Introduction to Matplotlib

Why plot?

A table of numbers hides trends that a plot reveals at a glance: a linear region, a sudden change, an outlier. Matplotlib is Python's standard plotting library. It can produce anything from a quick check of your data to figures ready for publication. Like NumPy, it has its own SPASCLE course; this lesson covers the essentials.

Most plotting uses the pyplot module, imported as plt by convention. When you run plotting code in SPASCLE's editor, the figure appears in the output panel. The first run takes a little longer while Matplotlib loads.

Your first line plot

plt.plot(x, y) draws a line through the points given by two sequences of the same length: x-coordinates and y-coordinates. Lists and NumPy arrays both work. plt.show() displays the finished figure.

import matplotlib.pyplot as plttemperature = [300, 400, 500, 600, 700, 800]       # Kresistivity = [1.72, 2.42, 3.11, 3.81, 4.50, 5.20]  # µΩ·cm, copperplt.plot(temperature, resistivity)plt.show()

Try it hands-on

The plot shows the trend, but nobody else could tell what it represents. A scientific plot always needs labelled axes with units.

Labels, titles and grids

plt.plot() also accepts a format for the line and markers. marker="o" adds a circle at each data point, which shows readers where the measurements actually are. linestyle="--" draws a dashed line, and color= sets the colour.

import matplotlib.pyplot as plttemperature = [300, 400, 500, 600, 700, 800]resistivity = [1.72, 2.42, 3.11, 3.81, 4.50, 5.20]plt.figure(figsize=(6, 4))plt.plot(temperature, resistivity, marker="o", color="#580A50")plt.xlabel("Temperature (K)")plt.ylabel("Resistivity (µΩ·cm)")plt.title("Electrical resistivity of copper")plt.grid(True)plt.show()

Try it hands-on

Several series and a legend

Call plt.plot() several times before plt.show() to draw several lines on the same axes. Give each a label=, then call plt.legend() to show which line is which. NumPy is handy here for generating smooth x-values with np.linspace() and calculating y-values from a formula.

import numpy as npimport matplotlib.pyplot as plttemperature = np.linspace(0, 400, 50)     # temperature rise, Klength = 1.0                              # mfor name, alpha in [("Aluminium", 23.1e-6), ("Copper", 16.5e-6), ("Steel", 12.0e-6)]:    expansion_mm = alpha * length * temperature * 1000    plt.plot(temperature, expansion_mm, label=name)plt.xlabel("Temperature rise (K)")plt.ylabel("Expansion of a 1 m bar (mm)")plt.title("Thermal expansion")plt.legend()plt.grid(True)plt.show()

Try it hands-on

Scatter plots

Use plt.scatter(x, y) when each point is an independent measurement and joining them with a line would suggest a continuity that isn't there, for example when comparing properties of different materials. plt.annotate() adds a text label next to a point.

import matplotlib.pyplot as pltnames = ["Mg", "Al", "Ti", "Fe", "Cu", "W"]density = [1.74, 2.70, 4.51, 7.87, 8.96, 19.25]      # g/cm³modulus = [45, 70, 116, 211, 130, 411]               # GPaplt.scatter(density, modulus, color="#580A50")for name, x, y in zip(names, density, modulus):    plt.annotate(name, (x, y), textcoords="offset points", xytext=(5, 5))plt.xlabel("Density (g/cm³)")plt.ylabel("Young's modulus (GPa)")plt.title("Stiffness versus density of metals")plt.grid(True)plt.show()

Try it hands-on

Bar charts

plt.bar(categories, values) compares a quantity across named categories.

import matplotlib.pyplot as pltmetals = ["Al", "Cu", "Fe", "Ti", "W"]melting_point = [660, 1085, 1538, 1668, 3422]   # °Cplt.bar(metals, melting_point, color="#580A50")plt.ylabel("Melting point (°C)")plt.title("Melting points of common metals")plt.show()

Try it hands-on