make a grouped bar graph with individual data points and std error bars inside the bars
71 views (last 30 days)
Show older comments
Dear Matlab,
I have delta and theta power spectrum data for 2 groups of participants. Control group N=11. Patient group N=6 (total 17 subjs).
GOAL: to make a bar graph with:
- individual data points and
- std error bars within the bars.
- I would also like to group them into 2 x-ticks (one for delta, one for theta [each with 2 bars for (1) controls and (2) patients])
- and also delta and theta to each be a different color
IDEAL GRAPH: something very similar to:
What I have to far, please see attached bar graph in standard blue.
I have attached sample data as well. columns 1,3 = controls. columns 2,4 = patient group.
Below is code I have so far:
% load SampleData
figure(8),clf, hold on
b1 = bar([1 2 3 4],mean(logbar,'Omitnan'));%,'FaceColor',[.7 .3 .9]);
errorbar([1,2,3,4],mean(logbar,'Omitnan'),std(logbar,[],'omitnan')/sqrt(17-1),'k.','LineWidth',1);
set(gca,'XTickLabel',{'Delta:Ctrl'; 'Delta:Dra';'Theta:Ctrl'; 'Theta:Dra'},'xtick',[1 2 3 4],'fontweight','bold')
ylabel('Relative Power (dB)','fontweight','bold','fontsize',11)
Thank you so much for any help and time, as always!
Best,
Joanne
1 Comment
Accepted Answer
Shree Charan
on 5 May 2023
Hi Joanne,
@Dave B‘s answer in https://www.mathworks.com/matlabcentral/answers/888187-grouped-bar-graph-with-individual-datapoints may be modified to suit the current requirements as follows.
figure(8),clf,
%split the data
ctrl = logbar(:,[1 3]);
dra = logbar(:,[2 4]);
%plot the bar graph
h = bar([mean(ctrl, 'Omitnan'); mean(dra, 'Omitnan')]');
hold on
h(1).FaceColor='r';
h(2).FaceColor='y';
%plot the error bars
errorbar(h(1).XEndPoints,mean(ctrl, 'Omitnan'),std(ctrl, 'Omitnan')/sqrt(17-1),'k.','LineWidth',2)
errorbar(h(2).XEndPoints,mean(dra, 'Omitnan'),std(dra, 'Omitnan')/sqrt(17-1),'k.','LineWidth',2)
%plot the individual points
scatter(repmat(h(1).XEndPoints(1), size(ctrl, 1), 1),ctrl(:, 1),60,'MarkerFaceColor','r','MarkerEdgeColor','k','LineWidth',1)
scatter(repmat(h(1).XEndPoints(2), size(ctrl, 1), 1),ctrl(:, 2),60,'MarkerFaceColor','r','MarkerEdgeColor','k','LineWidth',1)
scatter(repmat(h(2).XEndPoints(1), size(dra, 1), 1),dra(:, 1),60,'MarkerFaceColor','y','MarkerEdgeColor','k','LineWidth',1)
scatter(repmat(h(2).XEndPoints(2), size(dra, 1), 1),dra(:, 2),60,'MarkerFaceColor','y','MarkerEdgeColor','k','LineWidth',1)
%Give appropriate labels
set(gca,'XTickLabel',["Delta" "Theta"],'fontweight','bold')
ylabel('Relative Power (dB)','fontweight','bold','fontsize',11)
legend(["Ctrl", "Dra"])
hold off
The resulting graph would be as follows
You can further read up on “errorbar” and “repmat” function in the following MATLAB documentation:
2 Comments
Dave B
on 5 May 2023
I think I noticed one of your goals was to have two levels of tick labels, that can be a little difficult but a trick with tiledlayout makes it doable. Here's a slight modification of @Shree Charan's answer that adds the second row of labels.
load SampleData
clf
tiledlayout(1,2,'TileSpacing','none')
a1 = nexttile;hold on
b1 = bar([1 2],mean(logbar(:,1:2),'Omitnan'));
scatter(repmat(b1.XEndPoints,size(logbar,1),1), logbar(:,1:2),'filled','SeriesIndex',1,'MarkerEdgeColor','k','LineWidth',1)
errorbar([1 2],mean(logbar(:,1:2),'Omitnan'),std(logbar(:,1:2),[],'omitnan')/sqrt(17-1),'k.','LineWidth',1,'Marker', 'none');
set(a1,'XTickLabel',{'Ctrl'; 'Dra'},'xtick',[1 2],'fontweight','bold')
ylabel('Relative Power (dB)','fontweight','bold','fontsize',11)
xlabel('Delta')
a2=nexttile;hold on
b2 = bar([1 2],mean(logbar(:,3:4),'Omitnan'),'SeriesIndex',2);
scatter(repmat(b2.XEndPoints,size(logbar,1),1), logbar(:,3:4),'filled','SeriesIndex',2,'MarkerEdgeColor','k','LineWidth',1)
errorbar([1 2],mean(logbar(:,3:4),'Omitnan'),std(logbar(:,3:4),[],'omitnan')/sqrt(17-1),'k.','LineWidth',1,'Marker', 'none');
set(a2,'XTickLabel',{'Ctrl'; 'Dra'},'xtick',[1 2],'fontweight','bold','YColor','none')
xlabel('Theta')
linkaxes([a1 a2],'y')
More Answers (0)
See Also
Categories
Find more on Bar 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!