Can you "cache" matplotlib plots and show them dynamically?

Question

I want to create some matplotlib plots in batch, and then display them interactively, e.g something like this? (current code doesn't display the plots)

import matplotlib.pyplot as plt
from ipywidgets import interact
plots = {'a': plt.plot([1,1],[1,2]), 'b': plt.plot([2,2],[1,2])}

def f(x):
    return plots[x]

interact(f, x=['a','b'])  

Answer 1

Redraw figure

Possibly you want something like this, where the figure is cleared on each new selection and the artist in question is readded to the canvas.

%matplotlib notebook
import matplotlib.pyplot as plt
from ipywidgets import interact
plots = {'a': plt.plot([1,1],[1,2]), 'b': plt.plot([2,2],[1,2])}

def f(x):
    plt.gca().clear()
    plt.gca().add_artist(plots[x][0])
    plt.gca().autoscale()
    plt.gcf().canvas.draw_idle()

interact(f, x=['a','b']);

Result in a jupyter notebook:

Blitting

Unfortunately, the notebook backend currently does not support blitting. Using blitting, one could build the the plots beforehands and then just blit them to the axes. This could look like this:

import matplotlib.pyplot as plt
from matplotlib.widgets import RadioButtons

fig, ax = plt.subplots()
fig.subplots_adjust(left=0.18)
line1, = ax.plot([1,1],[1,2])
line2, = ax.plot([2,2],[1,2], color="crimson")
line2.remove()
fig.canvas.draw()
# store state A, where line1 is present
stateA = fig.canvas.copy_from_bbox(ax.bbox)

line1.remove()
ax.add_artist(line2)
fig.canvas.draw()
# store state B, where line2 is present
stateB = fig.canvas.copy_from_bbox(ax.bbox)


plots = {'a': stateA, 'b': stateB}
rax = plt.axes([0.05, 0.4, 0.1, 0.15])
check = RadioButtons(rax, ('a', 'b'), (False, True))

def f(x):
    fig.canvas.restore_region(plots[x])
    
check.on_clicked(f)

plt.show()

The above runs fine in an normal interactive figure. Once blitting is supported in the notebook backend in some future version of matplotlib, one could replace the RadioButtons and use interact in a notebook.