Bug report

Bug summary

Vertical annotations shrink subplots when a figure handle's tight_layout is set to True. The expectation is that annotations would overlap the subplots.

Looking at the documentation...

https://matplotlib.org/3.2.2/tutorials/intermediate/tight_layout_guide.html#caveats

tight_layout() only considers ticklabels, axis labels, and titles. Thus, other artists may be clipped and also may overlap

With that in mind, tight_layout should not consider Text annotations, and shouldn't adjust the subplot size when added.

Code for reproduction

import matplotlib.pyplot as plt
import numpy as np

x = np.linspace(0,np.pi*2,1000)
y = np.sin(x)
dy = np.cos(x)
dy2 = -np.sin(x)

for b in [False,True]:

    h, ax = plt.subplots(3,1,sharex = True)
    h.set_tight_layout(b)
    ax[0].plot(x,y)
    ax[1].plot(x,dy)
    ax[2].plot(x,dy2)

    ax[2].annotate(s = f'Annotation When Tight Layout is {b}', xy = (np.pi,-1), rotation = 'vertical')

Actual outcome

Expected outcome

In this case, tight_layout is set to FALSE just to see visually what it looks like with the text overlayed

Matplotlib version

• Operating system: CentOS Linux 7 (Core)
• Matplotlib version: 3.2.2
• Matplotlib backend (print(matplotlib.get_backend())): Qt5Agg
• Python version: 3.8.3
• Jupyter version (if applicable):
• Other libraries: numpy 1.18.5

conda

conda-forge