plot
[~,hc] = contourf(A); hold on

% first level corresponds to data minimum
% last level rarely corresponds to data maximum
ll = hc.LevelList

ll = 1×7

-16.3551 0 50.0000 100.0000 150.0000 200.0000 250.0000

% so make sure there are consistently enough boundaries to cover the data range
mx = max(A(:));
if mx > ll(end)
    ll = [ll mx];
end
ll % the full level list (the region boundaries)

ll = 1×8

-16.3551 0 50.0000 100.0000 150.0000 200.0000 250.0000 266.1712

% get the mean value of each region
nregions = numel(ll)-1;
lvlmean = zeros(1,nregions);
for k = 1:nregions
    if k < nregions
        % each region in a contourf() plot 
        % is associated with the lower boundary of the region
        mask = (A >= ll(k)) & (A < ll(k+1));
    else
        % make sure to include maximum values
        mask = (A >= ll(k)) & (A <= ll(k+1));
    end
    lvlmean(k) = mean(A(mask));
end
lvlmean

lvlmean = 1×7

-11.8971 25.6399 76.4761 127.1281 177.1200 222.5627 256.6190

... or you could choose to handle the region conditionals symmetrically if you wanted. That's up to you.