First Steps with Matplotlib

Some days ago I was asked which graph library I usually use…well, none daily.

I’ve colleagues using

D3.js or
matplotlib

Since I’m neither a JavaScript guru nor a Python expert it was the cool XKCD style that hooked me to matplotlib: Pimp your Charts with the fancy XKCD Style.

Usually, I’m fine with the raw data/numbers in my daily work. Let’s pick a basic example: X-ray a Git repository.

When did I publish my last post Wrap Java Libraries as OSGi Bundles with Maven?

$ git log -1 --format=%cd --relative-date content/Wrap\ Java\ Libraries\ as\ OSGi\ Bundles\ with\ Maven.md
7 weeks ago

What’s the inception date of the Git powered version of {devops blog}?

$ git log $(git rev-list --max-parents=0 HEAD) --format=%cd --relative-date
4 years, 11 months ago

When did the last change hit the repo?

$ git log -1 --format=%cd --relative-date
3 weeks ago

Let’s double-check the saying: a picture is worth a 1000 words and get through the nice XKCD showcase step by step:

Step 1 - Hello XKCD World

Prepare the empty basic figure with two axes only.

with plt.xkcd():
    fig = plt.figure()
    ax = fig.add_axes((0.1, 0.2, 0.8, 0.7))
    ax.spines['right'].set_color('none')
    ax.spines['top'].set_color('none')
    ax.set_xticks([])
    ax.set_yticks([])

For more details about tuing the axes check matplotlib.axes.

Step 1 - Hello XKCD World

Step 2 - The Timeline

Add two ticks to the x-axis to get a grip on the Blog Events we want to visualize.

ax.xaxis.set_ticks_position('bottom')
ax.set_xticks([0, 240])
ax.set_xticklabels(['4 years, 11 months ago', '7 weeks ago'])
ax.set_xlim([-25, 260])

The time we cover is roughly 250 weeks, so we set xlim to a range of [-25, 260] to put our data in between.

Step 2 - The Timeline

Step 3 - Scatter the Data

We put the data in a human readable array and feed it via numpy.ndarray.T in the required format to the matplotlib.pyplot.scatter function like follows:

data = np.array([
    [0, 1],
    [240, 1],
    [250, 1],
])
x, y = data.T
plt.scatter(x,y)

Step 3 - Scatter the Data

Step 4 - Finishing Touches

With matplotlib.pyplot.text we are trying to make the figure a bit more self-explanatory…

plt.text(0, 1, 'migration from\nPebble to Pelican', withdash=True,
    dashdirection=1,
    dashpad=10,
    dashlength=25,
    dashpush=10,
    dashrotation=45)

Step 4 - Finishing Touches

Seems like time for another post 🤔- maybe about my First Steps with Matplotlib...?

Header cover Photo by Samuel Zeller