How to create bar chart with x axis labels as grouped for different data

41 views (last 30 days)
Hello all, I want to plot a bar chart in matlab. I have plotted it in excel and attached the image for reference:
I have attached the excel file also. Any clue on how to get this.

Accepted Answer

Dave B
Dave B on 28 Aug 2023
Unfortunately MATLAB doesn't have a builtin feature for creating hierarchical rulers. An alternate workaround to the one dpb links to in the comment above is to use tiledlayout to make a separate axes for each bar chart. Here's an example that produces something somewhat similar to what you got in Excel:
tbl = readtable('Book1.xlsx','TextType','string');
Warning: Column headers from the file were modified to make them valid MATLAB identifiers before creating variable names for the table. The original column headers are saved in the VariableDescriptions property.
Set 'VariableNamingRule' to 'preserve' to use the original column headers as table variable names.
%% This chunk fills in the rows in the table to match the top row for the section
for i = 2:height(tbl)
for j = 1:width(tbl)
if ismissing(tbl{i,j})
tbl{i,j} = tbl{i-1,j};
end
end
end
tbl.SolutionAlgorithm = categorical(tbl.SolutionAlgorithm, ["SIMPLE" "SIMPLEC" "COUPLED"]);
groups = unique(tbl.TurbulenceModels,'stable');
%%
% This layout will hold each of the axes
layout = tiledlayout(1,numel(groups),'TileSpacing','tight');
% A background axes serves to create the illusion that tall of the bar
% charts live in the same axes, in reality there will be one Axes for
% each group
backax=axes(layout);
backax.Layout.TileSpan=[1 numel(groups)];
for i = 1:numel(groups)
% Note: this code uses axes and then sets the tile rather than using
% the nexttile function. The latter would potentially target 'backax'
% on the first iteration
ax = axes(layout);
% Make a bar chart for each group:
ax.Layout.Tile = i;
ind1 = tbl.TurbulenceModels==groups(i) & tbl.MeshingModels=="Tetrahedral";
ind2 = tbl.TurbulenceModels==groups(i) & tbl.MeshingModels=="Hexcore";
x = [tbl.SolutionAlgorithm(ind1) tbl.SolutionAlgorithm(ind2)];
y = [tbl.Cd(ind1) tbl.Cd(ind2)];
bar(x, y, 'BaseValue', .4, 'EdgeColor', 'none')
% This uses the Tex interpreter to add line breaks to the group names,
% unforturnately this has the side effect of left-justifying the groups
xlabel(groups(i).replace(" ","\newline"))
% Toggle the visibility of the axes as a whole (the white background
% and both rulers) off, then toggle the visibility of just the x axis
% back on.
set(ax,'XTickLabelRotation',90,'YTick',[],'Visible','off','YLim',[.4 1],'TickDir','none')
ax.XAxis.Visible='on';
end
% Because the individual axes objects have their own x axis, it can be
% disabled on the background axes
backax.XAxis.Visible=false;
% set the background axes y limits to match the last axes created, and add
% a label.
ylim(backax, ylim(ax));
ylabel(backax,'Cd')
% A legend must be associated with an axes, but it can be positioned with
% respect to the layout as follows:
leg = legend(ax, ["Tetrahedral meshing" "Hexcore meshing"],"Orientation",'horizontal');
leg.Layout.Tile='north';
% Make the aspect ratio a little wider for this long x axis
f=gcf;f.Position(3)=f.Position(3)*1.4;

More Answers (0)

Categories

Find more on Data Distribution 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!