Set axis tight only to x or y axis

Question

I have a plot look like this:

Obviously, the left and right side is a waste of space, so I set

plt.axis('tight')

But this gives me plot like this:

The xlim looks right now, but the ylim is too tight for the plot.

I'm wondering, if I can only set axis(tight) only to x axis in my case?

So the plot may look something like this:

It's certainly possible that I can do this manually by

plt.gca().set_xlim(left=-10, right=360)

But I'm afraid this is not a very elegant solution.

Still trying to figure out why they have weird behavior as the default here. — eric, Aug 20 '20 at 20:49
More generally, you may also want to show, e.g., a *subset* of the plotted xrange, and have the yrange be tight for that narrow (in x) view of what was plotted. Or vice versa. The accepted answer can be used for this too, ie. in general to set ytight or xtight independently (to use Matlab terminology)... — CPBL, Apr 09 '23 at 15:18

score 58 · Accepted Answer · edited May 24 '18 at 14:58

58

You want to use matplotlib's autoscale method from the matplotlib.axes.Axes class.

Using the functional API, you apply a tight x axis using

plt.autoscale(enable=True, axis='x', tight=True)

or if you are using the object oriented API you would use

ax = plt.gca()  # only to illustrate what `ax` is
ax.autoscale(enable=True, axis='x', tight=True)

For completeness, the axis kwarg can take 'x', 'y', or 'both', where the default is 'both'.

edited May 24 '18 at 14:58

Steven C. Howell

answered Jun 01 '16 at 06:50

Vadim Shkaberda

1

A small addition: I'm mostly using the object-oriented API, `autoscale` is actually a method of the `Axes` class. – Wisperwind Mar 17 '17 at 12:52
This does not seem to work with the y-axes. I tried `axis='y'` with no change. – Steven C. Howell Apr 16 '18 at 19:34
@StevenC.Howell Works fine for y-axes with matplotlib up to version 2.2.2 both on Linux (Python 2) and on Windows (Python 3, 32-/64-bit versions). – Vadim Shkaberda Apr 17 '18 at 21:56

eric · Answer 2 · 2021-01-11T16:25:44.457

I just put the following at the beginning of those scripts in which I know I'll want my xlims to hug my data:

import matplotlib.pyplot as plt
plt.rcParams['axes.xmargin'] = 0

If I decide to add some whitespace buffer to an individual plot in that same script, I do it manually with:

plt.xlim(lower_limit, upper_limit)

While the accepted answer works, and is what I used for a while, I switched to this strategy because I only have to remember it once per script.

2 Answers2