Adding caption below X-axis for a scatter plot using matplotlib

First, I feel weird posting an answer against the co-lead developer of matplotlib. Obviously, @tacaswell knows matplotlib far better than I ever will. But at the same time, his answer wasn't dynamic enough for me. I needed a caption that would always be based on the position of the xlabel, and couldn't just use text annotations.

I considered simply changing the xlabel to add a newline and the caption text, but that wouldn't clearly differentiate the caption, and you can't do things like change the text size or make it italic in the middle of a text string.

I solved this by using matplotlib's TeX capabilities. Here's my solution:

from matplotlib import pyplot as plt
from matplotlib import rc
import numpy as np
from pylab import *

rc('text', usetex=True)

file = open('distribution.txt', 'r')

txt="I need the caption to be present a little below X-axis"

x=[]
y=[]
for line in file:
    new=line.rstrip()
    mystring=new.split("\t")
    x.append(mystring[0])
    y.append(mystring[1])


fig = plt.figure()
ax1 = fig.add_axes((0.1,0.4,0.8,0.5))
ax1.set_title("This is my title")
ax1.set_xlabel(r'\begin{center}X-axis\\*\textit{\small{' + txt + r'}}\end{center}')
ax1.set_ylabel('Y-axis')
ax1.scatter(x,y, c='r')
plt.xlim(0, 1.05)
plt.ylim(0, 2.5)
plt.show()

I did the same thing with the random scatter plot from tacaswell's answer, and here's my result:

Scatter plot with caption

One warning: if you tweak this to take input string variables, the strings may not be properly escaped for use with TeX. Escaping LaTeX code is already covered on Stack Overflow, at https://stackoverflow.com/a/25875504/1404311 . I used that directly, and then could take arbitrary xlabels and captions.


You can simply use figtext. You can also change the value of x and y-axes as you want.

txt="I need the caption to be present a little below X-axis"
plt.figtext(0.5, 0.01, txt, wrap=True, horizontalalignment='center', fontsize=12)

Something like:

from matplotlib import pyplot as plt
import numpy as np

txt="I need the caption to be present a little below X-axis"

# make some synthetic data
x = np.linspace(0, 1, 512)
y = np.random.rand(512)*2.3 + .1

fig = plt.figure()
ax1 = fig.add_axes((0.1, 0.2, 0.8, 0.7))

ax1.set_title("This is my title")
ax1.set_xlabel('X-axis')
ax1.set_ylabel('Y-axis')

# make the edge colors match the facecolors
ax1.scatter(x,y, c='r', edgecolors='face')
# center text
fig.text(.5, .05, txt, ha='center')

# use OO interface    
ax1.set_xlim([0, 1.05])
ax1.set_ylim([0, 2.5])

# resize the figure to match the aspect ratio of the Axes    
fig.set_size_inches(7, 8, forward=True)

plt.show()

example result

might work. Making this easier to do is on the radar for mpl upstream, but we are still looking for someone to do it.