Problem with Matplotlib subplots with different number of rows and columns

Question

The code below creates a 4×4 grid of plots as expected:

import numpy as np
import matplotlib.pyplot as plt
x = np.arange(10)
rows, cols = 4, 4
_, axs = plt.subplots(rows, cols, figsize=(5,5))
for i in range(rows):
    for j in range(cols):
        axs[i,j].plot(x,x)

But things go wrong when I try to display subplots with 2 rows and 4 columns (same code with rows, cols = 2, 4):

When I would expect something like this:

If I try to add axs[i,j].set_aspect('equal', 'box') after each plot, I get:

Which is somehow better, but still leaves an empty line, just as if the default "tall rectangle" figure was leaving its footprint. Overall I have the impression that matplotlib wants its subplots to look "square-y".

Before I look into complicated layout matters, does anybody know about some straightforward way to come up with the result I’m expecting (which is not so crazy to expect, right?)

Asked By: Pierrick Bruneau

||

Source

Answer 1

You are decreasing the rows, but keeping the figure the same size. You would need to adjust the figsize. So since you cut the rows in half, cut the height in half too to stay proportional.

import numpy as np
import matplotlib.pyplot as plt
x = np.arange(10)
rows, cols = 2, 4
_, axs = plt.subplots(rows, cols, figsize=(5, 2.5))
for i in range(rows):
    for j in range(cols):
        axs[i,j].plot(x,x)

Answered By: chitown88

Answer 2

You should adjust the figsize:

_, axs = plt.subplots(rows, cols, figsize=(5, 5./2))

Answered By: Adrian Usler

Problem with Matplotlib subplots with different number of rows and columns

Question:

Answers: