Visualize scientific data

What you will learn

• Label axes with meaningful units.
• Plot multiple data series and add a legend.
• Distinguish measured data from model predictions.

The concept

A scientific figure should answer a question. Choose a plot type suited to the data, label axes with units, show relevant uncertainty when known, and avoid decorative effects that obscure the measurement.

MATLAB supports line plots, scatter plots, histograms, images, contour plots and 3D surfaces. Use hold on to add series to an axes and legend to explain them.

Worked example

Plot a noisy measurement alongside the underlying model.

t = linspace(0, 1, 250);
truth = exp(-2*t).*sin(2*pi*8*t);
measurement = truth + 0.08*randn(size(t));
plot(t, measurement, '.', 'DisplayName', 'Measured');
hold on;
plot(t, truth, 'LineWidth', 2, 'DisplayName', 'Model');
xlabel('Time (s)'); ylabel('Amplitude (a.u.)');
legend('Location', 'best'); grid on;

The random noise makes each run slightly different. For repeatable demonstrations, set a random seed with rng(1).

Try it yourself

Consider what the viewer needs to interpret the plot.

Where this is used

Visualization is central to spectroscopy, time-series analysis, optical imaging, sensor validation and publication-quality figures.