Is there a way to get shaded error regions on a scatter plot?
Show older comments
Hi folks,
I have a scatter plot with known errors for each datapoint. Is there a way to create a shaded region around the scatter plot? Below is my graph, the red and cyan regions are the errors and the blue is the data. Is there a way to shade them and make them a continuous line, like the graph below?
I've attached the table below, along with my code.
Thanks
figure;
hold on;
scatter(myData.Anisotropic_1, myData.AnisoMin);
scatter(myData.Anisotropic_1, myData.AnisoMax);
scatter(myData.Anisotropic_1, myData.("MPC aniso"), 'b');
alpha(0.1);
title('Axiovision versus MPC, with reproducibility (r) in shaded region',"FontWeight","bold");
xlabel('Axiovision', "FontWeight","bold");
ylabel('MPC', "FontWeight","bold");
hold off;
2 Comments
Image Analyst
on 5 Nov 2022
Edited: Image Analyst
on 5 Nov 2022
If you have any more questions, then attach your data and code to read it in with the paperclip icon after you read this:
Teshan Rezel
on 5 Nov 2022
Accepted Answer
It would help to have the data.
Try something like this —
x = linspace(0, 2*pi, 50);
y = sin(x) + 0.5;
err = rand(size(x))*0.5 + 0.5;
figure
scatter(x, y, 'filled')
hold on
patch([x, flip(x)], [y-0.25 flip(y+err)], 'b', 'FaceAlpha',0.25, 'EdgeColor','r', 'LineStyle','-.', 'MarkerSize',20)
hold off
grid
Experiment to get the desired result.
.
4 Comments
Teshan Rezel
on 5 Nov 2022
The vectors used for the patch call need to be sorted, since they are not originally. This sorts ‘myData.Anisotropic_1’ and then uses that index to reference the other vectors. Also, my code used row vectors, your uses column vectors, so that required a minor tweak. With those changes, it works.
Try this —
LD = load(websave('myData','https://www.mathworks.com/matlabcentral/answers/uploaded_files/1182658/myData.mat'));
myData = LD.myData
[patchx,pidx] = sort(myData.Anisotropic_1); % Sort, Return Index
patchy1 = myData.AnisoMax(pidx); % Use 'pidx'
patchy2 = myData.AnisoMin(pidx); % Use 'pidx'
xfill = [patchx; flipud(patchx)];
obs_anisofill = [patchy1; flipud(patchy2)];
figure;
hold on;
% % fill(xfill, obs_anisofill);
% % plot(myData.Anisotropic_1, myData.AnisoMin);
% % plot(myData.Anisotropic_1, myData.AnisoMax);
scatter(myData.Anisotropic_1, myData.("MPC aniso"), 'b');
h = lsline;
h.Color = 'b';
[p1, s1] = polyfit(get(h,'xdata'),get(h,'ydata'),1);
f1 = polyval(p1, myData.Anisotropic_1);
% rs1 = my_Rsquared_coeff(myData.("MPC aniso"), f1); % 'my_Rsquared_coeff' Doesn't Exist
% caption1 = sprintf('y = %0.2fx + %0.2f, R^2 = %.2f', p1(1), p1(2), rs1);
% text(min(myData.Anisotropic_1)+20, max(myData.("MPC aniso"))-40, caption1, 'FontSize', 10, 'Color', 'b', 'FontWeight', 'bold');
% plot(polyval(FitAniso, myData.Anisotropic_1));
patch(xfill, obs_anisofill, 'b', 'FaceAlpha',0.1, 'EdgeColor','r', 'LineStyle','-.', 'MarkerSize',20)
title('Axiovision versus MPC, with reproducibility (r) in shaded region',"FontWeight","bold");
xlabel('Axiovision', "FontWeight","bold");
ylabel('MPC', "FontWeight","bold");
hold off;
Some information (‘rs1’) is missing, so I commented it out and the lines using it out so the rest of the code will run.
Make appropriate changes to get the desired result.
.
Walter Roberson
on 6 Nov 2022
The alternative to sorting is to use boundary() with an alpha value to find the outline of the containing shape.
Star Strider
on 6 Nov 2022
I’ve done that in the past, especailly when the outline to be filled would not yield to sorting and remained essentially impossible to define. Here, sorting worked.
More Answers (0)
Categories
Find more on 2-D and 3-D Plots in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
