contourf plot value not corresponding to the legend bar value

I have a general question about a contourf() plot that I am plotting. The plot is shown in the screenshot below. What I am confused about is why different z-value (Level, in this case) have the same color. The plot only have two colors, blue and red, while there are many different values ranging from .22 to .26. For example, with Level value of .26, I am expecting it to be black, according to the bar on the right hand side, and .255 to be darker blue, and .2575 to be black as well, but it's only one shade of blue. It looks like anything above .25 is blue, and anything below is red. Can anyone shine some light on what is happening?

Here is my code.

close('all'), clear, clc
f1 = figure;
ax = axes(f1);
load('data.mat')
smplX = reshape(smplX,[],1);
smplY = reshape(smplY,[],1);
smplZ = reshape(smplZ,[],1);
xLimits = [round(min(smplX)-1) round(max(smplX)+1)];
yLimits = [round(min(smplY)-1) round(max(smplY)+1)];
[xq,yq] = meshgrid(xLimits(1):0.5:xLimits(2),yLimits(1):0.5:yLimits(2));
zq2 = griddata(smplX,smplY,smplZ,xq,yq,'linear');

Warning: Duplicate data points have been detected and removed - corresponding values have been averaged.

if any(isnan(zq2))
    % zq3 = griddata(smplX,smplY,smplZ,xq,yq,'nearest');
    % idx = isnan(zq2);
    % zq2(idx)=zq3(idx);
    zq2 = griddata(smplX,smplY,smplZ,xq,yq,'nearest');
end

Warning: Duplicate data points have been detected and removed - corresponding values have been averaged.

contourf(ax,xq,yq,zq2);
c = colormap(ax,flipud(turbo()));
ax.CLim = [0.22 max(zq2,[],'all')];
cInc = (ax.CLim(2)-ax.CLim(1))/256;
cInc = round((0.25 - ax.CLim(1))/cInc);
if cInc>size(c,1)
    cInc = size(c,1);
    disp(size(c,1));
end
% rectangle(ax,'Position',[-xLimits(1) yLimits(1) ...
%                           xLimits(2)-xLimits(1) ...
%                           yLimits(2)-yLimits(1)],...
%                         'FaceColor',...
%                         c(cInc,:),'LineWidth',2)
ax.XLimitMethod = "tight";
ax.YLimitMethod = "tight";
ax.Children = ax.Children(end:-1:1);
colorbar(ax,"eastoutside")

Answer 1

[Voss]

"It looks like anything above .25 is blue, and anything below is red."

This is happening because of the levels that contourf chose. Take a look at the levels by taking the second output from contourf, the contour object, and inspecting its 'LevelList' property:

f1 = figure;
ax = axes(f1);
load('data.mat')
smplX = reshape(smplX,[],1);
smplY = reshape(smplY,[],1);
smplZ = reshape(smplZ,[],1);
xLimits = [round(min(smplX)-1) round(max(smplX)+1)];
yLimits = [round(min(smplY)-1) round(max(smplY)+1)];
[xq,yq] = meshgrid(xLimits(1):0.5:xLimits(2),yLimits(1):0.5:yLimits(2));
zq2 = griddata(smplX,smplY,smplZ,xq,yq,'linear');

Warning: Duplicate data points have been detected and removed - corresponding values have been averaged.

if any(isnan(zq2))
    % zq3 = griddata(smplX,smplY,smplZ,xq,yq,'nearest');
    % idx = isnan(zq2);
    % zq2(idx)=zq3(idx);
    zq2 = griddata(smplX,smplY,smplZ,xq,yq,'nearest');
end

Warning: Duplicate data points have been detected and removed - corresponding values have been averaged.

[~,c_obj] = contourf(ax,xq,yq,zq2);
c = colormap(ax,flipud(turbo()));
ax.CLim = [0.22 max(zq2,[],'all')];
cInc = (ax.CLim(2)-ax.CLim(1))/256;
cInc = round((0.25 - ax.CLim(1))/cInc);
if cInc>size(c,1)
    cInc = size(c,1);
    disp(size(c,1));
end
% rectangle(ax,'Position',[-xLimits(1) yLimits(1) ...
%                           xLimits(2)-xLimits(1) ...
%                           yLimits(2)-yLimits(1)],...
%                         'FaceColor',...
%                         c(cInc,:),'LineWidth',2)
ax.XLimitMethod = "tight";
ax.YLimitMethod = "tight";
ax.Children = ax.Children(end:-1:1);
colorbar(ax,"eastoutside")

disp(c_obj.LevelList)

         0    0.0500    0.1000    0.1500    0.2000    0.2500

Those levels are not particularly suitable to your data. Only one of them (0.25) is within the range of data you care about. You can specify your own levels as input to contourf instead. For example:

f1 = figure;
ax = axes(f1);
load('data.mat')
smplX = reshape(smplX,[],1);
smplY = reshape(smplY,[],1);
smplZ = reshape(smplZ,[],1);
xLimits = [round(min(smplX)-1) round(max(smplX)+1)];
yLimits = [round(min(smplY)-1) round(max(smplY)+1)];
[xq,yq] = meshgrid(xLimits(1):0.5:xLimits(2),yLimits(1):0.5:yLimits(2));
zq2 = griddata(smplX,smplY,smplZ,xq,yq,'linear');

Warning: Duplicate data points have been detected and removed - corresponding values have been averaged.

if any(isnan(zq2))
    % zq3 = griddata(smplX,smplY,smplZ,xq,yq,'nearest');
    % idx = isnan(zq2);
    % zq2(idx)=zq3(idx);
    zq2 = griddata(smplX,smplY,smplZ,xq,yq,'nearest');
end

Warning: Duplicate data points have been detected and removed - corresponding values have been averaged.

[~,c_obj] = contourf(ax,xq,yq,zq2,'LevelList',[0 linspace(0.22,0.26,5)]);
c = colormap(ax,flipud(turbo()));
ax.CLim = [0.22 max(zq2,[],'all')];
cInc = (ax.CLim(2)-ax.CLim(1))/256;
cInc = round((0.25 - ax.CLim(1))/cInc);
if cInc>size(c,1)
    cInc = size(c,1);
    disp(size(c,1));
end
% rectangle(ax,'Position',[-xLimits(1) yLimits(1) ...
%                           xLimits(2)-xLimits(1) ...
%                           yLimits(2)-yLimits(1)],...
%                         'FaceColor',...
%                         c(cInc,:),'LineWidth',2)
ax.XLimitMethod = "tight";
ax.YLimitMethod = "tight";
ax.Children = ax.Children(end:-1:1);
colorbar(ax,"eastoutside")

disp(c_obj.LevelList)

         0    0.2200    0.2300    0.2400    0.2500    0.2600

I'm not sure that makes for a clear contour plot, but now you've got plenty of color variation.

[Quy]

Thanks!

[Voss]

You're welcome!