从 tkinter window 调用时使用 plt.close('all') 后 plt.show() 冻结

Question

The code is supposed to wait for user click, then it will close the plot and continue. This works great when not done from the tkinter loop, but if the plot function is invoked from tkinter, the program freezes until the tkinter window is closed.

So if you run this code, you will get a plot, then when you click on the plot, the plot will close, and program will display "Hello from: main program" which is what I want, but then, the tkinter window opens with a button. When you click the button, the plot will show again, but this time when you click inside the plot, the plot will close, but no text will be shown. Once you close the tkinter window, the text will display "hello from: tkinter."

This is not what I want, I want the text to display as soon as I click on the plot like on the first example. Can someone provide any guidance?

import tkinter as tk
from tkinter import *
from matplotlib import pyplot as plt

class patito():
    def __init__(self, location):
        self.x = [0,1,2,3,4,5,6]
        self.y = [1,2,1,2,1,2,1]

        self.fig, self.ax = plt.subplots()
        self.plot_this(location)
        

    def select_peak(self,event):
        plt.close("all")# close plot and continue to print("hello from:")

    def plot_this(self, location):
        plt.plot(self.x,self.y)
        self.fig.canvas.mpl_connect('button_press_event', self.select_peak)#call select_peak when clicking inside plot
        plt.show()
        print("hello from: " + location)

def call_patito(location):
    x = patito(location)

def tkinter_call_patito():
    x = patito("tkinter")
########################### Program starts Here ##############
    
call_patito("main program")
             
root = tk.Tk()
main_frame = Frame(root)
title_frame = Frame(root)

source_button =      Button(title_frame,justify=LEFT, text='Create patito object', command = tkinter_call_patito).grid(row=1, column = 1)
title_frame.pack()
main_frame.pack()
root.mainloop()

Answer 1

Ah, looks like I got it fixed by using the Qt5Agg backend for matplotlib. I tried WxAgg but didn't work for me.

import matplotlib
matplotlib.use('QT5Agg')