About
This page showcases the work of Tuo Wang that introduces packages to make ggplot2 plots more beautiful. You can find the original R code on Tuo's blog here.
Thanks to him for accepting sharing his work here! Thanks also to Tomás Capretto who translated this work from R to Python! 🙏🙏
As a teaser, here is the plot we’re gonna try building:
Load libraries
The first step is always to import some libraries. NumPy is going to be used to make some computations with arrays and load_penguins() is imported from the palmerpenguins library to have access to the popular dataset.
import matplotlib.pyplot as plt
import numpy as np
from matplotlib.lines import Line2D
from palmerpenguins import load_penguinsLoad and prepare the data
Today's data were collected and made available by Dr. Kristen Gorman and the Palmer Station, Antarctica LTER, a member of the Long Term Ecological Research Network. This dataset was popularized by Allison Horst in her R package palmerpenguins with the goal to offer an alternative to the iris dataset for data exploration and visualization.
penguins = load_penguins()
penguins.head(3)| species | island | bill_length_mm | bill_depth_mm | flipper_length_mm | body_mass_g | sex | year | |
|---|---|---|---|---|---|---|---|---|
| 0 | Adelie | Torgersen | 39.1 | 18.7 | 181.0 | 3750.0 | male | 2007 |
| 1 | Adelie | Torgersen | 39.5 | 17.4 | 186.0 | 3800.0 | female | 2007 |
| 2 | Adelie | Torgersen | 40.3 | 18.0 | 195.0 | 3250.0 | female | 2007 |
After dropping observations with missing values, it's necessary to compute the mean value for the numerical variables that will be displayed in the radar chart. Then, with the aid of the rescale() function, these summaries are rescaled to [0,1].
penguins = penguins.dropna()def rescale(x):
return (x - np.min(x)) / np.ptp(x)penguins_radar = (
penguins.groupby('species').agg(
avg_bill_length = ("bill_length_mm", np.mean),
avg_bill_depth = ("bill_depth_mm", np.mean),
avg_flipper_length = ("flipper_length_mm", np.mean),
avg_body_mass = ("body_mass_g", np.mean)
)
.apply(lambda x: rescale(x))
.reset_index()
)Let's have a final look to the data before starting to code.
penguins_radar| species | avg_bill_length | avg_bill_depth | avg_flipper_length | avg_body_mass | |
|---|---|---|---|---|---|
| 0 | Adelie | 0.000000 | 0.978584 | 0.000000 | 0.000000 |
| 1 | Chinstrap | 1.000000 | 1.000000 | 0.210846 | 0.019422 |
| 2 | Gentoo | 0.873549 | 0.000000 | 1.000000 | 1.000000 |
Basic radar plot
As usually, colors and other important values are declared before the code that actually produces the plot.
BG_WHITE = "#fbf9f4"
BLUE = "#2a475e"
GREY70 = "#b3b3b3"
GREY_LIGHT = "#f2efe8"
COLORS = ["#FF5A5F", "#FFB400", "#007A87"]# The three species of penguins
SPECIES = penguins_radar["species"].values.tolist()
# The four variables in the plot
VARIABLES = penguins_radar.columns.tolist()[1:]
VARIABLES_N = len(VARIABLES)
# The angles at which the values of the numeric variables are placed
ANGLES = [n / VARIABLES_N * 2 * np.pi for n in range(VARIABLES_N)]
ANGLES += ANGLES[:1]
# Padding used to customize the location of the tick labels
X_VERTICAL_TICK_PADDING = 5
X_HORIZONTAL_TICK_PADDING = 50
# Angle values going from 0 to 2*pi
HANGLES = np.linspace(0, 2 * np.pi)
# Used for the equivalent of horizontal lines in cartesian coordinates plots
# The last one is also used to add a fill which acts a background color.
H0 = np.zeros(len(HANGLES))
H1 = np.ones(len(HANGLES)) * 0.5
H2 = np.ones(len(HANGLES))Radar charts plot data points in a circular layout. Instead of horizontal and vertical axes, it has an angular and a radial axis for x and y, respectively. In this world, x values are given by angles and y values are a distance from the center of the circle.
For today's visualization, x axis represents the variables, and y axis their scaled mean. The locations for x, given in angles, should be manually calculated and passed to matplotlib. That is what is going on in n / VARIABLES_N * 2 * np.pi in the list comprehension above. The y values are already given in the data frame.
Want to see how it looks like? Let's do it!
# Initialize layout ----------------------------------------------
fig = plt.figure(figsize=(14, 10))
ax = fig.add_subplot(111, polar=True)
fig.patch.set_facecolor(BG_WHITE)
ax.set_facecolor(BG_WHITE)
# Rotate the "" 0 degrees on top.
# There it where the first variable, avg_bill_length, will go.
ax.set_theta_offset(np.pi / 2)
ax.set_theta_direction(-1)
# Setting lower limit to negative value reduces overlap
# for values that are 0 (the minimums)
ax.set_ylim(-0.1, 1.05)
# Plot lines and dots --------------------------------------------
for idx, species in enumerate(SPECIES):
values = penguins_radar.iloc[idx].drop("species").values.tolist()
values += values[:1]
ax.plot(ANGLES, values, c=COLORS[idx], linewidth=4, label=species)
ax.scatter(ANGLES, values, s=160, c=COLORS[idx], zorder=10)
Customize guides and annotations
The plot above looks quite nice for a start. But so many lines and labels for the axes are unnecesary (and also boring!). Let's remove these defaults and improve this chart with more beautiful custom annotations and guides.
# Set values for the angular axis (x)
ax.set_xticks(ANGLES[:-1])
ax.set_xticklabels(VARIABLES, size=14)
# Remove lines for radial axis (y)
ax.set_yticks([])
ax.yaxis.grid(False)
ax.xaxis.grid(False)
# Remove spines
ax.spines["start"].set_color("none")
ax.spines["polar"].set_color("none")
# Add custom lines for radial axis (y) at 0, 0.5 and 1.
ax.plot(HANGLES, H0, ls=(0, (6, 6)), c=GREY70)
ax.plot(HANGLES, H1, ls=(0, (6, 6)), c=COLORS[2])
ax.plot(HANGLES, H2, ls=(0, (6, 6)), c=GREY70)
# Now fill the area of the circle with radius 1.
# This create the effect of gray background.
ax.fill(HANGLES, H2, GREY_LIGHT)
# Custom guides for angular axis (x).
# These four lines do not cross the y = 0 value, so they go from
# the innermost circle, to the outermost circle with radius 1.
ax.plot([0, 0], [0, 1], lw=2, c=GREY70)
ax.plot([np.pi, np.pi], [0, 1], lw=2, c=GREY70)
ax.plot([np.pi / 2, np.pi / 2], [0, 1], lw=2, c=GREY70)
ax.plot([-np.pi / 2, -np.pi / 2], [0, 1], lw=2, c=GREY70)
# Add levels -----------------------------------------------------
# These labels indicate the values of the radial axis
PAD = 0.05
ax.text(-0.4, 0 + PAD, "0%", size=16, fontname="Roboto")
ax.text(-0.4, 0.5 + PAD, "50%", size=16, fontname="Roboto")
ax.text(-0.4, 1 + PAD, "100%", size=16, fontname="Roboto")
fig
Final chart
Chart clutter has been drastically reduced and the plot above looks much better, that's great! But there's still room for improvement. This plot needs a good title, a legend to make the colors insightful as well as a fix for the overlap between labels and figures.
# Create and add legends -----------------------------------------
# Legends are made from scratch.
# Iterate through species names and colors.
# These handles contain both markers and lines.
handles = [
Line2D(
[], [],
c=color,
lw=3,
marker="o",
markersize=8,
label=species
)
for species, color in zip(SPECIES, COLORS)
]
legend = ax.legend(
handles=handles,
loc=(1, 0), # bottom-right
labelspacing=1.5, # add space between labels
frameon=False # don't put a frame
)
# Iterate through text elements and change their properties
for text in legend.get_texts():
text.set_fontname("Roboto") # Change default font
text.set_fontsize(16) # Change default font size
# Adjust tick label positions ------------------------------------
XTICKS = ax.xaxis.get_major_ticks()
for tick in XTICKS[0::2]:
tick.set_pad(X_VERTICAL_TICK_PADDING)
for tick in XTICKS[1::2]:
tick.set_pad(X_HORIZONTAL_TICK_PADDING)
# Add title ------------------------------------------------------
fig.suptitle(
"Radar Plot of Penguin Species",
x = 0.1,
y = 1,
ha="left",
fontsize=32,
fontname="Lobster Two",
color=BLUE,
weight="bold",
)
fig
