Question

I am trying to have two inter-depedent x-axis in a matplotlib imshow() plot. I have bottom x-axis as the radius squared and I want the top as just the radius. I have tried so far:

ax8 = ax7.twiny()
ax8._sharex = ax7
fmtr = FuncFormatter(lambda x,pos: np.sqrt(x) )
ax8.xaxis.set_major_formatter(fmtr)
ax8.set_xlabel("Radius [m]")

where ax7 is the y-axis and the bottom x-axis (or radius squared). Instead of getting the sqrt (x_bottom) as the ticks at the top I just get a range from 0 to 1. How can I fix this?

Thanks a lot in advance.

Answer 1

You’re misunderstanding what twiny does. It makes a completely independent x-axis with a shared y-axis.

What you want to do is have a different formatter with a linked axis (i.e. sharing the axis limits but nothing else).

The simple way to do this is to manually set the axis limits for the twinned axis:

import numpy as np
import matplotlib.pyplot as plt
from matplotlib.ticker import FuncFormatter

fig, ax1 = plt.subplots()
ax1.plot(range(10))

ax2 = ax1.twiny()
formatter = FuncFormatter(lambda x, pos: '{:0.2f}'.format(np.sqrt(x)))
ax2.xaxis.set_major_formatter(formatter)

ax2.set_xlim(ax1.get_xlim())

plt.show()

However, as soon as you zoom or interact with the plot, you’ll notice that the axes are unlinked.

You could add an axes in the same position with both shared x and y axes, but then the tick formatters are shared, as well.

Therefore, the easiest way to do this is using a parasite axes.

As a quick example:

import numpy as np
import matplotlib.pyplot as plt
from matplotlib.ticker import FuncFormatter
from mpl_toolkits.axes_grid1.parasite_axes import SubplotHost

fig = plt.figure()
ax1 = SubplotHost(fig, 1,1,1)
fig.add_subplot(ax1)

ax2 = ax1.twin()

ax1.plot(range(10))

formatter = FuncFormatter(lambda x, pos: '{:0.2f}'.format(np.sqrt(x)))
ax2.xaxis.set_major_formatter(formatter)

plt.show()

Both this and the previous plot will look identical at first. The difference will become apparent when you interact (e.g. zoom/pan) with the plot.