Matplotlib: Making axes fit shape limits

I'm trying to draw a rectangle in matplotlib using the following code:

import matplotlib.pyplot as plt
import matplotlib.patches as patches

fig, ax = plt.subplots()
width = 20
height = 10

rect = patches.Rectangle((0,0),width, height, linewidth=4,edgecolor='r',facecolor='none')
ax.add_patch(rect)

plt.show()

Which results in:

The axes do not fit the rectangle limits in this case. I could solve it with:

import matplotlib.pyplot as plt
import matplotlib.patches as patches

fig, ax = plt.subplots()
width = 20
height = 10

ax.set_xlim(0,width)
ax.set_ylim(0,height)

rect = patches.Rectangle((0,0),width, height, linewidth=4,edgecolor='r',facecolor='none')
ax.add_patch(rect)

plt.show()

This gives me the following picture which solves the problem in this case:

However, as I am trying to plot many rectangles and other shapes in the same figure, I need a way that matplotlib smartly determines the proper axes limits itself, like the way it does when plotting normal diagrams.

You are looking for .autoscale(). You may use .margins(0) to remove any extra space that is added by default.

I.e.

import matplotlib.pyplot as plt
import matplotlib.patches as patches

fig, ax = plt.subplots()
width = 20
height = 10

rect = patches.Rectangle((0,0),width, height, linewidth=4,edgecolor='r',facecolor='none')
ax.add_patch(rect)
ax.margins(0)
ax.autoscale()

plt.show()