How to change the location of the symbols/text within a legend box?

Question

I have a subplot with a single legend entry. I am placing the legend at the bottom of the figure and using mode='expand'; however, the single legend entry is placed to the very left of the legend box. To my understanding, changing kwargs such as bbox_to_anchor changes the legend box parameters but not the parameters of the symbols/text within. Below is an example to reproduce my issue.

import numpy as np
import matplotlib.pyplot as plt

x = np.linspace(-10, 10, 21)
y = np.exp(x)
z = x **2

fig, axes = plt.subplots(nrows=1, ncols=2)
axes[0].plot(x, y, color='r', label='exponential')
axes[1].plot(x, z, color='b')
# handles, labels = axes[0].get_legend_handles_labels()
plt.subplots_adjust(bottom=0.125)
fig.legend(mode='expand', loc='lower center')
plt.show()
plt.close(fig)

This code produces . How can I change the position of the symbol and text such that they are centered in the legend box?

PS: I am aware that exponential is a bad label for this subplot since it only describes the first subfigure. But, this is just for examples-sake so that I can apply it to my actual use-case.

Answer 1

The legend entries are placed using a HPacker object. This does not allow to be centered. The behaviour is rather that those HPackers are "justified" (similar to the "justify" option in common word processing software).

A workaround would be to create three (or any odd number of) legend entries, such that the desired entry is in the middle. This would be accomplished via the ncol argument and the use of "dummy" entries (which might be transparent and have no associated label).

import numpy as np
import matplotlib.pyplot as plt

x = np.linspace(-10, 10, 21)
y = np.exp(x)
z = x **2

fig, axes = plt.subplots(nrows=1, ncols=2)
fig.subplots_adjust(bottom=0.125)

l1, = axes[0].plot(x, y, color='r', label='exponential')
axes[1].plot(x, z, color='b')


dummy = plt.Line2D([],[], alpha=0)
fig.legend(handles=[dummy, l1, dummy],
           mode='expand', loc='lower center', ncol=3)
plt.show()