[FIXED] Is there any way to change data from a matplotlib plot after it has been created and showed?

Issue

So I have a function that prints me some plot once called, and returns some other data. The code for the plot is this one

def somefunction(input):

     x = np.linspace(-5,5,100)
     fig, axs = plt.subplots(2,sharex=True)
     fig.suptitle("Some plots")

     axs[0].plot(x, x**2, "-b", label="square")
     axs[1].plot(x, x**3, "-y", label="cube")

     axs[0].set(ylabel="values")
     axs[1].set(xlabel="Timestamp (common)", ylabel="values")

     axs[0].legend()
     axs[1].legend()
 
     plt.show()
     

     return [1,2,3]

Now, what I want to do is to print this plot later again but with additional information. I thought about saving the figure created here as the output of the function. I tried to do this by adding this to the code:

def somefunction(input):

    x = np.linspace(-5,5,100)
    fig, axs = plt.subplots(2,sharex=True)
    fig.suptitle("Some plots")

    axs[0].plot(x, x**2, "-b", label="square")
    axs[1].plot(x, x**3, "-y", label="cube")

    axs[0].set(ylabel="values")
    axs[1].set(xlabel="Timestamp (common)", ylabel="values")

    axs[0].legend()
    axs[1].legend()

    plt.show()
    fig_out = fig

    return [1,2,3], fig_out

and then later I can just obtain the figure in the second component of the output of the function and change it as I want. Like:

figure = somefunction(input)[1]
#now perform any wanted changes in the plot and plot again
ax0 = figure.axes[0]
ax0.text(3, 8, 'New text updated in the figure', style='italic',
    bbox={'facecolor': 'red', 'alpha': 0.5, 'pad': 10})

plt.show()

This doesn't work. The figure is indeed, saved in the second component of the output, but it doesn't let me change anything about it. It's just there, and I can't change it, nor plot any changes made to the figure.

I also tried saving the axes instead of the figure, but same story. I can't seem to find a way to edit this plot after it was created. Is it even possible?

Solution

Your code is fine, if you don't try to show both plots at once. There are multiple options to solve it, for instance:

Option 1: show the plot at the end

from matplotlib import pyplot as plt

def somefunction(input):
    x = np.linspace(-5,5,100)
    fig, axs = plt.subplots(2,sharex=True)
    fig.suptitle("Some plots")

    axs[0].plot(x, x**2, "-b", label="square")
    axs[1].plot(x, x**3, "-y", label="cube")

    axs[0].set(ylabel="values")
    axs[1].set(xlabel="Timestamp (common)", ylabel="values")

    axs[0].legend()
    axs[1].legend()

    #plt.show() #<- DO NOT USE IT NOW

    return [1,2,3], fig

my_fig = somefunction(input)[1]
ax0 = my_fig.axes[0]
ax0.text(3, 8, 'New text updated in the figure', style='italic', bbox={'facecolor': 'red', 'alpha': 0.5, 'pad': 10})

plt.show()

Option 2: use block=False to indicate to wait until all figures are returned

def somefunction(input):
    x = np.linspace(-5,5,100)
    fig, axs = plt.subplots(2,sharex=True)
    fig.suptitle("Some plots")

    axs[0].plot(x, x**2, "-b", label="square")
    axs[1].plot(x, x**3, "-y", label="cube")

    axs[0].set(ylabel="values")
    axs[1].set(xlabel="Timestamp (common)", ylabel="values")

    axs[0].legend()
    axs[1].legend()

    plt.show(block=False) #<- USE BLOCK=FALSE

    return [1,2,3], fig

my_fig = somefunction(input)[1]
ax0 = my_fig.axes[0]
ax0.text(3, 8, 'New text updated in the figure', style='italic', bbox={'facecolor': 'red', 'alpha': 0.5, 'pad': 10})

plt.show()

Answered By - flyakite