Matplotlib Pan and Zoom only works for first instance - Not multiple and across sessions

[MarcSkovMadsen]

If I set interactive=True in the example in the Reference Gallery https://panel.holoviz.org/reference/panes/Matplotlib.html

The plot looks like below and there are errors in the console.

import panel as pn
pn.extension()
import numpy as np

from matplotlib.figure import Figure
from matplotlib import cm
# from matplotlib.backends.backend_agg import FigureCanvas  # not needed for mpl >= 3.1

Y, X = np.mgrid[-3:3:100j, -3:3:100j]
U = -1 - X**2 + Y
V = 1 + X - Y**2
speed = np.sqrt(U*U + V*V)

fig0 = Figure(figsize=(8, 6))
ax0 = fig0.subplots()
# FigureCanvas(fig0)  # not needed for mpl >= 3.1

strm = ax0.streamplot(X, Y, U, V, color=U, linewidth=2, cmap=cm.autumn)
fig0.colorbar(strm.lines)

mpl_pane = pn.pane.Matplotlib(fig0, dpi=144, interactive=True)
mpl_pane.servable()

Additional Context

There are some requests for interactive matplotlib. I.e. adding events. C.f. #2349. The question is if the current IPyWidgets based implementation should be fixed or a True Panel version of the interactive Matplotlib should be implemented?

[nghenzi]

it works for me. I think there is to add
pn.extension(sizing_mode="stretch_width")
or something like that. The matplotlib pane based on ipympl has the resize handle.

[MarcSkovMadsen]

Thanks @nghenzi . I can confirm that. Also to get the buttons look nice you need to add font awesome css.

import panel as pn
pn.extension()
import numpy as np

from matplotlib.figure import Figure
from matplotlib import cm
from matplotlib.backends.backend_agg import FigureCanvas  # not needed for mpl >= 3.1
FONTAWESOME_LINK = "https://cdnjs.cloudflare.com/ajax/libs/font-awesome/4.5.0/css/font-awesome.css"
pn.config.css_files.append(FONTAWESOME_LINK)
pn.extension(sizing_mode="stretch_width")

Y, X = np.mgrid[-3:3:100j, -3:3:100j]
U = -1 - X**2 + Y
V = 1 + X - Y**2
speed = np.sqrt(U*U + V*V)

fig0 = Figure(figsize=(8, 6))
ax0 = fig0.subplots()
FigureCanvas(fig0)  # not needed for mpl >= 3.1

strm = ax0.streamplot(X, Y, U, V, color=U, linewidth=2, cmap=cm.autumn)
fig0.colorbar(strm.lines)

mpl_pane = pn.pane.Matplotlib(fig0, dpi=144, interactive=True)
mpl_pane.servable()

[MarcSkovMadsen]

Unfortunately the interactivity is still not working. I cannot pan or zoom.

matplotlib_not_working.mp4

[MarcSkovMadsen]

matplotlib_not_working_2nd.mp4

import panel as pn
pn.extension()
import numpy as np

from matplotlib.figure import Figure
from matplotlib import cm
from matplotlib.backends.backend_agg import FigureCanvasAgg  # not needed for mpl >= 3.1
import matplotlib
FONTAWESOME_LINK = "https://cdnjs.cloudflare.com/ajax/libs/font-awesome/4.5.0/css/font-awesome.css"
pn.config.css_files.append(FONTAWESOME_LINK)
pn.extension(sizing_mode="stretch_width")

def get_mpl():
    Y, X = np.mgrid[-3:3:100j, -3:3:100j]
    U = -1 - X**2 + Y
    V = 1 + X - Y**2
    speed = np.sqrt(U*U + V*V)

    fig0 = Figure(figsize=(8, 6))
    ax0 = fig0.subplots()
    # FigureCanvasAgg(fig0)  # not needed for mpl >= 3.1

    strm = ax0.streamplot(X, Y, U, V, color=U, linewidth=2, cmap=cm.autumn)
    fig0.colorbar(strm.lines)

    return pn.pane.Matplotlib(fig0, dpi=144, interactive=True)

get_mpl().servable()

[MarcSkovMadsen]

I tried subscribing to click events as described here

def onclick(event):
        print('%s click: button=%d, x=%d, y=%d, xdata=%f, ydata=%f' %
            ('double' if event.dblclick else 'single', event.button,
            event.x, event.y, event.xdata, event.ydata))

    cid = fig0.canvas.mpl_connect('button_press_event', onclick)

But unfortunately it did not print anything.