来自绘图对象的matplotlib子图

Question

I've got a series of functions that return three plot objects (figure, axis and plot) and I would like to combine them into a single figure as subplots. I've put together example code:

import matplotlib.pyplot as plt
import numpy as np

def main():

    line_fig,line_axes,line_plot=line_grapher()
    cont_fig,cont_axes,cont_plot=cont_grapher()

    compound_fig=plot_compounder(line_fig,cont_fig)#which arguments?

    plt.show()

def line_grapher():
    x=np.linspace(0,2*np.pi)
    y=np.sin(x)/(x+1)

    line_fig=plt.figure()
    line_axes=line_fig.add_axes([0.1,0.1,0.8,0.8])
    line_plot=line_axes.plot(x,y)
    return line_fig,line_axes,line_plot

def cont_grapher():
    z=np.random.rand(10,10)

    cont_fig=plt.figure()
    cont_axes=cont_fig.add_axes([0.1,0.1,0.8,0.8])
    cont_plot=cont_axes.contourf(z)
    return cont_fig,cont_axes,cont_plot

def plot_compounder(fig1,fig2):
    #... lines that will compound the two figures that
    #... were passed to the function and return a single
    #... figure
    fig3=None#provisional, so that the code runs
    return fig3

if __name__=='__main__':
    main()

It would be really useful to combine a set of graphs into one with a function. Has anybody done this before?

Answer 1

If you're going to be plotting your graphs on the same figure anyway, there's no need to create a figure for each plot. Changing your plotting functions to just return the axes, you can instantiate a figure with two subplots and add an axes to each subplot:

def line_grapher(ax):
    x=np.linspace(0,2*np.pi)
    y=np.sin(x)/(x+1)

    ax.plot(x,y)


def cont_grapher(ax):
    z=np.random.rand(10,10)

    cont_plot = ax.contourf(z)

def main():

    fig3, axarr = plt.subplots(2)
    line_grapher(axarr[0])
    cont_grapher(axarr[1])

    plt.show()


if __name__=='__main__':
    main()

Look into the plt.subplots function and the add_subplot figure method for plotting multiple plots on one figure.