{"id":2351,"date":"2025-05-14T08:38:26","date_gmt":"2025-05-14T12:38:26","guid":{"rendered":"https:\/\/blogs.mathworks.com\/graphics-and-apps\/?p=2351"},"modified":"2026-01-30T20:33:07","modified_gmt":"2026-01-31T01:33:07","slug":"making-even-prettier-graphs","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/graphics-and-apps\/2025\/05\/14\/making-even-prettier-graphs\/","title":{"rendered":"Making even prettier graphs"},"content":{"rendered":"\r\n\r\n\r\n
\"Abby<\/td>\r\nGuest Writer: Abby Skofield<\/a>\r\n<\/strong>We're excited to welcome Abby Skofield back to the blog! Abby revisits the classic topic of \"Making Pretty Graphs\"\u2014this time bringing a modern twist with some of the latest MATLAB features. Abby shows how the workflow for customizing and polishing your figures has evolved\u2014making it easier than ever to create beautiful, publication-ready graphics right from your code. For Abby's complete bio, see our Contributors Page<\/a>.<\/td>\r\n<\/tr>\r\n<\/tbody>\r\n<\/table>\r\n
<\/div>\r\n
\r\n
\r\n\r\n
\r\n\r\n<\/div>\r\n
\r\n\r\n\"Open<\/a>\r\n\r\n<\/div>\r\n
Almost 20 years ago, my colleague Jiro Doke<\/a> published a blog article titled \"Making Pretty Graphs<\/a>,\" which I re-read and recommend to others more often than I do Jane Austen's collected works. Jiro, this means multiple times a year, believe it or not! In fact, I just shared his blog post with a new colleague today. As much as I love Jiro's original post, we've made some big changes to graphics in the intervening years. I realized that maybe it was time for me to pick up the torch and tackle Jiro's project with the latest version of MATLAB at my fingertips. Today I am going to share with you some examples of how you can use MATLAB code and the properties of our graphics objects to customize the way your MATLAB plots look. You'll probably make different aesthetic choices than I do, but the same techniques can be used to make your MATLAB plots even prettier<\/span>.<\/div>\r\n
Before joining MathWorks, I worked as a research assistant in a cognitive neuroscience lab, collecting and analyzing experimental data. Like Jiro, I used MATLAB to create data visualizations and, like many customers I've met during my subsequent 12 years at MathWorks, I used other tools to tweak these plots before including them in journal submissions and grant applications. You may yourself be a part of this fellowship, and welcome! If you're a wiz in Illustrator, or just a control freak, this workflow probably suits you pretty well. But perhaps you can save yourself some time the next time around by doing more of these tweaks in MATLAB.<\/div>\r\n
The big benefit of moving this work into MATLAB is that it is scriptable. Any customizations can be captured in code and you can re-run and regenerate the visualization as often as you want while iterating on the final look. And when you're done, you can export it to an image format like PNG, JPG or PDF. Let me show you what I'm talking about!<\/div>\r\n

Create Basic Plot<\/h2>\r\n
We'll start with some made up data, representing two groups of observed values as well as the results of a probability density estimation for each group of values. We can use histogram<\/span> to visualize the distribution of values and plot<\/span> to add in the density curves. What we get is a standard looking axes containing four charts, each of which is given its own color.<\/div>\r\n
\r\n
\r\n
[data1, data2, xpdf, pdf1, pdf2] = getFirstDataset;<\/div>\r\n<\/div>\r\n
\r\n
<\/div>\r\n<\/div>\r\n
\r\n
figure<\/div>\r\n<\/div>\r\n
\r\n
binEdges = 0:2:20;<\/div>\r\n<\/div>\r\n
\r\n
h1 = histogram(data1,binEdges,Normalization=\"pdf\"<\/span>,DisplayName=\"Observed\"<\/span>); % DisplayName stores the legend entry for this plot with the object<\/span><\/div>\r\n<\/div>\r\n
\r\n
hold on<\/div>\r\n<\/div>\r\n
\r\n
h2 = histogram(data2,binEdges,Normalization=\"pdf\"<\/span>,DisplayName=\"Predicted\"<\/span>);<\/div>\r\n<\/div>\r\n
\r\n
p1 = plot(xpdf,pdf1);<\/div>\r\n<\/div>\r\n
\r\n
p2 = plot(xpdf,pdf2);<\/div>\r\n
\r\n
\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n

Adjust Chart Properties<\/span><\/h2>\r\n
Many customers who share their published figures with us seem to prefer a lighter weight style, with fewer \"edges\" on their charts. The first thing I will do is clean up the edge lines around the histogram objects by setting the EdgeColor<\/span> to white. (You can also set most color properties, like EdgeColor<\/span>, to \"none\"<\/span>, though in the case of histogram<\/span> this makes it hard to see where the bin edges are.) I will also bump up the LineWidth<\/span> of the plot lines to help them stand out. Note that my example demonstrates how you can set a property on a single handle, as with like h1<\/span> and h2<\/span> below, using the dot-operator to access a property like EdgeColor<\/span>, or you can set a property on an array of multiple objects, as in [p1 p2]<\/span> below, using the set<\/span> command.<\/div>\r\n
\r\n
\r\n
h1.EdgeColor = \"white\"<\/span>;<\/div>\r\n<\/div>\r\n
\r\n
h2.EdgeColor = \"white\"<\/span>;<\/div>\r\n<\/div>\r\n
\r\n
<\/div>\r\n<\/div>\r\n
\r\n
set([p1 p2],LineWidth=4);<\/div>\r\n
\r\n
\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n

SeriesIndex and ColorOrder<\/h2>\r\n
The other thing I've often seen in customers' published examples is a simpler color palette. In my example, I know that each of the density curves should be associated with a different one of the Histogram objects, so I don't need<\/span> four different colors for the plots. I will use the SeriesIndex<\/span> property on the plot objects to tell MATLAB to group them, visually, with the respective Histogram. SeriesIndex<\/span> is how the system indexes into the Axes's ColorOrder<\/span> palette - charts with the same SeriesIndex<\/span> value will share the same color.<\/div>\r\n
\r\n
\r\n
p1.SeriesIndex = h1.SeriesIndex;<\/div>\r\n<\/div>\r\n
\r\n
p2.SeriesIndex = h2.SeriesIndex;<\/div>\r\n
\r\n
\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n
I can then use the colororder<\/span> function to update the color palette to one that works better for my dataset. Note that hex colors can be easily converted using the hex2rgb<\/span> function.<\/div>\r\n
\r\n
\r\n
redColor = hex2rgb(\"#C53B29\"<\/span>);<\/div>\r\n<\/div>\r\n
\r\n
greyColor = [0.4 0.4 0.4];<\/div>\r\n<\/div>\r\n
\r\n
colororder([greyColor;redColor])<\/div>\r\n
\r\n
\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n

Add Labels and Legend<\/h2>\r\n
Now that things are starting to look a little more dressed up, it's time to add some labels. Most of the labels you can add to MATLAB plots have their own function, like xlabel<\/span> or subtitle<\/span>.<\/div>\r\n
\r\n
\r\n
title(\"Figure 1\"<\/span>)<\/div>\r\n<\/div>\r\n
\r\n
subtitle(\"Probability distribution by distance\"<\/span>)<\/div>\r\n<\/div>\r\n
\r\n
xlabel(\"Distance\"<\/span>)<\/div>\r\n<\/div>\r\n
\r\n
ylabel(\"Probability\"<\/span>)<\/div>\r\n<\/div>\r\n
\r\n
xsecondarylabel(\"\\mum\"<\/span>) % Greek and other symbols like mu can be added using TeX markup with a backslash, as in: \\mu<\/span><\/div>\r\n<\/div>\r\n<\/div>\r\n
By passing in two specific handles to the legend command, I can create a legend with only those two entries. This helps keep things simple since I am only using two distinct colors to differentiate the Observed and Predicted values. Because I specified a DisplayName<\/span> in each of the histogram<\/span> commands above, these names will automatically be used in the legend.<\/div>\r\n
<\/div>\r\n
\r\n
\r\n
leg = legend([h1,h2]);<\/div>\r\n
\r\n
\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n
Most graphics commands, like legend<\/span> above, optionally return a handle to the object that is created by the command, which enables you to set properties on that object after it is created. In fact, I'll make that legend a little neater and tidier by turning off the Box border and reducing the width used for the legend icons.<\/div>\r\n
\r\n
\r\n
leg.Box = \"off\"<\/span>;<\/div>\r\n<\/div>\r\n
\r\n
leg.IconColumnWidth = 9;<\/div>\r\n
\r\n
\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n
<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n

Adjust Axes and Font Properties<\/h2>\r\n
There is one more object we haven't yet tweaked - the Axes. In MATLAB graphics, the Axes object encompasses the x<\/span>-axis, y<\/span>-axis and sometimes z<\/span>-axis which define the space in which the chart objects appear. The Axes object contains multitudes of properties - over 100! So if there's a tweak you're looking for, take a look in our excellent documentation. All graphics objects have documentation pages listing and explaining all of the object's properties, for instance:<\/div>\r\n