Accented characters in Matplotlib

Question

Is there a way to get Matplotlib to render accented chars (é,ã,â,etc)?

For instance, I'm trying to use accented characters on set_yticklabels() and Matplotlib renders squares instead, and when I use unicode() it renders the wrong characters.

Is there a way to make this work?

It turns out you can use u"éã", but first you have to set the file encoding:

# Using the magic encoding
# -*- coding: utf-8 -*-

After that Matplotlib correctly renders

u"é"

I also learned that you can use

import matplotlib.font_manager as fm
fp1=fm.FontProperties(fname="/path/to/somefont.ttf")
ax.title("é",fontproperties=fp1)

in case you need to render a characters that Matplotlib does not have.

Answer 1

Prefix the strings with u to tell Python that they are Unicode strings:

ax.set_yticklabels([u'é', u'ã', u'â'])

Answer 2

Sure. You can use TeX:

from matplotlib import rcParams
rcParams['text.usetex'] = True
ax = ... # Axes object
ax.set_yticklabels(['$\'{e}$', '$\tilde{a}$', '$\hat{a}$'])

Answer 3

I also had this problem specifically when I was trying to use the annotate function. Here was my error message:

ValueError: matplotlib display text must have all code points < 128 or use Unicode strings

And here's what I used to resolve this:

"accented string i.e. sāo paulo".decode('utf-8')

Answer 4

from matplotlib import rc

rcParams['text.latex.unicode']=True