23  Plotting guidelines

23.1 increase fontsize to legible sizes

import matplotlib.pyplot as plt
import matplotlib.colors as mcolors
import matplotlib
import numpy as np
import pandas as pd

Graph with default matplotlib values:

t = np.linspace(0, 10, 101)
y = np.sin(2.0*np.pi*t/3) + np.random.random(len(t)) + 4.0
fig, ax = plt.subplots()
ax.plot(t, y)
ax.set(title="This is a title",
       xlabel="time (days)",
       ylabel="price (US$)"
      );

You can use seaborn to easily change plot style and font size:

import seaborn as sns
sns.set_theme(style="ticks", font_scale=1.5)

t = np.linspace(0, 10, 101)
y = np.sin(2.0*np.pi*t/3) + np.random.random(len(t)) + 4.0
fig, ax = plt.subplots()
ax.plot(t, y)
ax.set(title="This is a title",
       xlabel="time (days)",
       ylabel="price (US$)"
      );

I recommend that you read seaborn’s Controlling figure aesthetics.

23.2 choose colors wisely

import math
import matplotlib.colors as mcolors
from matplotlib.patches import Rectangle


def plot_colortable(colors, *, ncols=4, sort_colors=True):

    cell_width = 212
    cell_height = 22
    swatch_width = 48
    margin = 12

    # Sort colors by hue, saturation, value and name.
    if sort_colors is True:
        names = sorted(
            colors, key=lambda c: tuple(mcolors.rgb_to_hsv(mcolors.to_rgb(c))))
    else:
        names = list(colors)

    n = len(names)
    nrows = math.ceil(n / ncols)

    width = cell_width * ncols + 2 * margin
    height = cell_height * nrows + 2 * margin
    dpi = 72

    fig, ax = plt.subplots(figsize=(width / dpi, height / dpi), dpi=dpi)
    fig.subplots_adjust(margin/width, margin/height,
                        (width-margin)/width, (height-margin)/height)
    ax.set_xlim(0, cell_width * ncols)
    ax.set_ylim(cell_height * (nrows-0.5), -cell_height/2.)
    ax.yaxis.set_visible(False)
    ax.xaxis.set_visible(False)
    ax.set_axis_off()

    for i, name in enumerate(names):
        row = i % nrows
        col = i // nrows
        y = row * cell_height

        swatch_start_x = cell_width * col
        text_pos_x = cell_width * col + swatch_width + 7

        ax.text(text_pos_x, y, name, fontsize=14,
                horizontalalignment='left',
                verticalalignment='center')

        ax.add_patch(
            Rectangle(xy=(swatch_start_x, y-9), width=swatch_width,
                      height=18, facecolor=colors[name], edgecolor='0.7')
        )

    return fig

When you plot with matplotlib, the default color order is the following. You can always specify a plot’s color by typing something like color="tab:red.

plot_colortable(mcolors.TABLEAU_COLORS, ncols=2, sort_colors=False);

You can write other words as color names, see below.

plot_colortable(mcolors.CSS4_COLORS)
plt.show()

This reminds me of this cartoon:

For almost all purposes, all these colors should be more than enough.

Be consistent!: if in one plot precipitation is blue and temperature is red, make sure you keep the same colors throughout your assignment.

Be mindful of blind-color people: A good rule of thumb is to avoid red and green shades in the same graph.

I’ll put a bunch of links below, this is for my own reference, but you are more than welcome to take a look.

• ColorBrewer
• Palettable
• xkcd colors
• Colormaps in Matplotlib

23.3 the best legend is no legend

t = np.linspace(0, 10, 101)
y1 = np.sin(2.0*np.pi*t/3) + np.random.random(len(t)) + 5.0
y2 = 0.7*np.sin(2.0*np.pi*t/1+1.0) + np.random.random(len(t)) + 2.0
y3 = 0.2*np.sin(2.0*np.pi*t/5+2.0) + 0.2*np.random.random(len(t)) + 3.5

fig, ax = plt.subplots(3, 1, figsize=(8,14))
fig.subplots_adjust(hspace=0.7)

# you can use legends
ax[0].plot(t, y1, color="darkblue", label="signal A")
ax[0].plot(t, y2, color="blue", label="signal B")
ax[0].plot(t, y3, color="xkcd:hot pink", label="signal C")
ax[0].set(title="you can have a legend",
          xlabel="time (days)",
          ylabel="signal (a.u.)"
         )
ax[0].legend()

# you can use an extra y axes
p1, = ax[1].plot(t, y1, color="darkblue")
ax[1].yaxis.label.set_color(p1.get_color())
ax[1].tick_params(axis='y', colors=p1.get_color())
ax[1].set(xlabel="time (days)",
          ylabel="signal A (a.u.)",
          title="you can have two y axes with different colors"
         )
ax1b = plt.twinx(ax[1])
p2, = ax1b.plot(t, y3, color="xkcd:hot pink", label="signal C")
ax1b.set(ylabel="signal B (a.u.)"
        )
ax1b.yaxis.label.set_color(p2.get_color())
ax1b.tick_params(axis='y', colors=p2.get_color())

# you can write directly on the graph
ax[2].plot(t, y1, color="darkblue", label="signal A")
ax[2].plot(t, y2, color="blue", label="signal B")
ax[2].plot(t, y3, color="xkcd:hot pink", label="signal C")
ax[2].set(xlabel="time (days)",
          ylabel="signal (a.u.)",
          ylim=[0.5,7],
          title="you can write directly on the graph"
         )
ax[2].text(-0.5, 5, "signal A", color="darkblue", ha="left")
ax[2].text(-0.5, 1, "signal B", color="blue", ha="left")
ax[2].text(-0.5, 4, "signal C", color="xkcd:hot pink", ha="left")

Text(-0.5, 4, 'signal C')

You can also make a colorbar to substitute a legend.

num_lines = 6

t = np.linspace(0, 2, 101)

# Get truncated colormap
cmap = plt.colormaps.get_cmap('jet')
bottom = 0.6; top = 1.0
truncated_cmap = mcolors.LinearSegmentedColormap.from_list("truncated_viridis", cmap(np.linspace(bottom, top, num_lines)))

# Create a figure and axis
fig, ax = plt.subplots()

# Plot the lines with increasing frequency
for i in range(num_lines):
    freq = i + 1
    y = np.sin(2.0 * np.pi * t * freq) + 2*freq
    ax.plot(t, y, color=truncated_cmap(i / (num_lines - 1)), label=f'Slope {slope}')

ax.set(xlabel="time (s)",
       ylabel="signal")

# Create a discrete colorbar
boundaries = np.linspace(0.5, num_lines + 0.5, num_lines + 1)
ticks = np.arange(num_lines) + 1
norm = mcolors.BoundaryNorm(boundaries, truncated_cmap.N)
sm = plt.cm.ScalarMappable(cmap=truncated_cmap, norm=norm)
sm.set_array([])  # fake up the array of the scalar mappable
cbar = plt.colorbar(sm, ticks=ticks, boundaries=boundaries, label='frequency', ax=ax)
cbar.ax.tick_params(which='both', size=0)
freq_list = [f"{x+1} Hz" for x in range(num_lines)]
cbar.set_ticklabels(freq_list)