python - Matplotlib returning a plot object

Question

I have a function that wraps pyplot.plt so I can quickly create graphs with oft-used defaults:

def plot_signal(time, signal, title='', xlab='', ylab='',
                line_width=1, alpha=1, color='k',
                subplots=False, show_grid=True, fig_size=(10, 5)):

    # Skipping a lot of other complexity here

    f, axarr = plt.subplots(figsize=fig_size)
    axarr.plot(time, signal, linewidth=line_width,
               alpha=alpha, color=color)
    axarr.set_xlim(min(time), max(time))
    axarr.set_xlabel(xlab)
    axarr.set_ylabel(ylab)
    axarr.grid(show_grid)

    plt.suptitle(title, size=16)
    plt.show()

However, there are times where I'd want to be able to return the plot so I can manually add/edit things for a specific graph. For example, I want to be able to change the axis labels, or add a second line to the plot after calling the function:

import numpy as np

x = np.random.rand(100)
y = np.random.rand(100)

plot = plot_signal(np.arange(len(x)), x)

plot.plt(y, 'r')
plot.show()

I've seen a few questions on this (How to return a matplotlib.figure.Figure object from Pandas plot function? and AttributeError: 'Figure' object has no attribute 'plot') and as a result I've tried adding the following to the end of the function:

• return axarr

• return axarr.get_figure()

• return plt.axes()

However, they all return a similar error: AttributeError: 'AxesSubplot' object has no attribute 'plt'

Whats the correct way to return a plot object so it can be edited later?

Answer 1

I think the error is pretty self-explanatory. There is no such thing as pyplot.plt, or similar. plt is the quasi standard abbreviated form of pyplot when being imported, i.e. import matplotlib.pyplot as plt.

Concerning the problem, the first approach, return axarr is the most versatile one. You get an axes, or an array of axes, and can plot to it.

The code may look like

def plot_signal(x,y, ..., **kwargs):
    # Skipping a lot of other complexity her
    f, ax = plt.subplots(figsize=fig_size)
    ax.plot(x,y, ...)
    # further stuff
    return ax

ax = plot_signal(x,y, ...)
ax.plot(x2, y2, ...)
plt.show()