Display data points in boxplots

I am using the boxplot function to create boxplots of my data. But I would also like to plot my data on top of the boxplots.

Can anyone please help with that?

Below is an example of my code

Data1 = [0.2658 0.1969 0.3702 0.2259 0.2575 0.2253 0.4486  0.5385  0.3982  0.2899]';
Data2 = [0.2589 0.2094 0.3575 0.2391 0.2547 0.1987 0.4156 0.5443 0.394 0.3]';
Data3 =  [0.2589 0.2094 0.3575 0.2391 0.2547 0.2087 0.4256 0.5243 0.3822 0.2983]';
group = [    ones(size(Data1));
         2 * ones(size(Data2))
         3 * ones(size(Data3))];
boxplot([Data1; Data2; Data3],group)
h = boxplot([Data1; Data2; Data3],group)
set(h,{'linew'},{2})
set(gca,'XTickLabel', {'Data1'; 'Data2'; 'Data3'})

Answer 1

[Adam Danz]

After some slight rearrangement of your code, I added a secton that optionally scatters the x coordinates around the centers of the boxes. Set the 'spread' to 0 to plot data points in the center of the boxes.

allData = {Data1; Data2; Data3}; 
h = boxplot(cell2mat(allData),group); % old version: h = boxplot([allData{:}],group);
set(h, 'linewidth' ,2)
set(gca,'XTickLabel', {'Data1'; 'Data2'; 'Data3'})

hold on
xCenter = 1:numel(allData); 
spread = 0.5; % 0=no spread; 0.5=random spread within box bounds (can be any value)

for i = 1:numel(allData)
    plot(rand(size(allData{i}))*spread -(spread/2) + xCenter(i), allData{i}, 'mo','linewidth', 2)
end

The x-scatter is random so the the x coordinates will differ each time the plot is created unless the rng seed is controlled.

[Gina Carts]

That's what I actually needed. Thanks a lot

[Adam Danz]

Glad I could help. Not sure if you caught my update a minute ago but I made a slight change so that the scatter can be controlled.

[Jonathan Glenning]

Thanks for your solution, But is there a way to make this work without using data sets that have identical amounts of data points within them? Whenever I try to use this I get:

"Error using horzcat

Dimensions of arrays being concatenated are not consistent."

I recognise that my data sets arent all equal in both column and row so this is likely tripping it up, but is there a solution to this?

[Adam Danz]

Sure; if the vectors stored in the cell array are vertical,

% Use this
h = boxplot(cell2mat(allData),group);

% instead of this
% h = boxplot([allData{:}],group);

[DM]

Is there a way to make each data appear in a different colour instead all of them being pink?

[Adam Danz]

The plot() command only allows you to assign one color per object. So you could plot the values within a loop where you can assign a different color on each iteration. Another method would be to use the scatter() function that allows you to assign different color for each point.