When I plot data using Matplotlib, the axes are always plotted by default as a box framing the plot. Let's say I am plotting data within axis limits -2 < x < 2 and -2 < y < 2, but I would like to draw axis lines inside this plot area through the origin, preferably with ticks and tick labels along these axis lines - not along the outer frame.
-
When asking us to solve a problem that requires writing code, it is best to first give some example code, include the plot and tell us what exactly you would like to change. Telling us what you have already tried and did not work would be nice, but not necessary.cel– cel2015-09-16 10:47:10 +00:00Commented Sep 16, 2015 at 10:47
-
3@cel: As someone who regularly deals with matplotlib, I think the question is quite clear.Hannes Ovrén– Hannes Ovrén2015-09-16 10:50:36 +00:00Commented Sep 16, 2015 at 10:50
Add a comment
|
2 Answers
This is well documented in the spines example (old link) / spine placement demo (new link).
You are going to turn off the right and top spines (e.g. spines['right'].set_color('none')), and move the left and bottom spines to the zero position (e.g. spines['left'].set_position('zero')).
import numpy as np
import matplotlib.pyplot as plt
fig = plt.figure()
x = np.linspace(-np.pi, np.pi, 100)
y = 2*np.sin(x)
ax = fig.add_subplot(111)
ax.set_title('zeroed spines')
ax.plot(x, y)
ax.spines['left'].set_position('zero')
ax.spines['right'].set_color('none')
ax.spines['bottom'].set_position('zero')
ax.spines['top'].set_color('none')
# remove the ticks from the top and right edges
ax.xaxis.set_ticks_position('bottom')
ax.yaxis.set_ticks_position('left')
Comments
I can at least give a half-complete answer. Yes, you can easily draw the axis lines. It is as simple as
plt.axvline(0)
plt.axhline(0)
The original axes will remain, but can be turned off with plt.axis('off').
It will also not give you any tick marks.
2 Comments
Thomas Arildsen
Thanks, that at least takes me some of the way. I will try to see if I can find a way of decorating those with ticks and labels.
Hannes Ovrén
I would suggest looking through the matplotlib documentation. There is quite a lot of things possible with matplotlib if you look past the
matplotlib.pyplot module, and instead check the "real" API. Drawing the coordinate axes is something that I have wanted to do myself several times, but never had the time to completely solve.