pyplot & the Object-Oriented API

pyplot — Quick, Stateful

import matplotlib.pyplot as plt

plt.plot(x, y, label='revenue')
plt.title('Monthly Revenue')
plt.xlabel('Month')
plt.ylabel('USD')
plt.legend()
plt.savefig('chart.png', dpi=300)
plt.show()

# pyplot tracks an implicit "current axes" — fine for one quick plot,
# ambiguous once multiple figures/subplots are involved

Object-Oriented — Explicit, Scales Better

fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(12, 5))

axes[0].plot(x, revenue)
axes[0].set_title('Revenue')

axes[1].scatter(x, users, c=values, cmap='viridis')
axes[1].set_title('Users')

fig.tight_layout()  # prevents titles/labels from overlapping
fig.savefig('report.png')

# ax.plot()/ax.set_title() are UNAMBIGUOUS about which subplot they
# affect — the recommended approach beyond a single throwaway plot

Common Chart Types

plt.hist(data, bins=30)
plt.bar(categories, values)
plt.scatter(x, y, c=values, cmap='viridis', s=sizes)

plt.style.use('ggplot')  # global preset styling for everything after this call