Common matplotlib snippets

• Creating figure(s)
• Setting axis style & width
• Constants
• Legend
• Saving the plot

Creating figure(s)

For single figure plot

fig = plt.figure(figsize=(20, 6))
# ...
plt.plot(...)

For multiple figure plot

fig, (plt1, plt2) = plt.subplots(
    1, # n rows
    2, # n columns
    figsize=(18, 6),
    gridspec_kw={
        'width_ratios': [1, 1], # 1:1 in this case
        # 👇 margin between subplots
        'wspace': 0.3,
        'hspace': 0.2
    }
)
# ...
plt.subplot(1, 2, 1) # start ploting on subplot at [0, 0]
# ...
plt.subplot(1, 2, 2) # start ploting on subplot at [0, 1]

Setting axis style & width

ax = plt.gca()
ax.spines['top'].set_linewidth(3)
ax.spines['right'].set_linewidth(3)
ax.spines['left'].set_linewidth(3)
ax.spines['bottom'].set_linewidth(3)

This can be set different in different subplots.

Constants

It is handy to set things like colors, line styles, font sizes as global constants, since it is common to modify the style of the figures multiple times.

H1 = 35
H2 = 30
colors = ['#324DC2', '#F3B194', '#58C8F5', '#6D738F', '#C24017']
hatches = ['x', '/', 'o', '\\', '.']
bar_width = 0.27

Legend

plt.legend(
    ax.get_legend_handles_labels()[0],
    ax.get_legend_handles_labels()[1],
    # 👆 upper lines can be manipulated to avoid repeating legend
    loc='upper center',
    fontsize=H2, # I prefer to set global constants for font size
    frameon=False, # no frame around legend
    ncol=2,
    handletextpad=0.2,
    columnspacing=4,
    borderpad=0.3,
)

Saving the plot

plt.savefig('./plot.pdf', bbox_inches='tight', dpi=300)