Testing event handling and picking with matplotlib

Question

As per subject, how to write tests for functions that deal with pick event handling in matplotlib?

In particular, given the following minimum working example, how to write a test that would offer 100% coverage?

import numpy as np
import matplotlib.pyplot as plt

def onpick(event):
    ind = event.ind
    print('you clicked on point(s):', ind)

def attach_handler_to_figure(figure):
    figure.canvas.mpl_connect('pick_event', onpick)

def main():
    plt.ion()
    x, y, c, s = np.random.rand(4, 100)
    fig, ax = plt.subplots()
    ax.scatter(x, y, 100*s, c, picker=True)
    attach_handler_to_figure(fig)

main()

The critical part is for me writing tests for the functions onpick and attach_handler_to_figure. Regarding the plotting I find this answer to be satisfying!

MORE INFO: I am not after a way of testing console output. What I am after are test functions, some sort of test_onpick and test_attach_handler_to_figure and test_main (well, in main the challenge is testing the line attach_handler_to_figure(fig)), that can be used by pytest or any other testing framework.

Is [Python: Write unittest for console print](https://stackoverflow.com/questions/33767627/python-write-unittest-for-console-print) what you're looking for? — ImportanceOfBeingErnest, Nov 02 '18 at 18:44
@ImportanceOfBeingErnest Not really, I'm after a way of simulating the click! But thanks for your input! — wizclown, Nov 02 '18 at 19:06

score 1 · Accepted Answer · answered Nov 02 '18 at 19:31

You can of course mock up a pick event. In the following I modified onpick to actually return someting. In order to test console output see Python: Write unittest for console print instead.

import numpy as np
import matplotlib.pyplot as plt

def onpick(event):
    ind = event.ind
    print('you clicked on point(s):', ind)
    return ind

def attach_handler_to_figure(figure):
    figure.canvas.mpl_connect('pick_event', onpick)

def main():
    #plt.ion()
    x, y, c, s = np.random.rand(4, 100)
    fig, ax = plt.subplots()
    ax.scatter(x, y, 100*s, c, picker=True)
    attach_handler_to_figure(fig)


def test_onpick():
    from unittest.mock import Mock

    main()

    event = Mock()
    event.ind = [2]

    ret = onpick(event)
    print(ret)
    assert ret == [2]

test_onpick()

Thank you for a great answer! The mock event is exactly the piece I was missing. — wizclown, Nov 02 '18 at 20:48

Testing event handling and picking with matplotlib

1 Answers1