Introduction to matplotlib

Overview

matplotlib is the workhorse of data science visualization. The module pyplot gives us MATLAB like plots.

You can install it via

pip install matplotlib

The most basic plot is done with the “plot”-function. It looks like this:

import matplotlib.pyplot as plt 
plt.plot([0, 1, 2, 3], [0, 1, 2, 3]) 
plt.show()

The plot function takes an x and y array and draws a blue line through all points.

You can of course draw each point independently without a line:

plt.plot([0, 1, 2, 3], [0, 1, 2, 3], "o")

or you can highlight the individual points while drawing the line

plt.plot([0, 1, 2, 3], [0, 1, 2, 3], marker='o')

You can find other markers here.

Color of the plot

The color of the line can be changed with the color parameter. The default color of a plot is blue. If you want e.g. red as the color, you can use ‘r’.

plt.plot([0, 1, 2, 3], [0, 1, 2, 3], color="r")

The following table shows the basic colors

In /matplotlib/_color_data.py you find additional colors, even colors from the XKCD color survey results

plt.plot(x, y, color="xkcd:nasty green")

Stroke width and style

changing the width of the plotted line is done via linewidth

plt.plot([0, 1, 2, 3], [0, 1, 2, 3], linewidth=7.0, color="xkcd:nasty green")

and the stroke style can be altered with the linestyle parameter

plt.plot([0, 1, 2, 3], [0, 1, 2, 3], linestyle=":", color="xkcd:nasty green", linewidth=7.0)

You can find other line styles here

Axis Labels

In school I learned that all axis of a plot must have labels. So let’s add them:

plt.ylabel('some other numbers') 
plt.xlabel('some numbers')

Saving the plot

If You want to save the plot as a png you can replace the show command with

plt.savefig('line_plot_01.png')

You can find the code examples in one JuPyter notebook in my github repo

When you’ve become familiar with the basic plot function you can dive into part 2