I want to move up my boxchart

I want to move up my boxchart a little bit. I want it to have two boxes for 0min right next to eachother, also for the other 3 exposure times right next to eachother. So, you will still have 4 catogories, but for every catagorie two boxes next to eachother. Because, it is a little bit hard to see now. Thank you!

%% Glass Ra

clear all, close all;

glass = xlsread('excelsensofarglass')
glass2 = xlsread('excelsensofarglass2')
glass3 = xlsread('excelsensofarglass3')
glass4 = xlsread('excelsensofarglass4')
glass5 = xlsread('excelsensofarglass5')
glass6 = xlsread('excelsensofarglass6')


G1 = [transpose(glass(1:4,1));transpose(glass2(1:4,1));transpose(glass3(1:4,1));transpose(glass4(1:4,1));transpose(glass5(1:4,1));transpose(glass6(1:4,1))]
G2 = [transpose(glass(6:9,1));transpose(glass2(6:9,1));transpose(glass3(6:9,1));transpose(glass4(6:9,1));transpose(glass5(6:9,1));transpose(glass6(6:9,1))] 
G4 = [transpose(glass(11:14,1));transpose(glass2(11:14,1));transpose(glass3(11:14,1));transpose(glass4(11:14,1));transpose(glass5(11:14,1));transpose(glass6(11:14,1))] 
G6 = [transpose(glass(16:19,1));transpose(glass2(16:19,1));transpose(glass3(16:19,1));transpose(glass4(16:19,1));transpose(glass5(16:19,1));transpose(glass6(16:19,1))]


figure(1)
hold on
% boxchart(G1)
boxchart(G2)
boxchart(G4)
% boxchart(G6)
ylim([0 1.6])
ylabel('Arithmetic average height [\mum]')
xlabel('duration of exposure')
title('Arithmetic average height glass')
legend('20 [\mum]','40 [\mum]','location','northwest')
set(gca,'XTickLabel',{'0min','60min','90min','240min'})

Answer 1

[Voss]

You can create a single boxchart per boxchart call, with a single column of data, and set its X location as desired.

To illustrate: first, the original, showing the overlapping boxcharts (using random data):

% Some random data:
G2 = rand(6,4);
G4 = rand(6,4);

figure%(1)
hold on

% boxchart(G1)
boxchart(G2)
boxchart(G4)
% boxchart(G6)

ylim([0 1.6])
ylabel('Arithmetic average height [\mum]')
xlabel('duration of exposure')
title('Arithmetic average height glass')
legend('20 [\mum]','40 [\mum]','location','northwest')
set(gca,'XTickLabel',{'0min','60min','90min','240min'})

Now, with multiple boxchart calls:

figure%(1)
hold on

[n_rows,n_cols] = size(G2);
x = zeros(n_rows,1);
colors = get(gca(),'ColorOrder');
for ii = 1:n_cols
    boxchart(x+ii-0.15,G2(:,ii),'BoxWidth',0.3,'BoxFaceColor',colors(1,:))
    boxchart(x+ii+0.15,G4(:,ii),'BoxWidth',0.3,'BoxFaceColor',colors(2,:))
end

ylim([0 1.6])
ylabel('Arithmetic average height [\mum]')
xlabel('duration of exposure')
title('Arithmetic average height glass')
legend('20 [\mum]','40 [\mum]','location','northwest')
set(gca, ...
    'XTick',1:4, ...
    'XTickLabel',{'0min','60min','90min','240min'})

[Chiel van Wanrooij]

This works perfectly. Thank you for your help!

[Voss]

You're welcome!

[Chiel van Wanrooij]

I have a follow up question. I maybe want to do four columns, I can make the width of all the boxes a bit smaller. However, than it is not visible anymore. Is it possible to widen the space bewteen the 4 catogories?

[Voss]

See below for a version with 4 groups:

% Some random data:
G1 = rand(6,4);
G2 = rand(6,4);
G4 = rand(6,4);
G6 = rand(6,4);

figure%(1)
hold on
[n_rows,n_cols] = size(G2);
x = zeros(n_rows,1);
colors = get(gca(),'ColorOrder');
x_offset = [-0.3 -0.1 0.1 0.3];
box_width = 0.2;
for ii = 1:n_cols
    boxchart(x+ii+x_offset(1),G1(:,ii),'BoxWidth',box_width,'BoxFaceColor',colors(1,:))
    boxchart(x+ii+x_offset(2),G2(:,ii),'BoxWidth',box_width,'BoxFaceColor',colors(2,:))
    boxchart(x+ii+x_offset(3),G4(:,ii),'BoxWidth',box_width,'BoxFaceColor',colors(3,:))
    boxchart(x+ii+x_offset(4),G6(:,ii),'BoxWidth',box_width,'BoxFaceColor',colors(4,:))
end

ylim([0 1.6])
ylabel('Arithmetic average height [\mum]')
xlabel('duration of exposure')
title('Arithmetic average height glass')
legend('10 [\mum]','20 [\mum]','40 [\mum]','60 [\mum]','location','northwest')
set(gca, ...
    'XTick',1:4, ...
    'XTickLabel',{'0min','60min','90min','240min'})

[Voss]

Here's a more general version.

You can change data, group_number_for_legend, and/or column_number_for_xticks, and the rest of the code will work as long as column_number_for_xticks has the same number of elements as data has pages (i.e., the number of matrices G1, G2, G4, ... used in data), and column_number_for_xticks has the same number of elements as each matrix has columns.

As they are defined here below, you get 5 x-ticks, each with a set of 6 boxcharts.

% Some random data, 6 matrices of size 10-by-5:
G1 = rand(10,5);
G2 = rand(10,5);
G4 = rand(10,5);
G6 = rand(10,5);
G9 = rand(10,5);
G11 = rand(10,5);

% concatenate all the data in a 3D array:
data = cat(3,G1,G2,G4,G6,G9,G11);

% numbers to display in the legend:
group_number_for_legend = [10 20 40 60 90 110];

% numbers to display in the xticklabels:
column_number_for_xticks = [0 60 90 240 420];


figure%(1)
hold on

[n_rows,n_cols,n_groups] = size(data);

colors = get(gca(),'ColorOrder');
colors = repmat(colors,ceil(n_groups/size(colors,1)),1);

box_width = 1/(n_groups+1);
if n_groups == 1
    x_offset = 0;
else
    x_offset = linspace(-0.5+box_width,0.5-box_width,n_groups);
end

x = zeros(n_rows,1);
for ii = 1:n_cols
    for kk = 1:n_groups
        boxchart(x+ii+x_offset(kk),data(:,ii,kk),'BoxWidth',box_width,'BoxFaceColor',colors(kk,:))
    end
end

ylim([0 1.6])
ylabel('Arithmetic average height [\mum]')
xlabel('duration of exposure')
title('Arithmetic average height glass')
legend(sprintfc('%d [\\mum]',group_number_for_legend),'location','northwest')
set(gca, ...
    'XTick',1:n_cols, ...
    'XTickLabel',sprintfc('%dmin',column_number_for_xticks))

Another run with a different setup, to create 4 sets of 3 boxcharts:

% Some random data, 3 matrices of size 10-by-4:
G1 = rand(10,4);
G2 = rand(10,4);
G4 = rand(10,4);

% concatenate all the data in a 3D array:
data = cat(3,G1,G2,G4);

% numbers to display in the legend:
group_number_for_legend = [10 20 40];

% numbers to display in the xticklabels:
column_number_for_xticks = [0 60 90 240];


figure%(1)
hold on

[n_rows,n_cols,n_groups] = size(data);

colors = get(gca(),'ColorOrder');
colors = repmat(colors,ceil(n_groups/size(colors,1)),1);

box_width = 1/(n_groups+1);
if n_groups == 1
    x_offset = 0;
else
    x_offset = linspace(-0.5+box_width,0.5-box_width,n_groups);
end

x = zeros(n_rows,1);
for ii = 1:n_cols
    for kk = 1:n_groups
        boxchart(x+ii+x_offset(kk),data(:,ii,kk),'BoxWidth',box_width,'BoxFaceColor',colors(kk,:))
    end
end

ylim([0 1.6])
ylabel('Arithmetic average height [\mum]')
xlabel('duration of exposure')
title('Arithmetic average height glass')
legend(sprintfc('%d [\\mum]',group_number_for_legend),'location','northwest')
set(gca, ...
    'XTick',1:n_cols, ...
    'XTickLabel',sprintfc('%dmin',column_number_for_xticks))

Answer 2

[Cameron]

I'm not sure if there is a way to do that. You can always use this instead.

% Left axes
ax1 = nexttile;
boxchart(ax1,G2)
set(ax1,'XTickLabel',{'0min','60min','90min','240min'});
% Right axes
ax2 = nexttile;
boxchart(ax2,G4)
set(ax2,'XTickLabel',{'0min','60min','90min','240min'});

Answer 3

[the cyclist]

It doesn't seem to me that the code you pasted could create the plot example you've posted (but maybe I'm misunderstanding something). Can you upload your G1, G2, ... data? (You can use the paper clip icon in the INSERT section of the toolbar.)

Without understanding more, my best suggestion would be instead of using the

boxchart(ydata)

syntax, use

boxchart(xgroupdata,ydata)

and give the two groups you want to separate slightly different x values (e.g. [-1,1], [-59 61], etc), so that they are offset from each other.