How to Change Space Between Subplots in Matplotlib
-
tight_layout()
Method to Change Matplotlib Space Between Subplots -
plt.subplots_adjust()
Method to Change Space Between Subplots in Matplotlib -
plt.subplot_tool()
Method to Change Space Between Subplots in Matplotlib -
Activate
constrained_layout=True
in Matplotlibsubplots
Function
We could use tight_layout()
, subplots_adjust()
and subplot_tool()
methods to change subplot size or space in Matplotlib. We can also improve space between Matplotlib space by setting constrained_layout=True
in the subplots()
function.
tight_layout()
Method to Change Matplotlib Space Between Subplots
The tight_layout()
method automatically maintains the proper space between subplots.
import numpy as np
import matplotlib.pyplot as plt
x = np.linspace(-3, 3, 100)
y1 = np.sin(x)
y2 = np.cos(x)
y3 = 1 / (1 + np.exp(-x))
y4 = np.exp(x)
fig, ax = plt.subplots(2, 2)
ax[0, 0].plot(x, y1)
ax[0, 1].plot(x, y2)
ax[1, 0].plot(x, y3)
ax[1, 1].plot(x, y4)
ax[0, 0].set_title("Sine function")
ax[0, 1].set_title("Cosine function")
ax[1, 0].set_title("Sigmoid function")
ax[1, 1].set_title("Exponential function")
fig.tight_layout()
plt.show()
If we didn’t use the tight_layout()
method, one row would overlap with the title of the next.
import numpy as np
import matplotlib.pyplot as plt
x = np.linspace(-3, 3, 100)
y1 = np.sin(x)
y2 = np.cos(x)
y3 = 1 / (1 + np.exp(-x))
y4 = np.exp(x)
fig, ax = plt.subplots(2, 2)
ax[0, 0].plot(x, y1)
ax[0, 1].plot(x, y2)
ax[1, 0].plot(x, y3)
ax[1, 1].plot(x, y4)
ax[0, 0].set_title("Sine function")
ax[0, 1].set_title("Cosine function")
ax[1, 0].set_title("Sigmoid function")
ax[1, 1].set_title("Exponential function")
plt.show()
plt.subplots_adjust()
Method to Change Space Between Subplots in Matplotlib
We can use the plt.subplots_adjust()
method to change the space between Matplotlib subplots.
import numpy as np
import matplotlib.pyplot as plt
x = np.linspace(-3, 3, 100)
y1 = np.sin(x)
y2 = np.cos(x)
y3 = 1 / (1 + np.exp(-x))
y4 = np.exp(x)
fig, ax = plt.subplots(2, 2)
ax[0, 0].plot(x, y1)
ax[0, 1].plot(x, y2)
ax[1, 0].plot(x, y3)
ax[1, 1].plot(x, y4)
ax[0, 0].set_title("Sine function")
ax[0, 1].set_title("Cosine function")
ax[1, 0].set_title("Sigmoid function")
ax[1, 1].set_title("Exponential function")
plt.subplots_adjust(left=0.125, bottom=0.1, right=0.9, top=0.9, wspace=0.2, hspace=0.35)
plt.show()
plt.subplots_adjust(left=0.125, bottom=0.1, right=0.9, top=0.9, wspace=0.2, hspace=0.35)
wspace
and hspace
specify the space reserved between Matplotlib subplots. They are the fractions of axis width and height, respectively.
left
, right
, top
and bottom
parameters specify four sides of the subplots’ positions. They are the fractions of the width and height of the figure.
plt.subplot_tool()
Method to Change Space Between Subplots in Matplotlib
This method launches a subplot tool window for a figure.
import numpy as np
import matplotlib.pyplot as plt
im1 = np.random.random((50, 50))
im2 = np.random.random((40, 50))
im3 = np.random.random((50, 40))
im4 = np.random.random((60, 50))
plt.subplot(221)
plt.imshow(im1)
plt.subplot(222)
plt.imshow(im2)
plt.subplot(223)
plt.imshow(im3)
plt.subplot(224)
plt.imshow(im4)
plt.subplot_tool()
plt.show()
It provides an interactive method for the user to drag the bar in the subplot_tool
to change the subplots’ layout.
Activate constrained_layout=True
in Matplotlib subplots
Function
constrained_layout
adjusts subplots and decorations automatically to fit them in the figure as best as possible.
constrained_layout
must be activated before or during subplot creation as it optimizes the layout before every drawing step.
import numpy as np
import matplotlib.pyplot as plt
a = np.linspace(0, 5, 100)
figure, axes = plt.subplots(2, 2, constrained_layout=True)
axes[0, 0].plot(x, np.exp(a))
axes[0, 1].plot(a, np.sin(a))
axes[1, 0].plot(a, np.cos(a))
axes[1, 1].plot(range(10))
axes[0, 0].set_title("subplot 1")
axes[0, 1].set_title("subplot 2")
axes[1, 0].set_title("subplot 3")
axes[1, 1].set_title("subplot 4")
plt.show()
Suraj Joshi is a backend software engineer at Matrice.ai.
LinkedIn